CEREBRAL ASYMMETRIES IN SENSORY AND PERCEPTUAL PROCESSING
ADVANCES IN PSYCHOLOGY 123 editors."
G. E. STELMACH R A. VROON
ELSEVIER Amsterdam
- Lausanne
- New
York
- Oxford
- Shannon
- Singapore
- Tokyo
CEREBRAL ASYMMETRIES IN SENSORY AND PERCEPTUAL PROCESSING
Edited by
Stephen CHRISTMAN Department of Psychology University of Toledo Toledo, Ohio, USA
1997
ELSEVIER Amsterdam
- Lausanne
- New York - Oxford.
Shannon
- Singapore
- Tokyo
NORTH-HOLLAND ELSEVIER SCIENCE B.V. Sara Burgerhartstraat 25 EO. Box 211, 1000 AE Amsterdam, The Netherlands
ISBN: 0 444 82510X 9 1997 Elsevier Science B.V. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the publisher, Elsevier Science B.V., Copyright & Permissions Department, RO. Box 521, i 000 AM Amsterdam, The Netherlands. Special regulations for readers in the U.S.A. - This publication has been registered with the Copyright Clearance Center Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923. Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the U.S.A. All other copyright questions, including photocopying outside of the U.S.A., should be referred to the copyright owner, Elsevier Science B.V., unless otherwise specified. No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. This book is printed on acid-free paper. Printed in The Netherlands
Table of Contents Preface
xi
Contributors
xviii
SECTION I: SPATIAL/TEMPORAL FREQUENCY PROCESSING
1. Hemispheric Asymmetry in the Processing of Spatial Frequency: Experiments Using Gratings and Bandpass Filtering Stephen Christman
Sinusoidal and Square-wave Stimuli 7 Compound Stimuli 12 Low-pass and Band-pass Filtered Stimuli Conclusions 23 References 26
19
2. Temporal Frequency Processing 31
Luciano Mecacci
Hemispheric Asymmetries of Spatio-temporal Interaction: Electrophysiological Evidence 35 Reading Disability and Impairment in Processing Basic Spatio-temporal Information 44 Evidence from Brain-injured Patients 46 Conclusion 49 References 49
3. Interhemispheric Transfer of Spatial and Temporal Frequency Information 55
Nicoletta Berardi and Adriana Fiorentini
Properties of Interhemispheric Commisures in Mammals Interactions between Sinusoidal Stimuli Presented in the Left or Right Visual Field 61
56
vi Discrimination of Spatial Phase in Complex Gratings Presented in the Left or Right Visual Field 67 Interhemispheric Transfer of Information on Chromatic Contrast 73 Discussion 74 References 76
SECTION II: OBJECT AND SPATIAL REPRESENTATIONS 4. Hemispheric Asymmetry for Components of Spatial Processing Joseph Hellige 83 The Categorical/Coordinate Distinction 84 The Search for Underlying Mechanisms of Hemispheric Asymmetry for Spatial Processing 88 The Speech/Attention-Shift Hypothesis 89 Are Categorical and Coordinate Spatial Relationships Processed Independently? 91 The Nature of Task-Relevant Visual Information 93 Extensions of the Categorical/Coordinate Distinction 112 Concluding Comments: More on Mechanisms and Future Directions 116 Notes 119 References 120
5. Computational Analyses and Hemispheric Asymmetries in Visual-Form Recognition Chad Marsolek and E. Darcy Burgund Visual Form Subsystems 126 Behavioral Evidence for Relatively Independent Subsystems Contradictory Internal Processing Strategies 13:5 Behavioral Evidence for Parts-based versus Holistic Processing 147 Conclusions and Implications 150 Acknowledgments 153 References 153
125
131
vii SECTION III: VISUAL ATTENTION
6. Amplification of Spatial Nonuniformities by Guided Search Mechanisms E. William Yund
161
162 Visual Search and the Guided Search Model 166 Spatial Nonuniformities in Visual Search General Discussion 184 Conclusions 190 Footnotes 190 Acknowledgements 192 References 193
7. Hemispheric Coordination of Spatial Attention James Enns and Alan Kingstone
197
Hemispheric Specialization in Visual Search? 199 106 Hemifield Differences in Unilateral vs. Bilateral Visual Displays Hemifield Competition in Object Identification 216 Discussion 220 What are the Implications for Understanding Spatial Attention? 226 228 Implications for Understanding Hemispheric Specialization Acknowledgments 229 References 229
8. Asymmetries in the Flanker Compatibility Effect Frederick Kitterle, Mark Ludorf, and Jeremy Moreland Expt. 1: Left-right Asymmetries in the FCE: M and W Letter Arrays 236 Experiment 1B - FCE with H, V letter arrays 243 Experiment 2 - Effects of letter case 245 Experiment 3 - Target-Flanker Spacing 250 General Discussion 252 References 258
233
viii SECTION IV: EFFECTS OF VISUAL FIELD LOCUS
9. The Relation Between Left-Right and Upper-Lower Visual Field Asymmetries Stephen Christman and Christopher Niebauer Simple Reaction Time 266 Resolution/Acuity 267 Local-Global Processing 269 Categorical/Coordinate Processing Stereopsis 272 Motion 274 Visual Search 276 Visual Attention 279 Pattern Recognition 281 Conclusions 283 References 290
263
270
SECTION V: AUDITORY PROCESSING
10. Hemispheric Specialization of Human Auditory Processing: Perception of Speech and Musical Sounds Robert Zatorre Phonetic Mechanisms in Speech Perception 301 Processing of Melodic Patterns 307 Auditory Imagery 312 Morphometry of Auditory Cortex via Structural MRI References 319
11. Perceptual and Cognitive Development: Electrophysiological Correlates Dennis Molfese and Dana Narter Voice Onset Time 328 Place of Articulation 336 Vowel Sounds 341
299
316
325
ix Electrophysiological Correlates of Infant Memory 342 Electrophysiological Correlates of Early Word Acquisition Acknowledgments 374 References 374
356
12. The Ipsilateral Auditory Pathway: A Psychobiological Perspective Kendall Hutson 383 Anatomy of the Ascending Auditory System Role of Ipsilateral Pathway in Behavior 405 Evoked Potential Studies 414 Role of Ascending Pathways in Physiology of the Inferior Colliculus 415 Consequences to Cognition 439 Conclusions 441 Footnote 442 Acknowledgment 443 References 443
385
SECTION VI: TACTUAL PROCESSING
13. Role of Sensory and Post-sensory Factors in Hemispheric Asymmetries in Tactual Perception Jo~'l Fagot, Agn~s Lacreuse, and Jacques Vauclair Anatomical Bases of Tactual Perception 470 Functional Asymmetries for Elementary Tactile Discriminations 471 Tactual Discrimination of Orientations 474 Retention of Sequence of Touches 475 Tactual Discrimination of Dot Patterns 475 Tactual Maze Learning 477 Haptic Discrimination of Spatial Forms 477 Exploratory Strategies for Nonsense Shape Discrimination Haptic Perception in Nonhuman Primates 483 General Discussion 485 References 488
469
478
SECTION VII: OLFACTORY PROCESSING 14. Laterality in Human Nasal Chemoreception Richard Dory, S. Bromly, P. Moberg, and T. Hummel
497
Anatomy of the Olfactory and Trigeminal Chemosensory Systems 499 Olfactory System 501 Trigeminal System 508 The Search for Anatomical Asymmetries in Brain Regions Related to Olfaction 510 The Search for Functional Asymmetry in Human Olfactory Pathways 511 Conclusions 527 Footnotes 530 Acknowledgements 531 References 531
Name Index
543
Subject Index
557
xi
Preface Since Justine Sergent (1982a) first proposed that the left versus right hemispheres were specialized for the processing of high versus low spatial frequencies, respectively, laterality researchers have increasingly come to recognize the importance of sensory and perceptual factors in determining observed patterns of hemispheric asymmetry. As Sergent and Joseph Hellige noted in a seminal 1986 paper, this growing realization mirrored comparable trends in mainstream cognitive research during the early 1970's; they quote Garner's (1970) comments: "too often has the nature of input been ignored, with the consequence of incorrect assessment of information processing at worst, or an inadequate picture at best". In important ways, laterality research was "catching up" with the rest of the field. In a similar vein, Hardyck (1986) argued that sensory and perceptual factors associated with lateral tachistoscopic presentation of input had come to "constitute a set of 'invisible effects' present in many experiments, but unanalyzable due to omnipresence across experiments" (p.226). While higher-order functions representing the end products of information processing (e.g., word and face recognition) have traditionally constituted the dominant focus of laterality research, it is apparent that a full account of cerebral lateralization needs to also consider the role of earlier information processing stages. An observer's ability to attend to, recognize, and remember material is a priori constrained by limitations in the ability to sense and perceive; sensory and perceptual processes serve as the "gateways" to higher order processing. Concerns with such lower-level factors are critical given the widespread use of lateral, tachistoscopic presentation in visual laterality work, which means that input processing is almost always data-limited, and the effects of even small hemispheric differences in sensory processing can potentially modulate the strength and direction of asymmetries in higher-level processes [the notion that small initial differences in sensory functions can "snowball" into functional asymmetries of considerable magnitude also plays a prominent role in developmental accounts of the origin of hemispheric asymmetries proposed by Kosslyn (1987) and Previc (1994)]. Thus, researchers interested specifically in higher-order processes will naturally be inclined to interpret any obtained hemispheric asymmetries in terms of precisely those higher-order processes of interest; to the extent, however, that such asymmetries are being partly or wholly determined by perceptual variables, researchers run the risk of reaching mistaken conclusions.
xii The purpose of this book is to provide a comprehensive overview of hemispheric differences in sensory and perceptual processing. Accordingly, the first section of this book deals directly with the intraand inter-hemispheric processing of spatial and temporal frequencies in the visual modality. Chapters by Christman and by Mecacci provide reviews of spatial and temporal frequency processing, respectively, while Berardi and Fiorentini describe constraints on the interhemispheric transfer of basic sensory information. The three chapters of this section may seen as an extension of two previous papers: Sergent and Hellige's 1986 paper, "Role of Input Factors in Visual-Field Asymmetries", and Christman's 1989 literature review, "Perceptual Characteristics in Visual Laterality Research" (both in the journal Brain and Cognition). These papers presented timely reviews of the influence of various visual input factors (e.g., exposure duration, size, eccentricity, luminance, etc.) on hemispheric processing. Since they were written, however, a substantial body of empirical research investigating sensory processing in the hemispheres has accumulated, providing more powerful and direct tests than could be provided by the necessarily post-hoc nature of the aforementioned review articles. While the first section focuses on the "raw" input to higher-order mechanisms, the second section addresses the initial interaction between sensory and cognitive mechanisms, dealing with how the left and right cerebral hemispheres differ in their computation and representation of sensory information: Hellige provides an overview of hemispheric differences in spatial representations, while Marsolek and Burgund present an important new theory of hemispheric differences in visual-form recognition that has roots in the distinctions between categorical and coordinate spatial representations discussed by Hellige. A key theme in both chapters is that extraction of sensory information from input is guided and constrained by perceptual goals; that is, hemispheric asymmetries are determined conjointly by the sensory information available in the input and by the types or ranges of sensory information that are required by the task at hand. The third section covers how attentional mechanisms modulate the nature of perceptual processing in the cerebral hemispheres. Processing of specific objects occurs, not in isolation, but in a rich context defined both by other objects in the environment and by internal expectations and goals on behalf of the observer. The cerebral basis for such phenomena as the attentional and grouping processes underlying local-global processing
xiii and visual search through many-element displays is addressed in the chapters by Yund and by Enns and Kingstone. In particular, Yund presents an extension of the Guided Search Model (Wolfe, 1994) to visual field differences in visual search, while Enns and Kingstone present a theory based on interhemispheric competition for limited attentional resources. Kitterle, Ludorf, and Moreland discuss a related phenomenon in the form of the "flanker" effect, reporting visual field differences in the effects of distractors on the processing of targets. Section four consists of a single chapter presenting a theme that does not fit tidily into any of the other sections. Namely, the chapter by Christman and Niebauer reviews evidence suggesting a functional linkage between upper and right visual field processing, on the one hand, and lower and left visual field processing on the other. Their chapter offers a challenege to the interpretation of lateral field differences as reflecting hemispheric differences. That is, to the extent that upper/lower field differences may not be directly interpretable in terms of retinal projection to different hemispheres, the question is raised whether the corresponding left/right differences may also reflect something beyond hemispheric differences as such. Conversely, it is also possible that upper/lower differences reflect hemispheric attentional biases along the vertical meridian; this would be consistent with previous demonstrations that manifestations of hemispheric asymmetries are not necessarily linked to retinal coordinates (e.g., Luh, Rueckert, & Levy, 1991). In any case, this chapter indicates that laterality researchers (who are interested in left-right differences) and vision researchers (who may be interested in the functional differences between near versus far vision as associated with the lower versus upper visual fields, respectively) cannot provide a complete picture without integrating their various approaches. Although vision represents the dominant sensory modality in humans, the other senses are important to consider, both in their own right and insofar as they play a role in rich, polymodal object representations. Consequently, the remaining sections cover sensory and perceptual level processing in other sensory modalities. First, the chapter by Zattore provides an overview of cerebral asymmetries in auditory processing, with special emphasis on the processing of speech and music and on auditory imagery. While research on cerebral differences in auditory processing in adults has made significant strides via the use of dichotic listening paradigms employing verbal and nonverbal material, theoretical interpretations of this work have tended to focus on higher-level cognitive
xiv mechanisms (e.g., lexical and semantic processes). Relatively less attention has been paid to the sensory and neural bases of such asymmetries, which are addressed in the chapter by Zatorre. Molfese and Narter review evoked potential studies of auditory processing, with special emphasis on research with infants and children. This focus on developmental issues is especially fitting in the context of audition, since auditory processing is more closely and exclusively linked to higher-order processes (namely, speech perception and language) than the other sensory modalities. In this sense, research on infants provides a purer picture of the sensory underpinnings of phonetic processing (e.g., without the obscuring effects of top-down influences). On a more cautionary note, Hutson's chapter reviews the neural bases for the representation of contralateral auditory hemispace in each hemisphere, with special reference to the common assumption that left versus right ear performance reflects right versus left hemisphere processing, respectively. The importance of attentional biases in overriding or obscuring structural differences in dichotic listening has been previously pointed out (e.g., Mondor & Bryden, 1992); Hutson offers evidence that even the structural linking of left ear-fight hemisphere and right ear-left hemisphere may be on tenuous ground in light of the lack of an orderly structural chiasm in the auditory system. Finally, the present book covers hemispheric processing of tactual/ haptic and olfactory stimulation; while relatively less research has been devoted to these modalities (with the exception, of course, of studies of hand dominance), their inclusion provides a breadth of treatment complementary to the depth of treatment of the visual and auditory modalities. The chapter by Fagot, Lacreuse, and V auclair provides a balanced review of processing in the tactile/haptic domain, the richness of which is often obscured by the emphasis on right-handed writing in humans; their chapter also includes data from nonhuman primate populations. Finally, Doty, Bromly, Moberg, and Hummel offer a thorough review of the neural bases for olfactory and chemosensory processing, the oldest of our senses and perhaps the most neglected in terms of research (I am aware of no previous review of cerebral asymmetries in chemoreception). The treatment of varied sensory and perceptual level asymmetries of hemispheric function in this volume is hoped to serve as a useful reference tool for laterality researchers interested in sensory level processing p e r se, as well as for those researchers focusing on higher-level processes who want to address the possible influence on such processes of
XV
lower-level asymmetries of hemispheric function. At best, it is hoped that such a compendium will shed light on possible analogs among hemispheric asymmetries in different sensory modalities; for example, some of the findings concerning asymmetries in the visual and auditory domains suggest a general hemispheric difference in processing higher versus lower resolution sensory information. In addition, a detailed consideration of sensory level asymmetries may serve to foster links between research into asymmetries in humans versus nonhumans. Functions such as language and face recognition do not exist in any directly comparable form in non-humans (with the possible exception of Great Apes); research into sensory processes in the left and right hemispheres promises to provide potentially important bridges between human and nonhuman laterality research. In the context of the preceding discussion contrasting the roles of cognitive processes versus sensory processes, it is worth noting that the distinction between higher-order and lower-order processing is somewhat artificial. That is, while at least some sensory processing must logically precede cognitive processing in the temporal domain (although there is extensive overlap), this does not necessarily entail any sharp qualitative demarcation between the contents and operations of early versus late information processing stages. There is a growing consensus among cognitive psychologists that basic principles of sensory and perceptual processing form the foundation of cognitive processes such as language, memory, and categorization. For example, Chatterjee, Maher, and Heilman (1995) argued that the assignment of thematic roles of agent versus patient (which map onto the grammatical categories of subject versus object) may be based on nonlinguistic, spatiotemporal representations. With regard to memory, Roediger, Weldon, and Challis (1989) reviewed evidence for the importance of "data-driven" processing in memory in which perceptual characteristics of both input and retrieval cues play critical roles in memory encoding and retrieval. Finally, Barsalou (1993) has developed a comprehensive model of human categorization and knowledge representation that is firmly grounded in a compositional system of perceptual symbols. It is worth noting that Barsalou's scheme in which simple perceptual symbols can be flexibly and recursively combined to form ever more elaborate representations bears more than a passing similarity to Corballis' (1991) "Generative Assembling Device".
xvi In closing, perhaps the ultimate aim of this book is to foster greater interaction and integration between neuropsychological and mainstream cognitive research. One of more attractive features of Sergent's initial formulation of the spatial frequency hypothesis was that it took a welldeveloped body of research on the visual processing of spatial frequency information and placed it in the context of laterality research; too often, laterality researchers have tended to create idiosyncratic accounts of hemispheric differences in function that lack operational definitions (e.g., the analytic-holistic dichotomy). It is hoped that a continued effort to ground laterality research in the empirical and theoretical findings gleaned from over 100 years of experimental psychology will be of benefit to both areas.
Acknowledgements I acknowledge the Department of Psychology and the College of Arts & Sciences at the University of Toledo for granting a sabbatical during which I wrote my two chapters and for providing a generous level of institutional support over the years. I would also like to thank the editors of North-Holland's Advances in Psychology series, Kees Michelson and David Hoole, for their patience and encouragement. The proofs for this book were prepared on a Power Macintosh 7100/80AV, using Microsoft Word 5. la; final preparation of figures was done using Aldus SuperPaint, v.3.0. I would like to thank Kathy Skurzewski for assistance in scanning figures. Special thanks are due my wife, Lori, for help in proofing some of the chapters and, more importantly, for putting up with my incessant work and worry. Finally, I wish to thank my children, Rayna and Sam, for tolerating (without too much rivalry) a temporary third "child" in the family in the form of this book. This book is dedicated to the memory of Justine Sergent. She published her first series of papers (e.g., Sergent, 1982a, 1982b, 1982c, 1982d) presenting the spatial frequency hypothesis the year that I entered graduate school and began pursuing her work. She was a truly exceptional source of inspiration for myself and the neuropsychology community as a whole, and she is missed. Stephen Christman Toledo, 1997
xvii References
Barsalou, L. (1993). Flexibility, structure, and linguistic vagary in concepts: Manifestations o f compositional system of perceptual symbols. In A.C. Collins, S.E. Gatlaercole, M.A. Conway, and P.E.M. Morris (Eds.), Theories of Memory. Hillsdale, NJ: Lawrence Erlbaum Assoc. Chatterjee, A., Maher, L., & Heilman, K. (1995). Spatial characteristics of thematic role representation. Neuropsychologia, 33, 643-648. Christman, S. (1989). Perceptual characteristics in visual laterality research. Brain and Cognition, 11, 238-257. Corballis, M. (1991). The Lopsided Ape: Evolution of the Generative Mind. New York: Oxford University Press. Garner, W.R. (1970). The stimulus in information processing. American Psychologist, 25, 350-358. Hardyck, C. (1986). Cerebral asymmetries and experimental parameters: Real differences and imaginary variations? Brain and Cognition, 5, 223-239. Kosslyn, S.M. (1987). Seeing and imagining in the cerebral hemispheres: A computational approach. Psychological Review, 94, 148-175. Luh, K.E., Rueckert, L.M., & Levy, J. (1991). Perceptual asymmetries for free viewing of several types of chimeric stimuli. Brain and Cognition, 16, 83-103. Mondor, T.A., & Bryden, M.P. (1992). On the relation between auditory spatial attention and auditory perceptual asymmetries. Perception & Psychophysics, 52, 393-402. Roedlger,-H., Weldon, M., & Challis, B. (1989). Explaining dissociations between implicit and explicit measures of retention: A processing account. In H. Roediger & F.I.M. Craik (Eds.), Varieties of Memory and Consciousness: Essays in Honour of Endel Tulving. Hillsdale, NJ: Lawrence Erlbaum Associates. Sergent, J. (1982a). The cerebral balance of power: Confrontation or cooperation? Journal of Experimental Psychology: Human Perception and Performance, 8, 253-272. Sergent, J. (1982b). About face: Left-hemisphere involvement in processing physiognomies. Journal of Experimental Psychology: Human Perception and Performance, 8, 1-14. Sergent, J. (1982c). Influence of luminance on hemispheric processing. Bulletin of the Psychonomic Society, 20, 221-223. Sergent, J. (1982d). Theoretical and methodological consequences of variations in exposure duration in laterality studies. Perception & Psychophysics, 3-1, 451-461. Sergent, J., & Hellige, J. (1986). Role of input factors in visual-field asymmetries. Brain and Cognition, 5, 174-199 Wolfi~, J. M. (1994). "Guided search" 2.0: A revised model of visual search. Psychonomic Bulletin & Review, 1,202-238.
xviii
Contributors Nicoletta Berardi Istituto di Neurofisiologia C.N.R. Pisa, Italy
Frederick Kitterle Department of Psychology Northern Illinois University
Steven Bromley Department of Otorhinolaryngology Univ. of Pennsylvania Medical Center
Agnrs Lacreuse Department of Psychology University of Georgia
E. Darcy Burgund Department of Psychology University of Minnesota
Mark Ludorf Department of Psychology Stephen F. Austin State University
Stephen Christman Department of Psychology University of Toledo
Chad Marsolek Department of Psychology University of Minnesota
Richard L. Doty Department of Otorhinolaryngology Univ. of Pennsylvania Medical Center
Luciano Mecacci Universit~ degli Studi di Firenze Italy
James Enns Department of Psychology University of British Columbia
Paul Moberg Department of Otorhinolaryngology Univ. of Pennsylvania Medical Center
Jorl Fagot Center for Research in Cognitive Neuroscience, Marseille, France
Jeremy Moreland Department of Psychology Stephen F. Austin State University
Adriana Fiorentini Dipartimento di Psicologia Generale Universita' di Firenze
Dennis Molfese Department of Psychology Southern Illinois University, Carbondale
Joseph Hellige Department of Psychology University of Southern California
Christopher Niebauer Department of Psychology University of Toledo
Thomas Hummel Department of Otorhinolaryngology Univ. of Pennsylvania Medical Center
Jacques Vauclair Center for Research in Cognitive Neuroscience, Marseille, France
Kendall Hutson Department of Psychology University of Toledo
E. William Y und Department of Neurology University of California, Davis
Alan Kingstone Department of Psychology University of Alberta
Robert Zatorre Montreal Neurological Institute McGill University
SECTION I: SPATIAL/TEMPORAL FREQUENCY PROCESSING
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
Chapter 1
Hemispheric Asymmetry in the Processing of Spatial Frequency: Experiments Using Gratings and Bandpass Filtering. Stephen D. Christman University of Toledo When Paul Broca first brought the existence of systematic asymmetries in language representation between the left and right cerebral hemispheres (LH and RH) to the attention of the 19th century medical community, the initial reaction was skepticism and disbelief. This was replaced within ten years by widespread acceptance (Harrington, 1987). Initial doubts centered on the prevailing assumption that bilateral symmetry was "perhaps the most general truth in all the science of animal construction" (Moxson, 1866); interestingly, however, a decade later, hemispheric asymmetry in humans was not only widely accepted, it was taken to be a hallmark of human superiority over other organisms: "Man is, of all the animals, the one whose brain in the normal state is the most asymmetrical... It is this that distinguishes us the most clearly from the animals" (Broca, 1877). While the existence of hemispheric differences has not come into serious question since, an unfortunate legacy of the 19th century viewpoint persisted until recently in the form of three implicit assumptions that guided laterality research conducted between 1880 and 1980: (i) that language and other high level cognitive functions were the only lateralized functions, (ii) that only humans
4
Christman
possessed language, and (iii) that, therefore, only humans exhibited significant degrees of hemispheric asymmetry. The last two decades have seen the dispelling of all three assumptions: Stanley Glick's book Cerebral Lateralization in Nonhuman Species (1985) cleared the way for a large growth in the number of studies of hemispheric asymmetries in nonhumans, and the work of researchers such as the Gardners has at least raised the possibility of rudimentary language acquisition in nonhuman primates (e.g., Gardner, Gardner, & Van Cantfort, 1989). The theme of this chapter (and, indeed, of many chapters in this volume) is that hemispheric asymmetry is not limited to higher-order functions and can be demonstrated in a wide variety of sensory and perceptual functions. The implications of hemispheric asymmetries in lower-order functions are important elements in the recent shift in theorizing about brain laterality from emphasis on all-inclusive dichotomies (e.g., Bradshaw and Nettleton's [1981] "analytic/holistic" dichotomy) to a growing realization that behavioral asymmetries (e.g., ear and visual field advantages) are determined by a multitude of factors, some involving cerebral lateralization and some not, some involving higherorder functions and others involving lower-order functions. Hellige (1993) provides an overview of this new componential approach to hemispheric asymmetry. Hemispheric asymmetries in lower-order functions also places the study of hemispheric asymmetry in an evolutionary context. The previous view that asymmetry was confined to higher-order (and especially linguistic) functions implied a sort of evolutionary discontinuity; the current view that asymmetry is present across a wide range of both species and functions places human asymmetry in a richer comparative context, allowing the potential use of animal models in studies of human asymmetry, and helping foster a reevaluation of the neural basis of higher-order asymmetries (c.f., the growing acknowledgment of the importance of non-cortical brain asymmetries). This chapter focuses on hemispheric differences in processing different ranges of spatial frequency content of visual input. Before discussing the relevant literature, however, it is useful to provide background on the role of spatial frequency in visual processing. The modem era in the visual sciences can be traced back to the seminal work of researchers such as Hubel and Wiesel (1962), who helped refine the use of single-cell recording techniques in the study of the neural basis
Spatial Frequency
5
of visual processing. Models of visual processing initially derived from this work posited the existence of various neuronal cell types selectively responding to specific visual features. For example, Hubel and Wiesel (1962, 1965) proposed three important types of cells in striate cortex: (i) simple cells, which respond best to lines, bar, or edges at particular orientations; (ii)complex cells, which respond best to bars or edges moving in specific directions in particular orientations; and (iii) hypercomplex cells, which respond not only to the orientation and direction of motion of stimuli, but also to specific stimulus sizes, lengths, and widths. More extreme versions of this approach have gone so far as to postulate the existence of "pontifical" or "grandfather" cells: cells that fire only when presented with a visual representation of some specific, complex object such as a face or hand (e.g., Barlow, 1972). The 1960s saw an alternative approach emerge which more or less replaced the single-cell feature detection framework. Campbell and Robson (1968) first proposed the existence of discrete pathways in the visual system, each sensitive to a limited range of spatial frequency components. These various pathways or channels were hypothesized to carry out a two-dimensional Fourier analysis of the visual scene, in which complex patterns are broken down into simple, sinusoidal components. Spatial frequency components can be described in terms of a number of dimensions. First, they consist of sinusoidal variations in luminance across space, with higher spatial frequencies involving more numerous cycles per unit distance (the spatial frequency of stimuli is typically described in terms of cycles per degree [cpd] of visual angle; high versus low frequency grating stimuli consist of thinner versus wider bars). Phenomenologically, high frequencies carry information about fine details, while low frequencies carry information about more global aspects of the visual scene. Second, they possess a specific orientation that is perpendicular to the axis of luminance variation. Third, they have some specific contrast, defined by the luminance difference between the lightest and darkest portions of the stimulus divided by the sum the luminances of the lightest and darkest portions of the stimulus. Finally, they have a specific phase, referring to the absolute position in space of the light and dark bars relative to some referent. For a more thorough coverage of the spatial frequency approach, the interested reader is directed to I)eValois and DeValois (1988); a concise but effective overview is provided in Hams (1980).
6
Christman
Thus, in the spatial frequency approach, the fundamental units of visual analysis are not discrete features, but spatially distributed sinusoidal frequency components. The spatial frequency approach has enjoyed great success, and is now a dominant approach to modeling visual processes. A nice example of the utility of the spatial frequency approach over the feature detection approach can be found in a study by DeValois, DeValois, and Yund (1979), who examined single-cell responses to gratings and checkerboard patterns. Checkerboards afford a dissociation between the predictions of the two approaches. Namely, the orientations of the explicit features (i.e., the edges) of a checkerboard are 0 ~ and 90 ~ while the orientations of the fundamental spatial frequency components are • 45 ~ Their procedure involved first identifying cells that produced optimal responding to a sinusoidal grating of some specific orientation (e.g., 0~ According to the feature detection approach, such a cell should exhibit optimal responding to a checkerboard pattern that contains edges oriented at 0 ~ while the spatial frequency approach would predict that such a cell would exhibit no response to such a checkerboard. Rather, that cell would respond optimally to a checkerboard whose edges were oriented at 45 ~ but whose fundamental Fourier component is at 0 ~ Their results confirmed the predictions of the spatial frequency approach: cells tuned to gratings at 0 ~ responded optimally to diagonally oriented checkerboard patterns. Additional evidence was reported concerning similar dissociations involving higher harmonic components and contrast. The ascending dominance of the spatial frequency approach in models of vision during the 1970s led to the first studies of visual field differences in spatial frequency processing, which will be discussed later (e.g., Blake & Mills, 1979; Rao, Rourke, & Whitman, 1981; Rijsdik, Kroon, & van der Wildt, 1980; Rovamo & Virsu, 1979). However, it was Justine Sergent who first formally proposed a model of cerebral hemispheric asymmetry in spatial frequency processing. She came to this hypothesis not through an interest in sensory psychophysics per se, but rather as an outgrowth of an interest in the cerebral bases for facial processing. In a thorough review of the literature, Sergent and Bindra (1981) proposed that perceptual characteristics of facial stimuli play an important role in determining which hemisphere exhibited superior performance. For example, experiments using faces consisting of line drawings and/or in which different faces differed by a single feature tended to yield LH advantages; photographs of faces and/or facial
Spatial Frequency
7
stimuli which differed on many features, on the other hand, tended to yield RH advantages. This led Sergent to the conclusion that a complete understanding of hemispheric differences in facial processing (and, more generally, in any higher-order type of process) required an understanding of potential hemispheric differences at lower .. sensory levels. More specifically, based on the fact that the facial studies that yielded LH versus RH advantages were biased towards the processing of fine versus coarse details, respectively, Sergent (1982) went on to propose that the LH versus RH were specialized for the processing of higher versus lower ranges of spatial frequency content of input. Initial tests of the spatial frequency hypothesis focused on indirect manipulations of the frequency content of input, with increases in size and retinal eccentricity, blurring, and decreases in luminance and exposure duration all attenuating the availability of high, relative to low, spatial frequencies; such manipulations were hypothesized to result in greater relative impairment of LH processing. Christman (1989) reviewed the relevant literature and concluded that there was moderate support for the spatial frequency hypothesis; however, he concluded that more definitive tests of the spatial frequency hypothesis would require the use of simpler grating stimuli or band-pass filtered stimuli. In the eight years since Christman's review was published, a substantial number of such studies have appeared, and are the focus of this chapter (the interested reader is also directed to a recent review by Grabowska and Nowicka [1996] that covers studies employing indirect manipulations of frequency content and/or electrophysiological measures that are beyond the scope of this chapter). Three domains will be reviewed: (i) studies employing single component stimuli, (ii) studies employing compound stimuli containing two or more components, and (iii) studies employing blurred or digitally filtered versions of more complex, naturalistic stimuli (e.g., letters, faces). The chapter will conclude with an evaluation of the current state of the spatial frequency hypothesis, along with recommended directions for future research.
I. Sinusoidal and Square-wave Stimuli A. Contrast Sensitivity/Detection The earliest studies of hemispheric processing of spatial frequency focused on contrast sensitivity, which refers to the threshold contrast
8
Christman
necessary to detect a spatial frequency component. In general, human contrast sensitivity for foveal vision peaks at about 2-4 cpd; the peak shifts to lower frequencies with increasing retinal eccentricity. A number of studies appeared in the late 1970s and early 1980s examining contrast sensitivity functions in the left and right visual fields. The prevailing trend was a finding of hemispheric symmetry in the processing of spatial frequency. Blake and Mills (1979) reported no hemiretinal or hemispheric differences in contrast thresholds for 2 and 6 cpd stimuli. Rao, Rourke, and Whitman (1981) reported an overall LVF superiority in sensitivity to non-flickering gratings that did not interact with spatial frequency (all visual field interactions in their study were confined to effects of temporal frequency, with 0-2 Hz vs. 4-16 Hz being associated with LVF vs. RVF advantages, respectively). Beaton and Blakemore (1981) reported no hemispheric differences in contrast sensitivity for a 3 cpd test grating. Fiorentini and Berardi (1984) reported no hemispheric differences in contrast sensitivity across the range of 0.7 to 7.0 cpd. Kitterle and Kaye (1985) reported hemispheric symmetry in contrast sensitivity, employing a procedural variation in which, rather than determining the lowest contrast at which a given frequency is visible, they determined the highest resolvable frequency at a given contrast level. Peterzell, Harvey, and Hardyck (1989) reported no hemispheric differences in contrast sensitivity functions with gratings ranging from 0.5 to 12 cpd. Finally, Kitterle, Christman, and Hellige (1990) also reported hemispheric symmetry in contrast sensitivity over a range from 0.75 to 12.0 cpd, although they did report a marginal overall LVF advantage in RT that did not interact with frequency. The above studies all involved detection of gratings at threshold, and none indicated hemispheric ~differences as a function of spatial frequency (although two studies, Rao et al. [1979] and Kitterle et al. [1990] reported overall LVF advantages, possibly reflecting a LVF advantage in simple RT; see Christman and Niebauer, this volume). Kitterle et al. (1990) pointed out that the lack of hemispheric asymmetry may have arisen from (i) the threshold contrast levels, and/or (ii) the use of detection tasks; accordingly, they examined suprathreshold detection and also reported no hemispheric differences (although they once again found an overall LVF advantage for RT). Thus, research examining hemispheric differences in contrast sensitivity and spatial frequency detection yields no evidence of hemispheric differences as a function of spatial frequency for threshold or suprathreshold stimuli.
Spatial Frequency
9
B. Discrimination
Discrimination of spatial frequency in the LVF vs RVF has also been examined. Berardi and Fiorentini (1984) examined discrimination performance for successively presented gratings in which one stimulus was fixed at 1.0 cpd and found no visual field differences. Szelag, Budhoska, and Koltuska (1987) employed square-wave gratings (which consist of a fundamental component of frequency f, along with the odd frequency harmonics, i.e., 3f, 5f, 7f, etc.) in a successive same/different task (with "different" trials involving a one octave difference in frequency) and reported no interaction between visual field and spatial frequency. Boles and Morelli (1988) also employed square-wave gratings in a successive same/different task and found no visual field differences (it is not clear from their methods section what ranges of frequency differences were tested). Grabowska, Semenza, Denes, and Testa (1989) used square-wave gratings of low, intermediate, and high frequencies in a successive discrimination task with left- versus rightbrain damaged patients; although the right-brain damaged patients exhibited greater overall impairment, this effect did not interact with frequency. Kitterle and Selig (1991) had subjects decide whether the second of two successively presented sinusoidal gratings was higher or lower in frequency than the first; they reported LVF vs. RVF advantages in the low (1-2 cpd) vs. high (4-12) frequency ranges. Finally, Niebauer and Christman (1997), using the same basic task as Kitterle and Selig (1991), manipulated the interstimulus interval (ISI; 100 vs. 3600 msec) and frequency difference (0.125 vs 1.0 octave) in a discrimination task employing sinusoidal gratings; they found LVF advantages for low frequency stimuli (1.0 cpd) across both ISis and both frequency differences; the complementary RVF advantages for higher frequencies (4.0 cpd) were found for all conditions except the 0.125 octave difference, 100 msec ISI condition. The results for discrimination tasks are somewhat mixed. However, there are potential problems with a number of the studies reporting no visual field differences. For example, the studies by Grabowska et al. (1989) and Szelag et al. (1987) used a constant phase for their gratings, which meant that Ss could have based responses on local luminance cues, and not the frequency of input as such. The two studies reporting visual field X frequency interactions (Kitterle & Selig, 1991; Niebauer & Christman, 1997) randomly varied the phase of the stimuli. Similarly,
10 Christman the procedures of Grabowska et al. (1989), Szelag et al. (1987), and Boles and Morelli (1987) involved uncontrolled luminance changes upon stimulus presentation; given evidence that such luminance changes can have complex and differential masking effects on different ranges of spatial frequency (e.g., Kitterle, Beasley, & Berta, 1984; Green, 1981), it is not clear how to evaluate their results. Again, the studies by Kitterle and Selig (1991) and Niebauer and Christman (1997) kept display luminance constant. Finally, with the exception of the study by Fiorentini and Berardi (1984), all studies reporting no visual field differences employed square-wave gratings, which contain broad ranges of frequency components and are therefore less than ideal for testing hemispheric differences in the processing of narrow and specific ranges of frequency. Thus, the evidence suggests that when sinusoidal stimuli are used and the procedures force Ss to base their responses on the frequency of input as such, there are RVF versus LVF advantages in discriminating stimuli of higher versus lower frequency, respectively. C. Identification
Few studies involving identification of spatial frequency have been conducted. Indeed, the author knows of only one relevant paper. Kitterle, Christman, and Hellige (1990) examined threshold and suprathreshold identification of spatial frequency as a function of visual field. Threshold data indicate that hemispheric differences depend on criteria used to define threshold. Their data indicate that visual field X spatial frequency interactions do not emerge until performance is at or above approximately 80% correct (this is close to the 75% criterion commonly used in psychophysical studies of threshold processing). Furthermore, they found that the visual field X frequency interaction reflected no hemispheric differences at lower frequencies and RVF advantages at higher frequencies. RT data yielded a trend towards a comparable visual field X frequency interaction (p<.06), reflecting no differences for low frequencies and RVF advantages for high frequencies. Their suprathreshold identification experiment employed 1.0 and 9.0 cpd sinusoidal gratings, with contrasts of 0.1, 0.2, and 0.4. They found a significant visual field X frequency interaction, with faster RVF vs LVF RTs for high versus low frequency gratings.
Spatial Frequency
11
D. Orientation judgments A number of studies have examined hemispheric differences in orientation judgments as a function of spatial frequency. Tei and Owens (1980) employed 4 cpd gratings in an adaptation paradigm: subjects viewed an adapting grating for 500, 1000, or 5000 msec, and were then presented with a test grating and judged whether the test and adapting gratings had the same orientation. They reported RVF advantages at the 1000 and 5000 msec adaptation durations. Beaton and B lakemore (1981) also employed an adaptation method to examine visual field differences in orientation processing. Using a baseline frequency of 3 cpd, they found no visual field differences in orientation selectivity. Given that these studies employed only one spatial frequency, they are of limited utility in addressing the existence of visual field X spatial frequency interactions. However, two other studies that also examined orientation processing in the left and right visual fields employed ranges of spatial frequency. Previc (1982) employed a go/no-go orientation identification task using gratings of 0.9 and 5.5 cpd. Although he reported overall RVF advantages in RT and accuracy, visual field did not interact with spatial frequency. Kitterle and Kaye (1985) reported that the visual field symmetries in contrast sensitivity held equally across vertically and obliquely oriented gratings. Finally, Fendrich and Gazzaniga (1990) presented simultaneous pairs of vertical and horizontal gratings with spatial frequencies ranging from 1 to 8 cpd. On each trial, the two gratings had the same frequency, and subjects judged whether or not they had the same orientation. Normal subjects yielded no main effect of visual field and no visual field X frequency interaction (two splitbrain patients were also tested, with one yielding an overall LVF advantage and the other yielding LVF vs RVF advantages at higher vs lower frequencies).
E. Single-component Stimuli: Conclusions To summarize the results obtained with sinusoidal stimuli, there is no evidence for visual field differences in the detection of different ranges of spatial frequency, nor in judgments of orientation as a function of spatial frequency. Tasks requiring the discrimination and identification of spatial frequency, however, appear to yield reliable visual field
12 Christman differences, with LVF vs RVF advantages in processing lower vs higher frequencies, respectively. With regard to the orientation experiments, the lack of visual field differences as a function of spatial frequency seems, on the surface, to contradict the spatial frequency hypothesis. However, deeper consideration shows that such experiments are, in fact, largely irrelevant to the hypothesis. This is so because an observer can identify the orientation of a grating without any explicit processing of that grating's frequency p e r se. That is, other cues, most importantly local luminance variations, can readily provide information about a grating's orientation; in this sense, all an observer has to do is find the orientation of a luminance edge without any reference to the frequency of the grating within which it is embedded. Indeed, an experiment will be discussed later in which subjects made orientation judgments on components of compound stimuli (Kitterle, Christman, & Conesa, 1993); in this experiment, visual field differences were found, presumably because the orientation judgment occurred subsequent to Ss (implicitly) identifying the relevant frequency component. The conclusion with regard to the detection, discrimination, and identification tasks stands the same as that reached by Kitterle and Christman (1991), who suggested that visual field differences in spatial frequency processing do not arise from hemispheric differences in the representation of low vs high spatial frequency channels; all evidence indicates that the two hemispheres receive identical sensory input. Rather, they argued that hemispheric differences arise as a result of the greater computational complexity involved in discrimination and identification decisions. For example, to detect a stimulus, all that the visual system need do is detect some threshold level of activity in any frequency channel; to discriminate and identify a stimulus, however, the relative patterns of activity across multiple frequency channels (which are fairly broadly tuned and exhibit substantial overlap) must be compared. Kitterle and Christman (1991) suggested that hemispheric differences arise in this process of interchannel comparison, with the RH versus LH better at relating and integrating the output of lower versus higher frequency channels.
II. Compound Stimuli The bulk of the work focusing on hemispheric asymmetry in spatial frequency processing has employed single-component stimuli. However,
Spatial Frequency
13
the use of stimuli composed of multiple spatial frequency components is an important step in evaluating the applicability of the spatial frequency hypothesis to the processing of the spectrally complex stimuli (e.g., letters, words, faces) employed in the majority of laterality research. There are a number of reasons why the findings obtained with singlecomponent stimuli may not fully generalize to more spectrally complex patterns. For example, different tasks performed on identical input may load more heavily on different ranges of frequency components. Indeed, many of the early tests of the spatial frequency hypothesis reported just such findings, with variations in task factors producing complex interactions with the spatial frequency content of input (e.g., Christman, 1987, 1990a; Hellige, Jonsson, & Corwin, 1984; Sergent, 1985, 1987). In addition, there is evidence for inhibition between spatial frequency channels (e.g., Gilbert & Wiesel, 1990; Lennie, 1980; Morrone, Burr, & Maffei, 1982). Thus, the way in which a given frequency component is processed in isolation may not generalize to how that component is processed in the context of additional frequency components. A. Task factors
A demonstration of the importance of task factors in influencing patterns of hemispheric asymmetry in spatial frequency processing is provided by Kitterle, Hellige, and Christman (1992). They presented observers with sine-wave and square-wave gratings with fundamental frequencies of 1.0 or 3.0 cpd. Observers made two different classifications. In one block of trials, their task was to indicate whether the bars of the stimulus were wide (i.e., 1.0 cpd) or narrow (i.e., 3.0 cpd); this task required the processing of the lowest, fundamental frequency component of stimuli. In a separate block of trials, observers judged whether the stimulus had sharp (i.e., square-wave) or fuzzy edges (i.e., sine-wave); this task required the processing of the higher harmonic frequencies. The "wide-narrow" task yielded a LVF advantage, while the "sharp-fuzzy" task yielded a RVF advantage. Thus, with spectrally complex stimuli such as square-wave gratings, LVF advantages are found when observers are required to selectively process the lowest frequency component (i.e., the fundamental); RVF advantages, however, are obtained when observers must selectively process higher frequency harmonic components.
14 Christman Kitterle, Christman, and Conesa (1993)examined the role of spatial frequency in the well-established hemispheric asymmetry for processing local versus global levels of structure (e.g., Robertson & Lamb, 1991; Van Kleeck, 1989). Adapting a task developed by Hughes (1986), they presented a set of four compound gratings yielded by the factorial combination of two spatial frequencies (1.5 and 9.0 cpd) and two orientations (45 ~ and 135~ in addition, the contrasts of the two frequency components were chosen so that they would yield equivalent RTs when presented in isolation. Two different tasks were used. In the focused attention task, observers indicated the orientation of either the high or low frequency component; in the divided attention task, observers indicated whether or not a target orientation (e.g., 45 ~ was present in one or both components. For both tasks, the low and high frequency components were comparable to global and local levels, respectively; furthermore, the irrelevant component could be either congruent and incongruent in orientation to the relevant component, providing an analog for interlevel interference. The results for the focused attention task provided a close fit with the local-global findings: responses were faster and more accurate to low frequency components than to high, and this advantage for low frequencies was larger for LVF trials; more importantly, interference was larger in the LVF when the irrelevant dimension was a low frequency component and was obtained only in the RVF when the irrelevant dimension was a high frequency component. Comparable results were obtained for the divided attention task, with faster and more accurate responding in the LVF vs RVF when the relevant orientation was present at only the low vs high frequency level, respectively. In addition to interpreting their results as supporting the notion that lower vs higher spatial frequencies underlie the processing of global and local levels, respectively, Kitterle et al. (1993) point out that, although the contrasts of the high and low frequency components had been chosen to equate RTs, responses to (and interference from) the low frequency component were faster (and larger) than for the high frequency component, suggesting that low frequency channels inhibited high frequency channels. In addition, the presence of visual field differences suggests that this interchannel inhibition differs in the left and fight hemispheres, with low frequency channels producing greater inhibition of high frequency channels in the left hemisphere. Thus, with compound stimuli, hemispheric differences may arise via hemispheric differences in
Spatial Frequency
15
processing specific ranges of frequency and/or via hemispheric differences in the magnitude of interchannel inhibition.
B. Relative frequency Related to the above discussion of interchannel inhibition is the issue of relative frequency. This refers to whether a given component is higher versus lower in frequency relative to other components present in the stimulus. Given the aforementioned evidence for greater low-onhigh frequency inhibition than the reverse, a given frequency component that is the lowest frequency contained in a stimulus will receive less inhibition than when the same component is of higher frequency relative to other components. Christman, Kitterle, and Hellige (1991) addressed this issue employing compound gratings. The basic task involved identifying which of two possible stimuli were presented on each trial. Baseline stimuli consisted of compound gratings with two components, and the other stimuli consisted of a 2.0 cpd component added to the two components of the baseline. In one condition, the baseline stimulus consisted of 0.5 and 1.0 cpd components and the alternative consisted of 0.5, 1.0, and 2.0 cpd components; thus, the 2.0 cpd component which differentiated the two stimuli was of high frequency relative to the other components. In a second condition, the baseline stimulus consisted of 4.0 and 8.0 cpd components and the alternative consisted of 2.0, 4.0, and 8.0 cpd components; thus, the 2.0 cpd component which differentiated the two stimuli was of low frequency relative to the other components. In other words, the only difference between the two stimuli in each condition was the presence or absence of the 2.0 cpd component; however, in one condition, that component was the lowest relative frequency, while in the other condition that component was the highest relative frequency. Results indicated that the ability of each hemisphere to process the 2.0 cpd component depended on its relative frequency: when it was the highest relative frequency, there was a RVF advantage in RT; however, when it was the highest relative frequency, there was a LVF advantage in RT. A number of important points can be made about this finding. First, it is unlikely that observers performed the task by simply assessing whether the 2.0 cpd component was present or not. This is so because the RT to three component stimuli was reliably longer than to two component stimuli; since "present" judgments are typically faster than
16 Christman absent judgments, such a strategy should have led to faster RTs to the three component stimuli. Thus, it appears that observers processed the compound stimuli as such. Second, the influence of relative frequency outweighed the influence of absolute frequency. For example, the mean absolute frequency of components in the high relative frequency condition (i.e., 0.5, 1.0, and 2.0 cpd) was lower than in the low relative frequency condition (i.e., 2.0, 4.0, and 8.0), yet a RVF advantage was obtained (the converse comparison holds for the LVF advantage in the low relative frequency condition). Third, the results of this study can be related to the results of Kitterle, Hellige, and Christman (1992), who reported LVF vs. RVF advantages in processing fundamental versus harmonic frequency components. That is, in the low relative frequency condition, the 2.0 cpd component was equivalent to a fundamental frequency and was processed better in the LVF; conversely, in the high relative frequency condition, the 2.0 cpd component was equivalent to a harmonic frequency and was processed better in the RVF. Finally, Ivry and Lebby (1993) reported anologous hemispheric differences in the processing of relative auditory frequency. C. Beat frequency Christman (1990b) attempted to dissociate between spatial frequency and spatial periodicity, taking advantage of a phenomenon called the "missing fundamental". This refers to a situation in which spatially periodic interactions among higher frequency components produce "beats" at a frequency lower than that of any of the constituent components. These "beats" consist of amplitude-modulated variations in contrast (phenomenologically, these appear as "washed-out" areas of the compound grating wherein no light or dark bars are visible) and have a spatial periodicity much lower than that of the actual frequency components. There is evidence that such periodicities are treated as real visual features by the visual system (e.g., Henning, Hertz, & Broadbent, 1975; but see DeValois, 1978 for a lack of such evidence). Christman (1990b) presented observers with two tasks. In the first, they distinguished between a low frequency stimulus composed of 1.5 + 2.0 cpd components and a high frequency stimulus composed of 7.0 + 8.0 cpd components; this task yielded expected LVF vs RVF advantages for low vs high frequency stimuli, respectively (see Fig. 1). The second task had observers distinguish between two high frequency
Spatial Frequency 400 -
380
,.,
LVF-RH
17
L
-
RT
"
360
-
340
!
!
1.5 + 2.0 cpd 570 -
Stimulus
7.0 + 8.0 cpd
LVF-RH
v
RVF-LH
A v
560 RT
550
540
O--------___
!
!
6.0 + 8.0 cpd
7.0 + 8.0 cpd StimulUs
Figure 1. Top panel: Discrimination between compound gratings composed of low versus high frequency components. Bottom panel: Discrimination between compound gratings composed of high frequency components (discriminable in terms of low versus high spatial periodicity).
stimuli, one composed of 7.0 + 8.0 cpd components and the other composed of 6.0 + 8.0 cpd components. Although both of these stimuli consist of high frequency components, they are distinguishable by their "beat" frequencies: the 7.0 + 8.0 cpd stimulus has amplitude-modulated variations in contrast with a periodicity of 1 cpd, while the 6.0 + 8.0 cpd stimulus has a m p l i t u d e - m o d u l a t e d variations in contrast with a
18 Christman periodicity of 2 cpd. Interestingly, even though in this latter condition all components were of high spatial frequency, a marginal LVF advantage was found for the 7.0 + 8.0 cpd stimulus and no visual field difference was obtained for the 6.0 + 8.0 cpd stimulus. These results highlight the fact that visual field differences are not a simple function of the spatial frequency of the components of compound gratings; rather, they suggest that the visual field differences obtained for the processing of spatial frequency may be a special case of a more general visual field difference in processing spatial periodicity; similarly, they provide another demonstration of how the processing of individual spatial frequency components occurs in a rich context of both other frequency components and changing task demands. D. Phase
Before discussing hemispheric differences in phase perception, some important aspects of spatial phase processing need to be mentioned. First, as discussed above, phase refers to the specific position in space of a spatial frequency component; however, one can distinguish between absolute phase (i.e., the position of a grating with respect to some specific arbitrary point in space) and relative phase (i.e., the position of a grating with respect to other gratings in the same region of space). Second, sensitivity to phase information decreases with increases in spatial frequency (e.g., Holt & Ross, 1980). Third, it is important to note that phase information is at least as important in pattern recognition as is the amplitude and orientation of spatial frequency components (e.g., Piotroski & Campbell, 1982). Lastly, there is evidence that the processing of phase information in the periphery may be qualitatively different from such processing in the fovea (e.g., Stephenson, Knapp, & Braddick, 1991), suggesting that studies of hemispheric differences in phase processing employing lateral stimulus presentation may have only limited utility in shedding light on how the hemispheres differ in processing foveal phase information. To date, only two studies have explicitly examined visual field differences in phase processing. Fiorentini and Berardi (1984) had observers discriminate between pairs of gratings consisting of the sum of two frequencies, f plus 3f (f equaled either 0.5 or 1.0 cpd). One grating of the pair had a constant phase relation between the two components, while the other grating had variable phase relations. They found that the
Spatial Frequency
19
just-discriminable phase difference was 20 ~ to 30 ~ smaller in the LVF. This effect held for both simultaneous and successive presentations. Rentschler, Christen, Christen, and Landis (1986) pointed out that the methods of Fiorentini and Berardi may have allowed observers to perform the discrimination on the basis of local contrast cues, as opposed to spatial phase per se. Accordingly, the constructed pairs of f plus 3f gratings in which the only difference was the relative phase of the components. They reported that the discrimination of peaks-subtract versus peaks-add gratings was superior in the LVF (this effect was limited, however, to 150 msec exposure durations; shorter durations did not yield visual field differences); however, discrimination of mirrorsymmetric gratings did not yield visual field differences. Thus, there is tentative evidence for LVF advantages in phase processing. Two important caveats must be offered, however. First and foremost, given the relative insensitivity to phase at higher frequencies, both of the above studies employed low frequency components; thus, it is not clear the extent to which the LVF advantages reflect a superiority in the processing of phase versus low spatial frequencies. Second, the fact that the visual field differences were dependent on such variables as exposure duration and specific types of phase relations casts doubt on a conclusion of overall LVF superiority in spatial phase processing.
III. Low-pass and Band-pass Filtered Stimuli The final area to be reviewed in this chapter consists of studies employing complex stimuli (e.g., letters, faces) that have been either blurred so as to remove higher spatial frequencies or digitally filtered so as to remove specific ranges of spatial frequency content. Use of such stimuli is important, as it provides a bridge between the types of lowerlevel sensory processing reviewed above and the higher-order pattern recognition processes that play important roles in many laterality experiments. Unfortunately, relatively few such studies exist, in large part owing to the methodological difficulties inherent to manipulations of spatial frequency content of complex (i.e., non-sinusoidal) stimuli. The simplest way to manipulate the spatial frequency content of complex stimuli is to simply blur the stimuli, which has the effect of attenuating high, relative to low, spatial frequency content. However, not all methods of blurring are appropriate; for example, simply defocusing the lens of a slide projector, while phenomenologically blurring an
20
Christman
image, has complex and nonlinear effects across a broad range of frequencies, and is therefore not appropriate as a method of manipulating frequency content. A more exact and widely used method has involved dioptric blurring, which involves changes in the focal length of an image passing through the lens, resulting in a blurred image being formed on the focal plane of the retina. Although this form of blurring attenuates sensitivity across a wide range of frequencies, the attenuation is specifically greater at higher frequencies and can quantified. Digital filtering represents a more precise method of controlling the spatial frequency content of complex input. This method involves computation of the Fourier transform of the input; specific bands of frequency content are then filtered from the resultant amplitude spectra, and the reverse Fourier transform is computed, yielding a band-pass filtered stimulus. While this method represents the optimal method of manipulating the spatial frequency content of input, it has rarely been used in laterality studies, owing in large part to the technical difficulties inherent to digital filtering. These include not only learning to use the growing body of digital filtering software (NIH's Image program is an excellent example), but also such issues as choosing the appropriate filtering parameters (e.g., shape and width of the filter) and correcting for non-linearities of display devices (e.g., gamma corrections). The first studies to employ precise forms of blurting were conducted by Justine Sergent (1985; 1987). Sergent (1985) reported RVF advantages for two facial tasks using unfiltered stimuli; the same tasks yielded LVF advantages when stimuli were filtered to remove all frequencies above 2 cpd. In a follow-up study, Sergent (1987) employed broad-pass and low-pass face versions (a third, quantized version was also used) and varied exposure duration. At the short exposure duration, LVF advantages were obtained for all face versions; at a longer exposure, however, the broad-pass version yielded a RVF advantage (presumably reflecting the fact that the extended exposure allowed the left hemisphere to extract higher frequencies), while the lowpass yielded no visual field differences. A similar study was conducted by Whitman and Keegan (1991), who digitally filtered faces so as to selectively remove either high or low frequency content, and found that removing high frequencies produced greater impairment of RVF processing (removal of low frequencies had comparable effects in both visual fields). Taken together, these results suggest that the right
Spatial Frequency
21
hemisphere is more efficient at extracting low frequencies from facial stimuli, especially under conditions of brief exposure durations. Furthermore, these results are consistent with those reported by Keenan, Whitman, and Pepe (1989), who presented faces either alone or masked with square wave gratings; low versus high frequency masks produced greater relative impairment of LVF versus RVF performance, respectively. Blurring manipulations have also been employed with non-facial stimuli. Jonsson and Hellige (1986) had subjects perform a physical identity letter comparison task on blurred and clear stimuli. Blurting produced greater impairment of RVF performance (with difficult comparisons, this effect was limited to accuracy measures, while it was confined to RT measures for easy comparisons). Michimata and Hellige (1987) extended this finding to a physical identity comparison task employing novel, nonlinguistic forms, again finding that blurring impaired RVF, but not LVF, performance. Christman (1990a) employed two levels of dioptric blur in a temporal integration task requiring digit recognition (the task involved breaking down a complete digit into two separate patterns, which were then flashed successively, separated by an interflash interval). The no blur condition yielded no visual field differences, while one diopter of blur (which removes all frequencies above roughly 9 cpd) impaired RVF performance but had no effect on LVF performance. Two diopters of blur (which removes all frequencies above roughly 3.5 cpd) impaired performance in both visual fields, but still yielded a LVF advantage. In an additional experiment, Christman (1990a) found that the effect of blur interacted with interflash interval: at a short interval, blur produced the expected pattern of greater impairment of RVF performance, whereas at a longer interval (which, due to the characteristics of visible persistence, put a premium on the processing higher frequencies) blur produced equivalent impairment of LVF and RVF performance. This finding echoes that of Sergent (1987), who also reported interactions between visual field, spatial frequency content, and temporal variables. Finally, two studies have examined the classification of band-pass filtered letter stimuli. Peterzell, Harvey, and Hardyck (1989) employed letters of three different sizes that had been Gaussian band-pass filtered to contain 1-octave ranges of spatial frequency content centered on five frequencies (1.0, 2.0, 4.0, 8.0, and 16.0 cpd), and reported no interactions for RT or accuracy between frequency and visual field
22
Christman
(indeed, the only visual field effect was for response bias, reflecting a bias for LVF trials to be yield "target" responses). Christman, Kitterle, and Niebauer (1997) argued that the lack of visual field effects in the Peterzell et al. study may have arisen from factors other than a lack of visual field difference in spatial frequency processing as such. In particular, they implicated two factors. First, the choice of target versus nontarget letters in the Peterzell et al. study may have allowed subjects to perform the task on the basis of simple feature detection; in this case, visual field differences would not be expected, as detection tasks have been shown to not yield frequency X visual field interactions (e.g., Kitterle, Christman, & Hellige, 1990). Second, Peterzell et al. had subjects run through over 2000 trials, raising the possibility that practice and/or fatigue effects may have obscured visual field differences. Accordingly, Christman et al. (1997) replicated the Peterzell et al. study using both their original target/nontarget classification, as well as a new classification that eliminated the potential featural confound, as well as limiting the total number of trials to 640 per subject. Both classification types yielded visual field X frequency interactions. For example, Experiment 1 from Christman et al. (1997), employing the new classification scheme, yielded a visual field X frequency interaction for RT (with LVF advantage at 1.0 cpd and RVF advantages at 8.0 and 16.0 cpd), although this effect was most pronounced for target stimuli and for large letters. Experiment 1 also yielded a marginal visual field X frequency interaction for sensitivity (i.e., d'), although the nature of the interaction was not fully consistent with the spatial frequency hypothesis: a marginal LVF advantage was found at 8.0 cpd, with no visual field differences at 4.0 or 16.0 cpd). Experiment 2 from Christman et al. (1997), employing Peterzell et al.'s original classification scheme, yielded a marginal visual field X frequency interaction in the predicted direction, although it was restricted to nontarget, large letters. There was also a visual field X frequency interaction for d ~ reflecting a marginal LVF versus RVF advantages at 4.0 versus 16.0 cpd, respectively (the 8.0 cpd did not yield visual field differences. Finally, both experiments yielded visual field differences in response bias, reflecting the same LVF "target" response bias reported by Peterzell et al. (1989). Thus, it appears that the lack of visual field effects reported by Peterzell et al. (1989) may have arisen in part from the number of trials employed. However, the visual field X frequency interactions in the Christman et al. study (1997) were generally of small size and marginal
Spatial Frequency
23
significance and were qualified by the fact that such interactions were often limited to targets only or certain letter sizes. The conclusions, therefore, with regard to the spatial frequency hypothesis of hemispheric asymmetry are, at best, mixed. Christman et al. (1997) point out, however, that the absence of visual field X frequency interactions for the response bias measure argues against Peterzell's (1991) contention that most, if not all, previous reports of visual field X frequency interactions arose from hemispheric differences in response bias and not sensitivity to different ranges of frequency content. IV. Conclusions
What is the status of the spatial frequency hypothesis fifteen years after Sergent's initial formulation? First, it is apparent that hemispheric differences in spatial frequency processing do not arise from hemispheric differences in the distribution of visual pathways carrying low versus high frequency information, as detection tasks do not yield any evidence for systematic hemispheric differences. However, tasks requiring the activity in different channels to be evaluated and compared (e.g., discrimination and identification) do yield consistent RH versus LH advantages for the processing of low versus high frequencies, respectively. This pattern of results also extends to the use of compound stimuli containing multiple frequency components. Tasks that bias attention to fundamental versus harmonic frequencies yield RH versus LH advantages, respectively (e.g., Kitterle, Hellige, and Christman, 1992; although this may be limited to stimuli whose fundamental frequencies are below some absolute value). Similarly, RH versus LH advantages for the processing of a given component depend on whether that component is of high or low frequency relative to other components of the stimulus (e.g., Christman, Kitterle, & Hellige, 1991; although again, this effect may limited to certain ranges of absolute frequency). The picture is less clear when it comes to the use of complex (e.g., non-sinusoidal) stimuli that have been filtered to remove specific ranges of frequency content. Studies employing filtered faces have yielded good support for the spatial frequency hypothesis, although the studies by Sergent (1985; 1987) indicated complex interactions between exposure duration and frequency in the determination of hemispheric differences. The studies by Peterzell et al. (1989) and Christman et al.
24
Christman
(1997), employing bandpass filtered letters, yielded marginal support at best: Peterzell et al. failed to find any evidence of visual field X frequency interactions, while the visual field X frequency interactions reported by Christman et al. were generally of marginal significance and did not appear to be very reliable. This lack of consistent and/or robust hemispheric differences as a function of spatial frequency in the processing of complex stimuli is mirrored by similar findings from studies employing indirect manipulations of frequency content (e.g., variations in retinal eccentricity, exposure duration, size, etc.). In a review of such studies, Christman (1989) concluded that there was moderate support for the spatial frequency hypothesis: 45 out of 75 relevant studies yielded visual field X frequency interactions in the predicted direction. However, it should be noted that, again, many of these interactions reflected marginal effect sizes; furthermore, nine of the 75 studies yielded visual field X frequency interactions in the opposite direction. How does one reconcile the robust support for hemispheric differences in spatial frequency processing obtained in studies employing sinusoidal stimuli with the equivocal support yielded by studies employing more complex stimuli? A number of possibilities exist. First and foremost, hemispheric asymmetries are multiply determined: a given experiment condition may involve a host of task and input factors varying in degree and direction of lateralization. In the studies employing sinusoidal stimuli, the only relevant input factor has been the spatial frequency of input as such; thus, in these studies, frequency emerges as the primary determinant of hemispheric processing differences. In studies employing more complex stimuli, however, a number of other input and task factors come into play and may potentially attenuate, obscure, and even reverse the usual hemispheric effects of frequency. For example, experiments using verbal materials appear to be particularly resistant to robust visual field X frequency effects, as indicated by the weak effects reported by Peterzell et al. (1989) and Christman et al. (1997), as well as by the fact that only 5 out of 13 studies employing words that were reviewed by Christman (1989) yielded the expected visual field X frequency interaction. In this case, it is possible that LH dominance for many aspects of verbal processing overrides any asymmetric effects of spatial frequency. Other factors may also account for the relatively weak nature of spatial frequency effects obtained with complex stimuli. Although the
Spatial Frequency
25
existence of visual pathways carrying specific ranges of spatial frequency content is well established, it is critical to bear in mind that these pathways also carry various types of temporal information. For example, there is a rough correspondence between low frequencies and transient temporal responsiveness and between high frequencies and sustained temporal responsiveness. Thus, temporal factors related to duration, onset and offset of stimuli may modulate spatial frequency effects; indeed, Mecacci (this volume) reviews a large body of literature showing that hemispheric processing of spatial frequency is very sensitive to temporal factors. Another important factor involves inhibition between channels carrying different ranges of spatial frequency information, with lower frequencies inhibiting higher frequencies (e.g., Gilbert & Wiesel, 1990; Lennie, 1980; Morrone, Burr, & Maffei, 1982). This means that the extraction of high spatial frequency information is dependent on the range and magnitude of lower frequency content of the input. Thus, with complex stimuli containing broad ranges of frequency content, the extraction of particular ranges of frequency content does not occur in isolation, but will be influenced by task requirements, temporal factors associated with input presentation, and by interactions with other bands of frequency content of the input. In closing, what are the implications of the spatial frequency hypothesis for theories of cerebral asymmetry in general? The studies employing sinusoidal stimuli leave little doubt that the left and right hemispheres differ in the efficiency with which they process the output of high and low spatial frequency channels. Thus, experimenters must pay careful attention to the specific choices of visual parameters such as resolution, luminance, contrast, and exposure duration. Until Sergent's initial formulation of the spatial frequency hypothesis, such factors were treated as "invisible" variables, and were not explicitly taken into account by laterality researchers; for example, laterality studies from the 1960s and 1970s regularly failed to report any luminance values for stimuli. On the other hand, the material reviewed above indicates that the influence of spatial frequency content on hemispheric asymmetries in the processing of more complex stimuli is, at best, moderate. In this sense, while researchers must select particular input parameters with care, it is clear that spatial frequency does not have a strong, overriding influence on hemispheric asymmetries, but is rather one of many determinants of such asymmetries.
26
Christman
Ultimately, perhaps the most important contribution of the spatial frequency hypothesis has been the forceful demonstration that hemispheric asymmetries are not confined to higher order processes. While this does not deny the special nature of higher order asymmetries in processes such as language, it suggests that such asymmetries likely have ontogenetic origins in pre-existing, lower level asymmetries of sensory and perceptual function. Even in cases where more recent, higher level functions largely override underlying perceptual asymmetries (as was suggested above for many types of verbal processing), consideration of the sensory bases may still inform models of higher level function. Indeed, this echoes a growing theme in both cognition and neuropsychology, as demonstrated by the debate on the extent to which linguistic semantic representations are grounded in perceptual versus conceptual systems (e.g., Barsalou, 1993), as well as Previc's (1991) recent model proposing that hemispheric asymmetries in high level verbal and spatial processing arise from relatively small hemispheric differences in sensory feedback during fetal and infant development. Finally, models of hemispheric differences in sensory function may provide a more direct foundation than models of higher-order function for attempts to understand the cytoarchitectonic bases of mental function. For example, Scheibel (1984) reported hemispheric differences in patterns of dendritic branching, with higher- versus lowerorder branching being more prevalent in the LH versus RH, respectively. Higher- versus lower-order branching is consistent with fine versus coarse grained neural networks, which, in turn, could provide a physiological foundation for the processing of high versus low frequencies, respectively.
References
Barlow, H.B. (1972). Single units and sensation: A neuron doctrine for perceptual psychology? Perception, 1,371-394. Barsalou, L.W. (1993). Flexibility, structure, and linguistic vagary in concepts: Manifestations of a compositional system of perceptual symbols. In A.F. Colllins et al. (Eds.), Theories Of memory. Lawrence Erlbaum Associates: Hove, England. Beaton, A., & Blakemore, C. (1981). Orientation selectivity of the human visual system as a function of retinal eccentricity and visual hemifield. Perception, 10, 273-282.
Spatial Frequency
27
Blake, R., & Mills, J. (1979). Pattern and flicker detection examined in terms of the nasal-temporal division of the retina. Perception, 8, 549555. Boles, D.B., & Morelli, M.L. (1988). Hemispheric sensitivity to spatial frequencies. Bulletin of the Psychonomic Society, 26, 552-555. Bradsliaw, J.L., & Nettleton, N.C. (1981) The nature of hemispheric specialization in man. Behavioral and Brain Sciences, 4, 51-91. Broca, P. (1877). Rapport sur un mEmoire de M. Armand de Fleury intitul6: De l'in(galitd dynamique des deux h(misph~res cdr(braux. Bulletins de l'Acadgmie de Mgdecine, 6, 508-539. Campbell, F.W., & Robson, J.G. (1968). Application of Fourier analysis to the visibility of gratings. Journal o f Physiology (London), 197, 551-566. Christman, S.D. (1987). Effects of perceptual quality on hemispheric asymmetries in visible persistence. Perception & Psychophysics, 41, 367-374. Christman, S.D. (1989). Perceptual characteristics and visual laterality research. Brain and Cognition, 11,239-257 Christman, S.D. (1990a). Effects of luminance "and blur on hemispheric asymmetries in temporal integration. Neuropsychologia, 28, 361374. Christman, S.D. (1990b). Hemispheric asymmetry in the processing of spatial frequency versus spatial periodicity. Presented at the TENNET conference, Montreal, May 10-12, 1990. Christman, S.D. (1993). On the complex relation between perceptual characteristics and hemispheric asymmetry. Brain and Cognition, 21, 123-129. Christman, S.D., Kitterle, F.L., & Hellige, J.B. (1991). Hemispheric asymmetry in the processing of relative versus absolute spatial frequency. Brain and Cognition, 16, 62-73. Christman, S.D., Kitterle, ~L., & Niebauer, C.L. (1997). Hemispheric asymmetries in the identification of band-pass filtered letters. Psychonomic Bulletin and Review, 4, 277-284. DeValois, K.K. (1978). Adaptation to the missing fundamental. Association for Research in Vision and Opthalmology, 243, 243. DeValois, R.L., & DeValois, K.K. (1988). Spatial vision. New York: Oxford University Press. DeValois, K.K., DeValois, R.L., & Yund, E.W. (1979). Responses of striate cortex cells to grating and checkerboard patterns. Journal of Physiology (London), 291,483-505. Fendrich, R., & Gazzaniga, M. (1990). Hemispheric processing of atial frequencies in two c o m m i s u r o t o m y patients. uropsychologia, 28, 657-664. Fiorentim, A., & Berardi, N. (1984). Right-hemisphere superiority in the discrimination of spatial phase. Perception, 13, 695-708. Gardner, R.A., Gardner, B.T., & Van Cantfort, T.E, (1989). Teaching sign language to chimpanzees. Albany, NY: SUNY Press. Gilbert, C.D., & Wiesel, T.N. (1990). The influence of contextual stimuli on the orientation selectivity of cells in the primary visual cortex of the cat. Vision Research, 3 0, 1689-1701.
~e
28 Christman Glick, S.D. (1985). Cerebral lateralization in nonhuman species. New York: Academic Press. Grabowska, A., & Nowicka, A. (1996). Visual-spatial frequency model of cerebral asymmetry: A critical survey of behavioral and electrophysiological studies. Psychogical Bulletin, 120, 434-449. Grab0wska, A., Semenza, C., Denes, G., & Testa, S. (1989). Impaired grating discrimination following right hemisphere damage. Neuropsychologia, 2 7, 259-263. Green, M.A. (1981). Spatial frequency effects of masking by light. Vision Research, 21, 861-866. Harrington, A. (1987). Medicine, mind and the double brain. Princeton, NJ: Princeton University Press. Harris, C.S. (1980). Visual coding and adaptability. Hillsdale, NJ: Lawrence Erlbaum Associates. Hellige, J.B., Jonsson, J.E., & Corwin, W.H. (1984). Effects of perceptual quality on the processing of human faces presented to the lefl~ and fight cerebral hemispheres. Journal of Experimental Psychology: Human Perception and Performance, 1 O, 90-107. Hellige, J.B. (1993). Hemispheric asymmetry. Cambridge, MA: Harvard University Press. Henning, G.B., Hertz, B.G., & Broadbent, D. (1975). Some experiments beanng on the hypothesis that the visual system analyses spatial patterns in independent bands of spatial frequency. Vision Research, 15, 887-897. Holt, J.J., & Ross, J. (1980). Phase perception in the high spatial frequency range. Vision Research, 20, 933-935, Hubel, D.H., & Wiesel, T.N. (1962). Receptive fields, binocular interaction, and functional architecture in the cat's visual cortex. Journal of Physiology, 160, 106-154. Hubel, D.H., & Wiesel, T.N. (1965). Receptive fields of and functional architecture in two nonstriate visual areas (18 and 19) of the cat. Journal o f Neurophysiology, 28, 229-289. Hughes, H.C. (1986). Asymmetric interference between components of suprathreshold compound gratings. Perception & Psychophysics, 40, 241-250. Ivry, R.B., & Lebby, P. (1993). Hemispheric difference in auditory perception are similar to those found in visual perception. Psychological Science, 4, 41-45. Jonsson, J.E., & Hellige, J.B. (1986). Lateralized effects of blurring: A test of the visual spatial frequency model of hemispheric asymmetry. Neuropsychologia, 24, 351-362. Keenan, P.A., Whitman, R.D., & Pepe, J. (1989). Hemispheric asymmetry in the processing of high and low spatial frequencies: A facial recognition task. Brain and Cognition, 11, 229-237. Kitterle, F.L.,Beasley, E.M., & Berta, J. (1984). The effect of luminance decrements upon the detection of sinusoidal gratings. Perception & Psychophysics, 3 5, 221-228. Kitterle, F., & Christman, S. (1991). Hemispheric Symmetries and Asymmetries in the Processing of Sine-wave Gratings. In F. Kitterle (Ed.), Advances in Cerebral Laterality Research: Data and Theory. Hillsdale, N.J." Lawrence Erlbaum Associates.
Spatial Frequency 29 Kitterle, F., Christman, S., & Conesa, J. (1993). Hemispheric differences in the interference among components of compound gratings. Perception and Psychophysics, 54, 785-793. Kitterle, F., Christman, S., & Hellige, J. (1990). Hemispheric differences are found in the identification, but not detection, of low vs. high spatial frequencies. Perception & Psychophysics, 48, 297-306. Kitterle, F.L., Hellige, J.B., & Christman, S.D. (1992). Hemispheric asymmetries depend on which spatial frequencies are task relevant. Brain and Cognition, 20, 308-314. Kitterle, F.L, & Kaye, R.S. (1985). Hemispheric symmetry in contrast and orientation sensitivity. Perception & Psychophysics, 37, 391-396. Kitterle, F.L. & Selig, S.M. (1991). Visual field effects in the discrimination of sine-wave gratings. Perception & Psychophysics, 50, 15-18. Lennie, P. (1980). Parallel visual pathways: A review. Vision Research, 20, 561-594. Michimata, C., & Hellige, J.B. (1987). Effects of blurring and stimulus size on the lateralized processing of nonverbal stimuli. Neuropsychologia, 25, 397-407. Morrone, M.C., Burr, D.C., & Maffei, L. (1982). Functional implications of cross-orientation inhibition of cortical cells: I. Neurophysiological evidence. Proceedings of the Royal Society of London: Series B, 216, 335-354. Moxson, W. (1866). On the connexion between loss of speech and paralysis of the right side. The British and Foreign MedicoChirurgical Review, 3 7, 481-489. Niebauer, C.L., & Christman, S.D. (1997). Categorical versus coordinate processing of spatial frequency relations. Manuscript submitted for publication. Peterzell, D. (1991). On the nonrelation between spatial frequency and cerebral hemispheric competence. Brain and Cognition, 15, 62-68. Peterzell, D., Harvey, Jr., L., & Hardyck, C. (1989). Spatial frequencies and the cerebral hemispheres: Contrast sensitivity, visible persistence, and letter classification. Perception & Psychophysics, 46, 2143-455. Piotrowski, L.N., & Campbell, F.W. (1982). A demonstration of the visual importance and flexibility of spatial-frequency amplitude and phase. Perception, 11,337-346. Previc, F.H. (1982). Visual pattern recognition in the cerebral hemispheres: The role of spatial filtering. Perceptual and Motor Skills, 55, 1319-1326. Previc, F.H. (1991). A general theory concerning the prenatal origins of cerebral lateralization in humans. Psychological Review, 98, 299324. Rao, S.M., Rourke, D., & Whitman, R.D. (1981). Spatio-temporal discrimination of frequency in the right and left visual fields: A preliminary report. Perceptual and Motor Skills, 53, 311-316. Rentschler, I., Chi-isten, L., Christen, S., & Landis, T. (1986). Features versus spatial phase in a tachistoscopic laterality experiment. Perception & Psychophysics, 39, 205-209.
30
Christman
Rijsdijk, J.P., Kroon, J.N., & van der Wildt, G.J.(1980). Contrast sensitivity as a function of position on the retina. Vision Research, 20, 235-241. Robertson, L., & Lamb, M. (1991). Neuropsychological contributions to theories of part/whole organization. Cognitive Psychology, 23, 299330. Rovamo, J., & Virsu, V. (1979). An estimation and application of the human cortical magnification factor. Experimental Brain Research, 3 7, 495-510. Scheibel, A.B. (1984). A dendritic correlate of human speech. In N. Geschwind & A. Galaburda (Eds.), Cerebral dominance. Harvard University Press: Cambridge, MA. Sergent, J. (1982). The cerebral balance of power: Confrontation or cooperation? Journal of Experimental Psychology: Human Perception and Performance, 8, 253-272. Sergent, J. (1985). Influence of task and input factors on hemispheric involvement in face processing. Journal of Experimental Psychology:
Human Perception and Performance, 11,846-861.
Sergent, J. (1987). Failures to confirm the spatial-frequency hypothesis: Fatal blow or healthy complication? Canadian Journal of Psycho logy_,41, 412-428. Sergent, J., & Bindra, D. (1981). Differential hemispheric processing of faces: Methodological considerations and reinterpretation. Psychological Bulletin, 89, 541-554. Stephenson, C.M., Knapp, A.J., & Braddick, O.J. (1991). Discrimination of spatial phase shows a qualitative differences between foveal and peripheral processing. Vision Research, 31, 1315-1326. Szelag, E., Budhoska, W., & Koltuska, B. (1987). Hemispheric differences in the perception of gratings. Bulletin of the Psychonomic Society, 25, 95-98. Tei, B.E, & Owen D.H. (1980). Laterality differences in sensitivity to line orientation as a function of adaptation duration. Perception & Psychophysics, 28, 479-483. Van Kleeck, M. (1989). Hemispheric differences in global versus local processing of hierarchical visual stimuli by normal subjects: New data and a meta-analysis of previous studies. Neuropsychologia, 27, 1165-1178. Whitman, R.D., & Keegan, J.F. (1991). Lateralization of facial processing: A spatial frequency model. International Journal of Neuroscience, 60, 177-185.
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
31
Chapter 2
Temporal Frequency Processing Luciano Mecacci Universit
degli Studi di Firenze, Italy
The temporal dimension of visual stimulus was considered a decisive factor in affecting the direction of visual-field asymmetries since the early proposals of the spatial-frequency hypothesis. Justine Sergent, in one of her first papers on sensory resolution of the two cerebral hemispheres (1982b), found that the variation of exposure duration made the visual field superiority (in terms of reaction-time speed) shift from one side to another. At short durations (40 msec), the left visual field (LVF)/right hemisphere (RH) was superior in a facial categorization task. However, at long durations (200 msec), right visual field (RVF)/ left hemisphere (LH) became superior. The effect of exposure duration, replicated in many other successive works, was interpreted in the framework of research carried out in the Seventies on space and time domains in spatial-frequency analysis (see a review in Sergent & Hellige, 1986). On one hand, there was the general neurophysiological model of two visual systems tuned to different "combinations" of spatial (SF) and temporal frequencies (TF): the transient system (originating in Y-cells of retina) sensitive to low SF and high TF, and especially activated by motion and flickering stimuli; the sustained system (X cells) sensitive to high SF and low TF, and especially activated by spatial properties of stimuli (Braddick, Campbell, & Atkinson, 1978; Kulikowski & Tolhurst, 1973; Maffei, 1978) (Fig. 1). On the other hand, at the behavioral level, there was a correlate of this neural specialization in speed of reaction to gratings of different SF: reaction times (RTs) were shorter to low SF (transient system) than to high SF (sustained system) (Breitmeyer, 1975; Breitmeyer & Ganz, 1976).
32
Mecacci
rr
..
_
-,
A
100
30-
\x\ ;~
I0
9 w,,,,q
e~--l
C 0
I
Spatial
3
10
Frequency
30
(cpd) B
r I-~ .Im,,l
C 0
3oo-
r~
-
0,"
,,
O
\
\ O~%
IO0:
% %
o~ % % i
30
• / 1
0
!
r
i
! i
1
Temporal
1
"
~
I
" !
!
!
frequency
! 'I !
10
''
"1
w--
30
(Hz)
Figure 1. Interaction between SF and TF as processed by the two visual channels or systems (perception of spatial pattern vs. temporal modulation) described by Kulikowski & Tolhurst (1973). Sensitivity for perception of spatial pattern is indicated by solid circles, sensitivity for detection of flicker is indicated by open circles. Stimulus was a grating of high SF (12 c/deg) in graph A and low SF (0.8 c/deg) in graph B. Contrast of gratings was sinusoidally modulated in time.
Temporal Frequency
33
Moreover results on the effects of spatial filtering of faces were showing that low SF were sufficient for processing accurately and efficiently this kind of visual information (Harmon, 1973; Tieger & Ganz, 1979). On the basis of these neurophysiological and psychophysical data, Sergent's path of reasoning was the following. Faces, with high-SF content removed by short exposure durations, continue to be recognized since the remaining low SF are sufficient for their processing. If RH is quicker than LH in responding to faces presented at short durations in the LVF, then RH is more sensitive than LH in responding to low SF. Consequently, "the RH does have an advantage over the LH in processing faces, not because it contains a specific 'face processor', but because of its better tuning to the low range of SF" (Sergent, 1982b, pp. 458-459). The same reasoning might be applied to assess the LH specialization in processing visual information based essentially on high-SF content. In short, Sergent (1983) considered physical properties of visual information, such as duration, luminance, contrast, etc. as a way to verify the general hypothesis that the two hemispheres are specialized in analyzing different SF ranges which are necessary for higher-order or categorical information processing (LH for high-SF processing and RH for low-SF processing). Thus the fundamental dichotomy depended on the space dimension, even if Sergent stressed the interaction of different visual input factors in determining the extent and direction of hemisphere asymmetries. Recent research has tried to confirm the spatial-frequency hypothesis (see reviews in the special issue of Brain and Cognition, 1986, vol. 5, no. 2; Christman, 1989; Hellige, 1990, 1993, 1995) and has approached the question through a different and updated neurophysiological framework. Kosslyn, Chabris, Marsolek, & Koenig (1992) have suggested that the contrast sensitivity peak is at low SF for RH and high SF for LH since the two hemispheres would process visual inputs preferentially from two distinct classes of receptive fields (large and overlapping fields for RH; small and nonoverlapping fields for LH) (Fig. 2). The hemispheric functional difference would emerge especially at low contrast (top horizontal dotted line in the figure), while at high contrast (bottom dotted line) the asymmetry would disappear because both hemispheres process inputs from the two types of receptive fields. The distinction between the two types of visual receptive fields is based on current neurophysiological evidence of two visual pathways (magnocellular and parvocellular) which do not correspond directly to
34
Mecacci
>,, 9~=,,N
> ,, t . , 4
1000 -
r,~
-9, , ,
,....,
. . . .
100 r
iO-
>
Ri Hemisphere ~
9I,,,,,I
Left Hemisphere
c~ I
0
=
0.1
"i"
1
i
10
.
.
.
.
!
100
Spatial frequency (c/deg) Figure 2. In the model by Kosslyn, Chabris, Marsolek, and Koenig (1992) "the modulation transfer functions of the high-level areas are slightly shifted, so that the peak sensitivity for the RH is at a lower SF, which could reflect its use of more input from larger receptive fields (and vice versa for the LH). At low contrast, the performances of the two hemispheres would be well separated, as indicated by the relatively small amount of overlap at the top horizontal dotted line. This difference could arise because the hemispheres differ in the degree to which they use inputs from units with different-sized receptive fields. However, at high contrast (when less sensitivity is required, as indicated by the bottom dotted line), both hemispheres use input from units with a wide range of receptive field sizes; hence the two distributions have a large amount of overlap, and the performance of the two hemispheres would not be well separated" (Kosslyn et al., 1992, pp. 572-573).
the past transient and sustained systems, although some affinities may be found. Magnocellular pathway (large receptive fields) would be more responsive to low-SF stimuli modulated at high TF, while parvocellular pathways (small receptive fields) would be more responsive to low TF (DeYoe & Van Essen, 1987; Livingstone & Hubel, 1988; Merigan & Maunsell, 1993). Livingstone has suggested that "the magnocellular
Temporal Frequency
35
ganglia may project preferentially to the fight hemisphere" (Kosslyn et al., 1992, p. 569). Thus new candidates as neurophysiological bases of hemispheric asymmetries in visual information processing might be represented by magnocellular and parvocellular pathways, engaged preferentially in RH and LH processing respectively (for a comparison between the spatial-frequency hypothesis and new neurophysiological models, see Brown & Kosslyn, 1995). Another important aspect concerned the processing level at which hemispheric asymmetries might emerge. Both Sergent's original version of the spatial-frequency hypothesis and the neurophysiological interpretation by Kosslyn and co-workers assume that hemispheric asymmetry arose "beyond the sensory level" in discrimination and categorization tasks (Sergent, 1982a, p. 266) or in "higher areas involved in memory and comparison" (Kosslyn et al., 1992, p. 572). On the contrary, at sensory levels in detection tasks, no hemispheric asymmetries emerged (as shown by Kitterle, Christman, & HeUige, 1.990). However in a series of electrophysiological works, evidence was obtained that hemispheric asymmetries might be found also at suprathreshold contrast levels. TF were a fundamental parameter to make these asymmetries emerge. More than a straight dichotomy in terms of SF ranges only (low SF in RH, high SF in LH), a functional asymmetry in terms of the interaction between SF and TF was observed. The main difference from the spatial-frequency hypothesis was that electrophysiological data showed an opposite direction of hemispheric specialization: high SF (and low TF) in RH and low SF (and high TF) in LH. This interaction was found depending on other factors like contrast level and field size, and subjects' handedness. Data were gathered generally in healthy subjects, and some relevant results were obtained also in brain-injured patients (a previous review in Mecacci, 1993).
Hemispheric asymmetries of spatio-temporal interaction: Eiectrophysiological evidence The electrophysiological technique of visual evoked potentials (VEPs) has been widely used in research on SF analysis (Regan, 1989). VEPs, that is the averaged brain electrical activity recorded on the scalp in response to a visual stimulus, have been employing in humans as a noninvasive technique to assess the neural correlates of several phenomena related to SF analysis. For instance, VEP amplitude augments as
36
Mecacci
function of contrast of sine-wave gratings, in the same way charge frequency of simple cells in cat's visual cortex increases as function of contrast (Campbell, Maffei, & Piccolino, 1973). Moreover psychophysical contrast sensitivity may be predicted by means of VEP technique: in fact the extrapolation of the regression line- between VEP amplitude and logarithm of contrast - to zero amplitude gives a contrast value which is very close to the threshold assessed by means of psychophysical methods (Campbell & Maffei, 1970; Fiorentini & Maffei, 1973). The curvilinear relationship between SF values and contrast sensitivity is verified plotting the VEP amplitude recorded for gratings of different SF and the same contrast value (Campbell & Maffei, 1970). The selective adaptation to a grating with a specific SF, found in humans by psychophysical methods (Blakemore & Campbell, 1969) and in the cat with cellular recordings (Maffei, Fiorentini, & Bisti, 1973), is confirmed by VEP recordings in humans (Mecacci & Spinelli, 1976). All these VEP data were obtained recording from one active electrode located on a central-posterior part of the scalp. There were not special reasons to record from two lateralized electrodes since asymmetrical electrophysiological responses were not expected for elementary visual stimuli such as gratings of different SF. Over the Seventies, indeed, the dichotomy between low-order and high-order processing was widely accepted in literature. In 1979 Moscovitch stated: "Information processing at an early precategorical, and presumably peripheral level is similar in both hemispheres. Thus both hemispheres are equally efficient in extracting and storing information about physical features of the stimulus array. Hemispheric asymmetries in information processing emerge only at a higher level of analysis in which relational or categorical features are represented". Recordings from lateralized electrodes were not made simply since asymmetrical brain responses for "early precategorical" visual information were not expected so that even the relevance of some results gathered from time to time on VEP hemispheric asymmetries was not sufficiently evaluated. These results were been founding in clinical neurophysiology field rather than in the area of basic vision research. VEP were usually recorded in response to checkerboards, following a phase-reversal rather than an on-off stimulation (Shagass, Amadeo, & Roemer, 1976). The check size (an indirect index of SF values; see Kelly, 1976), presented at low TF (1 Hz), was not always varied systematically so that the greatest amplitude was generally recorded over RH on temporal rather occipital
Temporal Frequency
37
leads (Vella, Butler, & Glass, 1972). However sometimes it may be found also on occipital derivations (Cohn, Kircher, Emmerson, & Dustman, 1985). Using high TF (6 Hz or above) the largest amplitude was recorded on LH as shown by Klemm et al. (1980, 1983). In our research VEP hemispheric asymmetries were found to depend on various parameters: electrode location; contrast; spatial-temporal frequency interaction; stimulation mode; visual field size; handdominance; eye-dominance (Mecacci & Spinelli, 1984, 1987; Mecacci, Spinelli, & Viggiano, 1990; Rebai, Bagot, & Viggiano, 1993; Rebai, Mecacci, Bonnet, & Bagot, 1986, 1989; Spinelli & Mecacci, 1986; Spinelli & Mecacci, 1990a, 1990b; Viggiano, 1992). Electrode location. Hemispheric asymmetries emerged from electrodes located on temporal lobes (2 cm above the inion and 5 cm laterally to the midline in the left or fight direction) while no asymmetry was recorded from electrodes located on occipital lobes (2 cm above the inion and 2-2.5 cm laterally to the midline in the left or right direction). This finding would indicate the existence of at least two distinct levels of processing: in a first level the visual information is processed symmetrically by the two hemispheres, while in a second level hemispheric asymmetries may emerge (Fig. 3). Contrast. In occipital leads VEP amplitude augments as function of contrast, while VEP latency decreases and VEP phase "advances" (Burr & Morrone, 1987) in both hemispheres (Fig. 4). In temporal leads VEP hemispheric asymmetries are not detectable neither near the threshold nor at low contrast values and become clear only at high contrast values, when VEP amplitude augments as function of contrast only in one hemisphere. On the contrary, there is no hemispheric asymmetry for VEP phase (Fig. 4). Two relevant conclusions may be drawn from these data. First, VEP hemispheric asymmetries emerge at high contrast values and disappear at threshold in agreement with several psychophysical data (Fiorentini & Berardi, 1984; Kitterle & Kaye, 1985; Kitterle, 1986). VEP data coming from temporal leads should indicate a neural activity partly different in somewhat from that engaged at threshold levels. Second, VEP hemispheric asymmetries may reflect the activity of different neuron populations of the two hemispheres in relation to contrast sensitivity. In fact, three limbs may be isolated in the curve relating VEP amplitude and contrast at low, intermediate, and high values. At low contrast the curve slope is shallow, at intermediate contrast the curve is high and at high contrast the slope has zero value or is
38
Mecacci
A
11
2c
lOcm
LH J4.
RH !
i
~1 500
.
0
msec
B 11
LH
--'
r
10
-
5
.
LH
r
E 8
..
r
RH
SO,b
~..,dP
;;>
v 0
, I
,
, I,..LI. S
A
10
Contrast
9
9 ! ....
SO
!
I00
;>
o II
10
50
I00
Contrast
F i g u r e 3. (A) VEPs by checkerboards phase-reversed at 8 Hz. Recordings on occipital (left half) and temporal (fight half) leads in LH and RH. In temporal leads, responses (VEP second harmonic: 16 peaks in 1 sec) were clear only in LH. (B) VEP recorded on temporal leads of LH and RH in two right-handed subjects for different values of stimulus contrast. Stimulus was a checkerboard phase-reversedat 8 Hz. VEP amplitude (microvolts) augmented as function of contrast in LH more than in RH (data by Mecacci & Spinelli reported in Mecacci, 1993).
Temporal Frequency
1 c/deg
A
2 c/deg
5 c/deg
e
Occipital
1
4
1
o~
9
. . . .
Jll+__,l I
I
l
1o
l[,
. . . .
1~o
l
CL
I0
100
'I
! 9
11oo
I
lo
Ioo
Temporal I
........
I tO
....
' - -,-~ii 100
4
2
2
0 ;0
100
Occipital
B 110
r~
4
11o
IN
9
0
a
1110 9
.110
.110
10
100
10
I00
Temporal
l
>
s n o "1
I91 0
180
,1110
9 q.'
|
"
10
-.-.I
100
9
I
-
+ ; ,-;~,
.....~
|0
* - : .:.+-,
'00
,~)
.....
-;;o
39
40
Mecacci
Figure 4 (previouspage). VEP amplitude in microvolts (A) and VEP phase in deg (B) as function of contrast and SF (1, 2, and 5 c/deg) in occipital and temporal leads of LH (open symbols) and RH (filled symbols) of one right-handed subject. VEP amplitude augmented symmetrically as function of contrast on occipital leads, while on temporal leads it augmented in LH more than in RH. No significant differences between the two hemispheres were found in VEP phase as function of contrast in both occipital and temporal leads.
negative. Saturation or adaptation mechanisms in cortical neurons were considered as explanation of the roll off at high contrast (Maffei, Fiorentini, & Bisti, 1973; Movshon & Lennie, 1979; Albrecht, Farrar, & Hamilton, 1984). The other two limbs, at low and intermediate contrast, were explained with the selective activity of two different visual subsystems and pathways: low contrast/central cells and high contrast/peripheral cells (Cambell & Maffei, 1970); low contrast/ sustained system and high contrast/transient system (Murray & Kulikowski, 1983); low contrast/magnocellular system and high contrast/ parvocellular system (Nakayama & Mackeben, 1982; Bobak, BodisWollner, Harnois, & Thornton, 1984; Tyler & Apkarian, 1985). Since the hemispheric asymmetries emerged particularly at medium-high contrast, the hypothesis was advanced (Spinelli & Mecacci, 1990a)that relative neural populations (peripheral, parvo or transient cells, according to the various classifications) might be work differentially in the two hemispheres. Note that the results we are discussing were obtained with gratings phase-reversed at high TF (8 Hz). Spatial-temporal interaction. The first striking phenomenon we observed (Mecacci & Spinelli, 1984) was that VEP were differentially affected in the two hemispheres (temporal leads) by varying the TF of stimuli. When gratings and checkerboards are phase-reversed, VEP contain a number of "main" peaks that is double the stimulation frequency (in response to a grating phase-reversed at 8 Hz, 16 peaks are detectable in 1-sec sweep of VEP). Thus as TF augments, the number of VEP peaks increases. This synchronization was observed much more clearly on the left temporal lead than on right one (in right-handed subjects) (Fig. 5). Subsequent research showed that the TF effect was not independent on the SF of stimuli. In fact the two hemispheres were found to be sensitive to particular combinations of SF and TF: at low TF larger amplitudes were recorded over RH especially for high SF; at high TF larger amplitudes were recorded over LH especially for low SF (Fig.
Temporal Frequency
41
6). However several works found that these interactive effects are more or less strong depending on other factors, as it will be specified later. Mode of stimulation. Two main modes of stimulation are used to record VEP by elementary stimuli such as gratings and checkerboards. In on-off mode, the stimulus is preceded and followed by a blank field (off part) generally having the same luminance and duration of the pattern. In phase-reversal mode, the pattern does not disappear from the screen: black stripes of gratings or black checks of checkerboards became white and vice versa at a fixed temporal rate. At a rate of one or two presentations of the stimulus each second, a typical negativepositive-negative (NPN) waveform is evident, with P peak having a latency of 100 msec about. Thus at low TF (1-2 Hz), on-off and phasereversal VEP have a similar waveform. From 4 Hz about on, only phasereversal VEP acquire a sinusoidal form with a number of peaks double than the TF of stimulation (Fig. 5). Kulikowski (1974, 1977) suggested that in on-off stimulation the sustained (X) system is particularly activated for responding to spatial or pattern components of the stimulus, while in phase-reversal the transient(Y) system is particularly engaged in responding to temporal or motion components. Results showing larger amplitude at high TF on LH were obtained with phasereversal stimulation (Mecacci & Spinelli, 1984; Rebai et al., 1986; Klemm et al., 1980, 1983) (Fig. 6). In the work by Rebai, Bagot, and Viggiano (1993) the effects of the two modes of stimulation were tested with gratings presented at a low TF (1 Hz). Larger VEP amplitudes were recorded when the on-off stimulation was used (especially for high SF), while no asymmetries emerged for the phase-reversal stimulation. In this latter condition asymmetry is found particularly at high TF, with larger amplitudes on LH (Viggiano, 1992). Finally, it is worth noting that hemispheric asymmetry in processing basic visual information was confirmed by research on regional cerebral blood flow (rCBF). Wendt, Risberg, Stenberg, Ros6n, and Ingvar (1994) measured rCBF in response to checkerboards phase-reversed at 1 Hz. Larger activation in RH than in LH was observed when subjects were sober, while the asymmetry disappeared after alcohol ingestion. Data were discussed in relation to our and other research groups' results on VEP asymmetry, and the hypothesis that ethanol inebriation unbalances the functional specialization of the two hemispheres was advanced.
42 Mecacci
LH
RH .5 1
2 3
4
6
8
2.5 0
500
msec Figure 5. VEP recorded in LH and RH of one right-handed subject. The stimulus was a checkerboard phase-reversed at different TF (from 0.5 to 12.5 Hz). The relation between VEP amplitude and TF is much more evident in LH than in RH (data by Mecacci & Spinelli reported in Mecacci, 1993).
Temporal Frequency
Occipital
43
Temporal
1.0
r, /
0.5
.
0
'k
//,
i
1.0 /
0.5
Is--S
l
m
0 1.0
9
!
I
t
t
It
t
..t
~
U
I I
E
I
0.5
,m
o
o ~
c~
0
1.0
0.5
0
1.0
I
I
!
!
t
f~
0
l t -!
t
t~_t!
,m
Q
0.5
!
4
t
f
6 8
I
tf
12 18 4
6 8
12 18
T e m p o r a l f r e q u e n c y (Hz) F i g u r e 6. Second harmonic amplitude (relative units) of VEP recorded on occipital and temporal leads of LH (filled circles) and RH (open circles). Stimulus was
44
Mecacci
a sine-wave grating phase-reversed at different TF. The top four pairs of graphs show the individual responses recorded in four subjects, graphs at bottom show the mean results over eight subjects (from Rebai, Mecacci, Bagot, & Bonnet, 1986).
Visual field size. As is well known in the literature, VEP amplitude augments as function of visual field size (Fukui, Kato, & Kuroiwa, 1986; Jeffreys & Axford, 1972; Yiannikas & Walsh, 1983). This amplitude increment was found to be asymmetrical in temporal leads when the visual field expanded beyond the foveal portions of the retina (Mecacci, Spinelli, & Viggiano, 1990). With a small visual field, larger amplitudes are recorded on RH for small check sizes (high SF) presented at low TF, while larger amplitudes are recorded on LH when the visual field is enlarged and large check sizes (low SF) are presented at high TF (Fig. 7). Hand-dominance and eye-dominance. The picture we traced above is made more complex by the fact that the degree and direction of VEP asymmetries depend on also the handedness and eye-dominance of subjects. In the work by Spinelli and Mecacci (1990) three groups of subjects were compared (14 right-handers with right-eye dominance, 12 right-handers with left-eye dominance, and 10 left-handers with left-eye dominance). The typical VEP asymmetry (larger amplitudes on RH for low TF and on LH for high TF) was found in right-handers with right eye-dominance. In right-handers with left-eye dominance the asymmetry was not significant. In left-handers an asymmetry having a different direction than in right-handers was found (larger amplitudes in RH for high TF). This inverse-direction asymmetry in left-handers was confirmed by Mecacci, Spinelli, and Viggiano (1990) and had been already described by Klemm et al. (1980). In the work by Rebai et al. (1989), LH superiority in left-handers emerged for low SF and high TF. Since left-handers were not tested in relation to their eye-dominance, the results may not be generalized to the population of left-handers. Also familial handedness seems to affect the degree of VEP asymmetries (Rebai, Lannou, Bernard, Bonnet, & Rocchetti, unpublished results).
Reading disability and impairment in processing basic spatiotemporal information The hypothesis that reading disability is due to deficits at low levels of visual processing had been accepted by some authors, although it was
Temporal Frequency 10
1 Hz
45
8 Hz
i
q~
I
a,
o,a,
,ml
E m >, 0
I L
I,
I
I
t
t
.
t
f
t
t
t
I
L__
t
___t..
t
Figure 7. VEP amplitude for checkerboards of different check sizes (in abscissa, from left to fight: 16, 11.5, 8, 5.6, 4, and 2.8 min of arc; contrast 50%) phase-reversed at 1 and 8 Hz, and presented in different conditions of visual field (from left to fight: full-field of 6 deg; foveal stimulation of 1 deg; full-field of 6 deg; annular stimulation given by coveting the central 1 deg of field). For small check sizes (corresponding to high SF) larger amplitudes were recorded in RH at 1 Hz especially with foveal stimulation. For large check sizes (low SF) larger amplitudes were recorded in LH at 8 Hz.
generally discarded during the Seventies (see relative reviews in Benton, 1975; Vellutino, 1979). However, in the early '80s, Lovegrove and coworkers proposed that a high percentage of specifically-disabled readers was affected by a low-level visual deficit, in particular in the transient system (after the paper in Science by Lovegrove, Bowling, Badcock, & Blackwood, 1980, many other works were published; see a first review in Lovegrove, Martin, & Slaghuis, 1986). Psychophysical data showed that disabled readers had a specific impairment in processing low SF and high TF. Considering that transient system has a crucial role in the integration of information from successive fixations (such as in reading: Breitmeyer, 1983), Lovegrove suggested that an abnormal functioning of the transient system might be one of the main cause of reading disability. Abnormalities of VEPs for gratings at low SF were recorded in children with reading disability (May, Lovegrove, Martin, & Nelson, 1991), supporting the hypothesis of a deficit in the transient system.
46
Mecacci
Also differences in VEPs for checkerboards of different check size were found between controls and poor readers by Mecacci, Sechi, & Levi (1983), and were interpreted as evidence of basic visual impairments in reading disability (the results are discussed also in the framework of their hypothesis by Lovegrove et al., 1986, pp. 255-256). The question was updated in terms of the distinction between the magnocellular and parvocellular pathways. In the work by Livingstone, Rosen, Drislane, and Galaburda (1991) on VEPs in dyslexic individuals, abnormalities of VEPs were found for checkerboards at high TF, low luminance and low contrast. Results were interpreted as the correlate of a loss of magnocellular neurons and the main cause of reading disability. This hypothesis was tested by Victor, Conte, Burton, and Nass (1993) recording transient and steady-state VEPs in dyslexics, patient controls, and normals. Contrast, luminance and TF of checkerboards were varied. However the results by Livingstone et al. (1991) were not confirmed. The hypothesis of an impairment of the magnocellular pathway in specific reading disability requires further investigation, using more adequate psychophysical or electrophysiological procedures to differentiate the contributions of the parvocellular vs. magnocellular visual pathways (see the work by Spinelli, Angelelli, De Luca, & Burr, 1996).
Evidence from brain-injured patients Impairment of SF and TF processing has been frequently shown in several cases of individuals affected by brain lesions. Some investigations were carried out by means of psychophysical methods, other works have been conducted by means of VEP recording (on the use of EP technique in neuropsychological research, see Viggiano, 1996). Spatial contrast sensitivity was found to be seriously reduced in patients with lesions in the visual cortex (Bodis-Wollner, 1976; Bodis-Wollner & Diamond, 1976; a review of these results in Regan, 1989). If a general impairment in the processing of spatial and temporal parameters of basic visual information may be expected, the effect of damaged hemisphere (left or right) on the type and severity degree of deficit remains an open question. In the work by Hess, Zihl, Pointer, and Schmidt (1990), contrast sensitivity for sine-wave gratings of different spatial frequencies was tested in 62 patients. No relation was found between the side of lesion and the range of SF (low or high) for which a deficit in contrast sensitivity was ascertained. Other works have found a
Temporal Frequency
47
special impairment of spatial contrast sensitivity in patients with RH than LH lesions. In the work by Kobayashi, Mukuno, Ishikawa, and Tasaki (1985) on 23 patients with unilateral lesions, the contrast sensitivity impairment was more serious in patients with right parieto-occipital lesions and hemispatial agnosia syndrome than in patients with lesion to the same areas of left hemisphere. In the work by Grabowska, Semenza, Denes, and Testa (1989), the performance in a discrimination test (subjects had to judge whether the SF of the second square-wave grating was the same as that for the first one presented 2 sec before for a 300msec duration) was more impaired in patients with fight lesions (N=19) than in patients with left lesions (N=24), and in control subjects (N=28). Another group of investigations focused on the impairment of contrast sensitivity and processing of SF and TF in patients affected by unilateral spatial neglect syndrome (briefly, hemineglect). Hemineglect is generally associated with lesions of right parietal areas, in particular the cortex of the right inferior parietal lobe. These patients ignore stimuli presented in the left part of visual field and do not explore this side of the space by means of eye movements. Although hemineglect syndrome is now generally explained by deficit at higher levels of information processing, some works have tried to verify the hypothesis that a basic sensory-perceptual deficit is the main cause of this neuropsychological impairment (for a review, Bisiach & Vallar, 1988). To test whether basic visual information is impaired in hemineglect patients, a series of investigations was carried out by Spinelli and her coworkers. In a first study (Spinelli, Guariglia, Massironi, Pizzamiglio, & Zoccolotti, 1990), contrast sensitivity and performance in a discrimination test (subjects had to judge whether the bars in the upper and lower parts of the stimulus - divided in two equal parts - were equal or different) were tested in 26 patients with lesions to right hemisphere (15 patients with hemineglect syndrome). A general impairment in contrast sensitivity, especially in the range of low SF, was found in brain-injured patients, but no special deficit was shown by hemineglect patients. In a second study (Spinelli & Zoccolotti, 1992), the contrast sensitivity for stationary and moving sine-wave gratings was psychophysically tested in 17 patients (5 with hemineglect) and 5 control subjects. Both the impairment of contrast sensitivity for low SF in patients and the absence of special deficit in hemineglect patients were confirmed. However the question remained whether impairments might be found in relation to TF processing interacting with SF analysis. A systematic investigation on
48
Mecacci
VEPs by checkerboards varying in both check size (range: 12-72 min of arc)and TF (phase-reversal mode; range: 1.96-16.6 Hz), was carried out in 20 patients (10 fight-damaged with hemineglect, 4 right-damaged without neglect, 6 left-damaged) and 6 controls (Viggiano, Spinelli, & Mecacci, 1995). In the condition of peripheral stimulation, no significant difference was found between patients and controls. In hemineglect patients, VEPs by stimuli presented to left (neglected) hemifield had smaller amplitudes compared to VEPs by stimuli to right hemifield, but the difference did not reach the statistical significance (p=0.08). In the condition of central stimulation, VEPs recorded in patients had smaller amplitudes than in controls (p<0.005), but the differences between the three groups of patients were not significant. The question was further investigated from the perspective of functional differences between parvocellular and magnocellular pathways. In a first study was found that VEPs of hemineglect patients showed a systematic delay of latency (Spinelli, Burr, & Morrone, 1994), and the hypothesis was advanced that the delay might be caused by damage to the magnocellular pathway, known to respond with shorter latency than the parvocellular pathway. In another work (Spinelli, Angelelli, De Luca, & Burr, 1996), VEPs by sine-wave gratings (0.6 cpd), modulated either in luminance (yellow-black) or in chromaticity (equiluminant red-green), were phase-reversed sinusoidally in contrast at 1-4 Hz (coloured stimuli in this temporal condition were expected to activate especially the parvocellular pathway) or 4-10 Hz (luminance stimuli in this other temporal condition were expected to activate the magnocellular pathway). VEPs were recorded in 10 patients with right-hemisphere lesions and hemineglect syndrome. VEPs by coloured stimuli (at low TF) were similar in amplitude in both hemifields. For VEPs by luminance stimuli a trend was observed to record smaller amplitudes for stimuli presented at high TF in the left (neglected) hemifield. For what regards the latency, no difference was found between the two hemifields when VEPs by coloured stimuli were recorded. On the contrary, a remarkable delay of latency was found for luminance stimuli: VEPs by stimuli in the left hemifield had longer latencies than VEPs by stimuli in the right. Spinelli et al. (1994) drew the conclusion that the parvocellular pathway (activated by chromatic stimuli, counterphased at low TF) was saved in hemineglect patients, while the magnocellular pathway (activated by luminance stimuli, at high TF) was damaged.
Temporal Frequency
49
Conclusion
Data on VEP by gratings and checkerboards show that hemispheric asymmetries are present also for the processing of elementary visual information. Instead of a strict hemispheric functional differentiation in terms of SF analysis, a more complex interaction between the spatial and temporal dimensions was found. Indeed the two-visual-system model shown in Fig. 1 seems to fit in with VEP results (high SF-low TF processed by RH and low SF-high TF processed by LH) more than the original Sergent's model (low SF and high SF processed by RH and LH, respectively, without considering the range of TF). This interactive model and its hemispheric functional organization has to be further verified in relation to current research on parvocellular and magnocellular pathways. Moreover, VEP data show that the lateralized processing of elementary visual stimuli does not occur at occipital early levels. In terms of Kosslyn's model (see Brown & Kosslyn, 1995), VEP hemispheric asymmetries associated to the processing of spatialtemporal interaction might arise in the "intermediate level" or preprocessing system in occipital-temporal regions. Here visual basic information would be organized into perceptual groups as inputs for the later high-level processing in the inferior temporal lobes, while at the preceding "low level" in occipital cortex no hemispheric differentiation of spatial-temporal analysis would be required. Impairments in spatialtemporal analysis, found in neuropsychological syndromes, might be located at this intermediate level, beyond the sensory level.
References
Albrecht, D. G., Farrar, S. B., & Hamilton, D. B. (1984); Spatial contrast adaptation characteristics of neurones in the cats visual cortex. Journal of Physiology, 34 7, 713-739. Benton, L. A. (1975). Developmental dyslexia: Neurological aspects. In W. J. Freidman (Ed.), Advances in neurology. Vol. 1: Current reviews of higher nervous system dysfunction. New York: Raven Press. Bisiach, E., & Vallar, G. (1988). Hemineglect in humans. In F. Boiler & J. Grafman (Eds.), Handbook of neuropsychology (pp. 195-222). Amsterdam: Elsevier. Blakemore, C., & Campbell, F. W. (1969). On the existence in the human visual system of neurones selectively sensitive to the
50
Mecacci
orientation and size of retinal images. Journal of Physiology, 203, 237-260. Bobak, P., Bodis-Wollner, I., Harnois, C., & Thornton, J. (1984). VEPs in humans reveal high and low spatial contrast mechanisms. Investigative Ophthalmology & Visual Science, 25, 980-983. Bodis-Wollner, I. (1976). Vulnerability of spatial frequency channels in cerebral lesions. Nature, 261, 309-311. Bodis-Wollner, I., & Diamond, S.P. (1976). The measurement of spatial contrast sensitivity in case of blurred vision associated with cerebral lesions. Brain, 99, 695-710. Bradd!ck, O., Campbell, F. W., & Atkinson, J. (1978). Channels in vision: Basic aspects. In R. Held, H. W. Leibowitz, & H.-L. Teuber (Eds.), Perception (Handbook of sensory perception. Vol. 8) (pp. 338). Berlin-Heidelberg-New York: Springer-Verlag. Breitmeyer, B. G. (1975). Simple reaction time as a measure of the temporal response properties of transient and sustained channels.
Vision Research, 15, 1411-1412.
Breitmeyer, 13. G. (1983). Sensory masking, persistence and enhancement in visual exploration and reading. In K. Rayner (Ed.), Eye movements in reading: Perceptual and language processes. New York: Academic Press. Breitmeyer, B. G., & Ganz, L. (1976). Implications of sustained and transient channels for theories of visual-pattern masking, saccadic suppression, and information processing. Psychological Review, 83, 1-36.
Brown, H. D., & Koss!yn, S. M. (1995). Hemispheric differences in visual object processing: Structural versus allocation theories. In R. J. Davidson & K. Hugdahl (Eds.), Brain asymmetry (pp. 77-97). _ Cambridge, Mass.: The MIT Press. Burr, D., & Morrone, C. (1987). Inhibitory interaction in the human visual system revealed in pattern evoked potentials. Journal of Physiology, 389, 1-21. Campbell, F.W., & Maffei, L. (1970). Electrophysiological evidence for the existence of orientation and size detectors in the human visual system. Journal of Physiology, 204, 635-652. Campbell, F. W., Maffei, L., & Piccolino, M. (1973). The contrast sensitivity of the cat. Journal of Physiology, 229, 719-731. Christman, S. (1989). Perceptual characteristics in visual laterality research. Brain and Cognition, 11, 238-257. Cohn N B., Kircher, J., Lmmerson, R. J., & Dustman, R. E. (1985). Pattern reversal evoked potentials" Age, sex, and hemispheric asymmetry. Electroencephalography and Clinical Neurophysiology, 62, 3 99-4O5. DeYoe, E. A., & Van Essen, D.C. (1988). Concurrentprocessing streams in monkey visual cortex. Trends in Neuroscience, I1, 219-226. Fiorentini, A., & Berardi, N. (1984). Right-hemisphere superiority in the discrimination of spatial phase. Perception, 13, 695-708. Fiorentini, A., & Maffel, L. (1973). Contrast perception and electrophysiological correlates. Journal of Physiology, 2 3 / , 61-69. Fukui, R., Kato, M., & Kuroiwa, Y. (1986). Effects of central scotomata on pattern reversal visual evoked potentials in patients with
Temporal Frequency
51
maculopathy and healthy subjects. Electroencephalography and Clinical Neurophysiology, 63, 317-326. Grabowska, A., Semenza, C., Denes, G., & Testa, S. (1989). Impaired grating discrimination following right hemisphere damage. Neuropsychologia, 2 7, 259-263. Harmon, L. D. (1973). The recognition of faces. Scientific American, 227, 71-82. Hellige, J. B. (1990). Hemispheric asymmetry. Annual Review of Psycho logy, 41, 55-80. Hellige, J. B. (1993). Hemispheric asymmetry: What's right and what's left. Cambridge, Mass.: Harvard University Press. Helhg" e, J. B. (1995). Hemispheric asymmetry for components of visual information processing. In R. J. Davidson & K. Hugdahl (Eds.), Brain asymmetry (pp. 99-121). Cambridge, Mass.: The MIT Press. Hess, R. F., Zihl, J., Pointer, J. S., & Schmidt, C. (1990). The contrast sensitivity deficit in cases with cerebral lesions. Clinical Vision Science, ~, 203-215. Jeffreys, D. A., & Axford, J. G. (1972). Source locations of patternspecific components of human visual evoked potentials. I. Component of striate cortical origin. Experimental Brain Research, 16, 1-21. Kelly, D. H. (1976). Pattern detection and the two-dimensional Fourier transform: Flickering checkerboards and chromatic mechanisms. Vision Research, 16, 277-287. Kitterle, F. L., & Kaye, R. S. (1985). Hemispheric symmetry in contrast and orientation sensitivity. Perception & Psychophysics, 37, 391-396. Kitterle, F. L. (1986). Psychophysics o f lateral tachistoscopic presentation. Brain and Cognition, 5, 131-162. Kitterle, F. L., Christman, S., & Hellige, J. B. (1990). Hemispheric differences are found in the identification, but not the detection, of low vs. high spatial frequencies. Perception & P_sychop_hysics,48, 297-306. Klemm, W. R., Gibbons, W. D., Allen, R. G., & Richey, E. O. (1980). Hemispheric lateralization and handedness correlation of human evoked "steady-state" responses to patterned visual stimuli. Physiological Psychology, 8, 409-416. Klemm, W.R., Goodson, R. A., & Allen, R. G. (1983). Contrast effects of the three primary colors on human visual evoked potentials. Electroencephalography and Clinical Neurophysiology, 55, 557-566. Kobayashi, S., Mukuno, K., Ishikawa, S., & Tasaki, V. (1985). Hemispheric lateralization of spatial contrast sensitivity. Annals of Neurology, 17, 141-145. Kosslyn, S. M., Chabris, C. F., Marsolek, C. J., & Koenig, O. (1992). Categorical versus coordinate spatial relations: Computational analyses and computer simulations. Journal of Experimental Psychology: Human Perception and Performance, 18, 562-577. Kulikowski, J. J. (1974). Human averaged occipital potentials evoked by pattern and movement. Journal of Physiology, 242, 70P-71P. Kulikowski, J. J. (1977). Separation of occipital potentials related to detection of pattern and movement. In J. E. Desmedt (Ed.), Visual evoked potentials in man: New developments (pp. 184-196). Oxford: Clarendon Press.
52
Mecacci
Kulikowski, J. J., & Tolhurst, D. J. (1973). Psychophysical evidence for sustained and transient detectors in human vision. Journal of Physiology, 232, 149-162. Livingstone, M., & Hubel, D. (1988). Segregation of form, color, movement, and depth: Anatomy, physiology, and perception. Science, 240, 740-749. Livingstone, M. S., Rosen, G. D., Drislane, F. W., & Galaburda, A. M. (1991). Physiological and anatomical evidence for a magnocellular defect in developmental dyslexia. Proceedings of the National Academy of Sciences of the U.S.A., 88, 7943-7947. Lovegrove, w. J., Bowling, A., Badcock, D., & Blackwood, M. (1980). Specific reading disability: Differences in contrast sensitivity as a ftinction of spatial frequency. Science, 210, 439-440. Lovegrove, W., Martin, F., & Slaghuis, W. (1986). A theoretical and experimental case for a visual deficit in specific reading disability. Cognitive Neuropsychology, 3, 225-267. Maffei, L. (1978). Spatial frequency channels: Neural mechanisms. In R. Held, H. W. Leibowitz, & H.-L. Teuber (Eds.), Perception (Handbook of sensory perception. Vol. 8) (pp. 39-66). BerlinHeidelberg, New York: Springer-Verlag. Maffei, L., Fiorentini, A., & Bisti, S. (1973). Neural correlate of percej~tual adaptation to ,gratings. Science, 182, 1036-1038. May, J. ~J., Lovegrove, W. J., Martin, F., & Nelson, P. (1991). Patternelicited visual evoked potentials in good and poor readers. Clinical Vision Science, 6, 13 l- 136. Mecacci, L. (1993). On spatial frequencies and cerebral hemispheres: Some remarks from the electrophysiological and neuropsychological points of view. Brain and Cognition, 2 2, 199-212. Mecacci, L., Sechi, E., & Levi, G. (1983). Abnormalities of visual evoked potentials by checkerboards in children with specific reading disability. Brain and Cognition, 2, 135-143. Mecacci, L., & Spinelli, D. (1976). The effects of spatial frequency adaptation on human evoked potentials. Vision Research, 16, 477479. Mecacci, L., & Spinelli, D. (1984). Hemispheric asymmetry for basic functions: Electrophysiological evidence. International Journal of Psychophysiology, 2, 223 [Abstract]. Mecacci, L., & Splnelli, D. (1987). Hemispheric asymmetry of pattern reversal visual evoked potentials in healthy subjects. International Journal of Psychophysiology, 4, 325-328. Mecacci, L., Spinelli, D., & Viggiano, M. (1990). The effects of visual field size on hemispheric asymmetry of pattern reversal visual evoked potentials. International Journal of Neuroscience, 51, 141-151. Meiigan, W. H., & Maunsell, J. H. R. (1993). How parallel are theprimate visual pathways? Annual Review of Neuroscience, 16, 369-402. Moscovitch, L. (1979). Information processing and the cerebral hemispheres. In M. S. Gazzaniga (Ed.), Handbook of behavioral neurology, vol. 2. Neuropsychology (pp. 379-446). New York: Plenum Press. Movshon, J. A., & Lennie, P. (1979). Pattern selective adaptation in visual cortical neurones. Nature, 278, 850-852.
Temporal Frequency
53
Murray, I. J. & Kulikowski, J. J. (1983). VEPs and contrast. Vision Research, 2 3, 1741-1743. Nakayama, K., & Mackeben, M. (1982). Steady state visual evoked ,potentials in the alert primate. Vision Research, 22, 1261-1271. Rebai, M., Bagot, J.-D., & Viggiano, M. P. (1993). Hemispheric asymmetry In transient visual evoked potentials induced by the spatial factor of the stimulation. Brain and Cognition, 23, 263-278. Rebai, M., Lannou, J., Bernard, C., Bonnet, C., & Rocchetti, G. (unpublished results). Hemispheric asymmetries of visual evoked ,po.tentials in relation to handedness and familial left-handedness. Rebai, M., Mecacci, L., Bonnet, C., & Bagot, J.-D. (1986). Hemispheric asymmetries in the visual evoked potentials to temporal frequency: Preliminary evidence. Perception, 1-5, 589-594. Rebai, M., Mecacci, L., Bonnet, C., & Bagot, J.-D. (1989). Influence of spatial frequency and handedness on hemispheric asymmetry in visually steady-state evoked potentials. Neuropsychologia, 27, 315324. Regan, D. (1989). Human brain electrophysiology: Evoked potentials and evoked magnetic fields in science and medicine. New York/Amsterdam/London: Elsevier. Sergent, J. (1982a). The cerebral balance of power: Confrontation or cooperation? Journal of Experimental Psychology: Human Perception & Performance, 8, 253-272. Sergent, J. (1982b). Theoretical and methodological consequences of variations in exposure duration in visual laterality studies. Perception & Psychophysics, 31, 451-461. Sergent, J. (1983). The role of input in visual hemispheric asymmetries. Psychological Bulletin, 93, 481-512. Sergent, J., & Hellige, J. B. (1986). Role of input factors in visual-field asymmetries. Brain and Cognition, 5, 174-199. Shagass, C., Amadeo, M., & Roemer, R. A. (1976). Spatial distribution of potentials evoked by half-field pattern-reversal and pattern-onset stimuli. Electroencephalography and Clinical Neurophysiology, 41, 609-622. Spinelli, D., & Mecacci, L. (1986). Hemispheric asymmetry of visual evoked potentials. Perception, 15, 39 [Abstract]. Spinelli, D., & Mecaccl, L. (1990a). Contrast and hemispheric asymmetry: An electrophysiological investigation. International Journal of Neuroscience, 50, 113-119. Spinelli, D., & Mecacci, L. (1990b). Handedness and hemispheric asymmetry of pattern reversal visual evoked potentials. Brain and Cognition, 13, 193-210. Spinelli, D., & Zoccolotti, P. (1992). Perception of moving and stationary gratings in brain damaged patients with unilateral spatial neglect. Neuropsychologia, 30, 393-401. Spinelli, D., Angelelli, P.,De Luca, M., & Burr, D. (1996). VEP in neglect patients have longer latencies for luminance but not for chromatic patients. NeuroReport, 7, 815-819. Spinelli, D., Burr, D. C., & Morrone, C. (1994). Spatial neglect is associated with increased latencies of visual evoked potentials. Visual Neuroscience, 11, 909-918.
54
Mecacci
Spinelli, D., Guariglia, C., Massironi, M., Pizzamiglio, P., & Zoccolotti, P. (1990). Contrast sensitivity and low spatial frequency discrimination in hemi-neglect patients. Neuropsychologia, 28, 727-732. Tieger, T., & Ganz, L. (1979). Recognition of faces in the presence of two-dimensional sinusoidal masks. Perception & Psychophysics, 26, 163-167. Tyler, C. W., & Apkarian, P. A. (1985). Effects of contrast, orientation and binocularity in the pattern evoked potential. Vision Research, 25, 755-774. Vella, E. J., Butler, S. R., & Glass, A. (1972). Electrical correlate of right hemisphere function. Nature New Biology, 236, 125-126. Vellutino, F. R. (1979). Dyslexia: Theory and research. London: The MIT Press. Victor, J. D., Conte, M. M., Burton, L., & Nass, R. D. (1993). Visual evoked potentials in dyslexics and normals" Failure to find a difference in transient or steady-state responses. Visual Neuroscience, 10, 939-96. Viggiano, M. P. (1992). Effetti della modalit~ di stimolazione sui otenziali evocatl visivi da reticoli. Archivio di Psicologia, urologia e Psichiatria, 53, 231-239. Viggiano, M. P. (1996). Event-related potentials in brain-injured patients with neuropsychological disorders: A review. Journal of Clinical and Experimental Neuropsychology, 18. Viggiano, M. P., Spinelli, D., & Mecacci, L. (1995). Pattern reversal visual evoked potentials in patients with hemineglect syndrome. Brain and Cognition, 2 7, 17-35. Wendt, P.E., Ristierg, J., Stenberg, G., Ros6n, I., & Ingvar, D.H. (1994). Ethanol reduces asymmetry of visual rCBF responses. Journal of Cerebral Blood Flow and Metabolism, 14, 963-973. Yiannikas, C., & Walsh, J. C. (1983). The variation of the pattern shift visual evoked response with the size of the stimulus field. Electroencephalography and Clinical Neurophysiology, 55, 427-435.
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
55
Chapter 3
lnterhemispheric Transfer of Spatial and Temporal Frequency Information. Nicoletta Berardi*,t Adriana Fiorentini* *Istituto di Neurofisiologia C.N.R., Pisa, Italy and tDipartimento di Psicologia Generale Proc. Svil. Soc., Universita' di Firenze, Italy The two cerebral hemispheres are not separate entities: interhemispheric commissural fibers connect both cortical and subcortical regions of the brain. The classical experiments on split-brain patients suggest that a role of commissural fibers is to integrate the activity of the two hemispheres and to compensate for the presence of hemispheric specializations. The loss of interhemispheric communication is evident in the inability of the split-brain subjects to interrelate stimuli seen in left and right visual fields or to verbally name or describe objects and patterns presented in the left half field of vision (Gazzaniga, Bogen & S perry, 1965). The principle of supplemental complementarity (Sperry, 1962) proposes that the pattern of organization of the commissural connections is a means by which the activity of each cerebral hemisphere is supplemented in orderly manner with different and complementary information about concurrent activity in the other hemisphere. This is important in view of the fact that each hemisphere receives information
56
Berardi and Fiorentini
about the opposite half of the sensory space and controls the contralateral half of the musculature. In the visual system, where each visual hemifield is represented in the contralateral hemisphere, the interhemispheric commissures may play a role for those visual functions that exhibit a hemispheric specialization, as the recognition of complex forms and visual spatial localization (see Davidoff, 1982, and Chapters 1 and 2, this Volume). Another role of the visual commissures is related to the requirement of assuring stimulus equivalence across retinal translation: the interhemispheric commissures may grant continuity across the vertical meridian. Most of our knowledge about the properties of interhemispheric transfer of visual information is derived from anatomical, electrophysiological and behavioural data on cats and monkeys. Before reviewing and discussing the experiments on interhemispheric transfer in humans, it is useful to briefly summarize the relevant findings in animals.
Properties of interhemispheric commissures in mammals In mammals, interhemispheric fibers connect areas of the cerebral (neo)cortex, the hippocampal formations and various subcortical structures, including the colliculi. The interhemispheric commissure of the neocortex is the corpus callosum, formed by a very large number of fibers originating from a relatively small proportion of neurons in each of various cortical regions. Its posterior portion, the splenium, connects visual areas of the cortex. The anterior commissure, that is much smaller than the callosum, connects regions of the paleocortex (the amigdalae and the olfactory bulbs), but in the primate also contains fibers originating in the visual (inferotemporal) cortex. In non-human primates as in other mammals, callosal connections between regions of the visual cortical areas of the two hemispheres are usually restricted to a strip along the projection of the vertical meridian of the visual field (Van Essen & Zeki, 1978). This is particularly true for the region of transition between striate (V 1 or 17) and prestriate (V2 or 18) visual areas, where the vertical meridian of the visual field is represented (Fig. 1). In V 1 (and V2) of either cerebral hemisphere, within a few millimeters from the V1-V2 border, there are callosal projecting neurons that send fibers to neurons of the contralateral V1 (V2) and callosal recipient neurons that receive callosal input from neurons of the contralateral V 1 (V2) (Kennedy, Dehay & Bullier, 1986).
Interhemispheric Transfer
A
FRONTAL ~
PARIETAL
LO
BE
TEMPORAL [
~!!!!J~i~ii~~ LOBE
B
Figure 1: Cortical visual areas and callosal connections.
57
58
Berardi and Fiorentini
The receptive field (i.e. the area of the visual field where a visual stimulus is able to elicit a response from the neuron) of callosal recipient neurons are formed by a portion in the contralateral visual field, due to the direct retino-geniculate input, and a callosal portion that extends into the ipsilateral visual field (see Fig. 2) (Berlucchi & Rizzolatti, 1968; Berlucchi, Tassinari & Antonini, 1986; Berlucchi et al., 1987; Berlucchi & Antonini, 1989; Gross & Mishkin, 1977; Payne, 1990). Thus the callosal connections extend into the ipsilateral hemifield the representation of the contralateral hemifield in either hemisphere. At the border between visual areas V1 and V2, the extent of callosal connections in terms of distance from the vertical meridian and of the size of the callosal receptive fields produces a vertical strip of the visual field receiving bilateral inputs. This strip extends for a few degrees on either side of the vertical midline in the cortical projection of the cat's area centralis, and is wider for the projection of more peripheral retinal regions (Payne, 1990). In the monkey, commissural receptive fields straddling the vertical meridian have been described for the inferotemporal cortex (Gross & Mishkin, 1977). The anatomical data available for the callosal projections at the transition zone between V1 and V2 suggest that here the width of the vertical strip with bilateral inputs corresponds to a visual angle considerably smaller in the monkey than in the cat (Kennedy, Dehay & Bullier, 1986). The receptive fields of most callosal neurons are binocular and their direct and callosal inputs respond preferentially to stimuli of the same orientation (see Berlucchi et al. 1987 for review). Callosal neurons seem to be grouped in columns and thus integrated into the columnar architecture of the visual cortex (Kennedy et al., 1986). This leaves open the possibility that adjacent callosal columns are joined by horizontal intracortical fibers, that may contribute to extending the areas of the visual field that are bilaterally interconnected. The interhemispheric commissures are not the only means that may assure a bilateral representation along the vertical midline. Anatomical evidence indicates that in the nasal retina of the monkey, ganglion cells that project contralaterally are mixed with cells projecting ipsilaterally, and this region of naso-temporal overlap should assure that a small area along the midline of the visual field is represented in both hemispheres (Stone, Leicester & Sherman, 1973; Bunt, Minckler & Johanson, 1977; Leventhal, Ault & Vitek, 1988; Fukuda et al. 1989). Whether these anatomical connections have a functional role is uncertain, however.
o ~
o
~.
s~ ~.
oo
>
o
o
~.
Q~
o~
,,I ~
pJ
L
C~
m
I l l
~
Q~
u
m
s~
mt
Dmlo CJ~
qn~b
60
Berardi and Fiorentini
Using a technique that reveals the metabolic activity of cortical neurons, Tootell et al. (1988) did not find any evidence for a dual representation of the vertical meridian in the visual cortex of the monkey, at the V l-V2 border. A projection of the foveal region of the nasal retina onto both hemispheres, if present in man, should result in a "macular sparing" in commissurotomy patients: even after section of the cortical interhemispheric commissures, a dual representation of a midline region should be preserved. Although examples of foveal sparing were reported following lesions of the geniculo-striate pathway in humans (Koerner & Teuber, 1973), several attempts to find perceptual evidence of nasotemporal overlap either in normal subjects (reaction time experiments, Harvey, 1978; Haun, 1978) or in subjects lacking a corpus callosum (Lines, 1984; Fendrick & Gazzaniga, 1989; Sugishita et al., 1994) failed to find evidence for a region of dual representation extending more than 1/4-1/2 deg from the fixation point. This naso-temporal overlap within the fovea should be sufficient to ensure that information about tiny binocular disparities, as required, for instance, for fine stereopsis, reach either hemisphere, but, taking into account the small size of foveal receptive fields, it is obviously not sufficient for tasks requiring a somewhat larger extension of the bilateral area. Thus, interhemispheric commissures may play the major role in providing perceptual continuity at the vertical meridian. If so, it is important to know the information transfer capacity of the interhemispheric visual connections, which may not be the same for the fibers interconnecting primary visual areas and those interconnecting extrastriate visual areas of increasing hierarchical order (Fig. 1). This has to be expected for at least two reasons. First, different areas may have different functional roles, being specialized for the processing of different parameters of the visual stimuli and receiving specific subsets of the V 1 input. Second, the organization and size of the receptive fields vary substantially from the primary visual area to the extrastriate areas in the temporal and parietal streams. For instance, in primates, the neurons of the inferotemporal cortex have very large receptive fields which include the fovea but extend far from the vertical meridian into both the contralateral and the ipsilateral visual field (Gross, Bender & Mishkin, 1977) and seem to be preferentially stimulated by specific complex forms (Fujita et al., 1992; Tanaka, 1993; Wang et al., 1996). Both the callosum and the anterior commissure assure the interhemispheric
Interhemispheric Transfer
61
connections in this area and subserve transfer of learning of visual discrimination tasks in monkeys (Seacord, Gross & Mishkin, 1979). As regards the callosal connections at the border between V1 and V2, areas where neurons have well defined properties of spatial and temporal filters, it makes sense to try and define the interhemispheric transfer capacity in terms of filtering properties, i.e. in terms of spatial and temporal frequencies transmitted. For the aforementioned reasons, however, one has to bear in mind that whatever the limits found at the level of the callosal transfer at the V1/V2 border, these do not necessarily apply to the properties of connections between visual areas at further levels of processing. I. Interactions between sinusoidal stimuli presented in the left and right visual field. A. Visual Evoked Potential experiments in human subjects In man, electrophysiological evidence suggests that the response to a simple visual stimulus (sinusoidal grating) presented laterally can be influenced by a stimulus presented in the opposite visual field, and that these inter-stimulus influences are restricted to an area extending only a few degrees from the midline (Berardi et al. 1989). The evidence is derived from experiments where Visual Evoked Potentials (VEP)were recorded in response to a horizontal sinusoidal grating reversed in contrast at a fixed temporal frequency and located in the right or left visual field at a small distance (30 min) from the fixation point. The amplitude of the response to this grating of moderate contrast was substantially reduced in the presence of a second horizontal grating of high contrast, reversed at a somewhat different temporal frequency, and located on the opposite side of the fixation point (Fig. 3). For this effect to occur, the high contrast grating had to be displaced no more than 2 deg from the vertical meridian and to be accurately aligned with the contralateral grating along the horizontal meridian. The effect of the high contrast grating on the response to the low contrast grating was present both when the former was in the left visual field and the latter in the fight, and vice versa, and also when the gratings were presented dichoptically, i.e. separately to the two eyes. The latter finding indicates that the interaction is generated at a cortical level.
62
Berardi and Fiorentini
TEST
CONTROL
CONTROL
B'[-'I BINOCULAR
DICHOPTIC
(
0
...... II'
T 2
l
l
I
T
0
T 2
rT.F7
2 #V[
I
7
I
I
0
T
__T__ 2
"
]
T
I
50 ms I--
0
~i
T 2 _
I
T
l-
0
I
I
r
r
Figure 3: Examples of VEP responses to a reference grating under conditions of dichoptic vs. binocular viewing as a function of contralateral stimulation.
The most likely explanation for this effect is that it is subserved by connections between cortical regions of the two hemispheres and the callosal connections seem to be good candidates to account for the effect. An explanation in terms of a bilateral retinal projection of the central region of the visual field seems unlikely because of the relatively
Interhemispheric Transfer
B
A Z
0
63
2_
:::) Z ILl I,"-"
"=: 1.5_ I11 I" I.U
..,,
I
!
I
1
2
4
SPATIAL
FREQUENCY
(c/deg)
4
8 10
7 20
TEMPORAL FREQUENCY (Hz]
Figure 4: Attenuation of VEP response by contralateral stimulus as a function of spatial and temporal frequency.
large interstimulus distance (2.5 deg) still capable of producing an interaction (see Berardi et al. 1989, for discussion of this point). Consistent with this interpretation, in a subject with callosal agenesis no VEP interaction was found between gratings located in the opposite visual hemifields, under conditions producing substantial interaction for normal subjects (Fiorentini et al. 1992). The influence of the contralateral stimulus on the VEP response to the low contrast grating has been found to occur only within a limited range of spatial and temporal frequencies (Fig. 4): gratings with a spatial frequency of 4 c/deg or more and temporal frequency of 10 Hz or more fail to produce any VEP reduction across the midline. This contrasts with interactions between spatially superimposed gratings, which take place across a wide range of spatial and temporal frequencies (Fiorentini, Pirchio & Spinelli, 1983, Regan, 1983) and suggests a limitation in the interhemispheric transfer of visually elicited neural activity at the level responsible for the generation of VEPs in response to sinusoidal stimuli.
64
Berardi and Fiorentini
B. Psychophysical experiments in human subjects The possibility for sinusoidal stimuli located in the left and right visual field, close to the vertical midline, to interfere with each other, as observed in VEP experiments, has been demonstrated also in psychophysical experiments, with findings that confirm limitations in the interhemispheric transfer of spatial and temporal frequency information. The detection threshold for a 2 c/deg grating separated 0.5 deg from the midline is increased by about 0.2 log units in the presence of a high contrast grating in the mirror symmetric position from the fixation point (Berardi et al. 1989). No such increase in detection threshold is usually observed for non-superimposed gratings located in other positions. Following this report, an extensive investigation (Kitterle, Christman & Conesa, 1995) provided further evidence that sinusoidal gratings of low spatial (1 c/deg) and low temporal frequency (5 Hz) in the opposite visual fields, close to the midline, interact with each other, while this does not occur at a higher spatial (9 c/deg) or temporal frequency (10 Hz). This evidence was derived using reaction time (RT), where RT for the detection of a near-threshold grating located in one visual field was measured in the presence of an "adapting" grating of high contrast located in the mirror-symmetric position of the opposite visual field. For gratings of low spatial and temporal frequency, RT significantly increased with respect to that measured in the absence of the adapting contralateral stimulus (Fig. 5). Since RT increases when contrast is reduced, this finding was interpreted as being due to a reduction in perceived contrast of the test grating produced by the contralateral adapting stimulus. In addition, in both detection and identification tasks, the simultaneous brief presentation of two gratings of low spatial frequency at 0.5 deg on either side of the vertical midline, yielded shorter reaction times than expected merely on the basis of probability summation, indicating that the two stimuli "coactivated" in producing the response. Thus, both facilitory and inhibitory interactions may occur between sinusoidal stimuli presented symmetrically and at short distances from the vertical meridian. In either case, the effects are conditional upon the parameters of the stimuli, suggesting that the cross-talk between the two hemispheres in these types of task is limited to sinusoidal gratings of relatively low spatial and temporal frequency and, according to the VEP evidence, to mirror-symmetric and well aligned stimuli. There seems to be no hemispheric asymmetry for any of the observed effects.
Interhemispheric Transfer
65
500-
A
475
o.:/ 4~0 450
42S-
400 400.
350
.
.
.
.
.
.
.
.
375
.
0
4
~
i
---~o
SPATIAL FREQUENCY
Figure 5: Reaction time as function of target spatial frequency in the absence (circles) or presence (squares) of a 1 c/deg (left panel) or 9 c/deg (fight panel) adapting grating presented to opposite visual field. Open symbols indicate target grating in RVF and adapting grating in LVF; filled symbols indicate target grating in LVF and adapting grating in RVF. Based on data from Kittede, Christman, and Conesa, 1995.
C. Experiments in cats and monkeys
The electrophysiological and psychophysical findings in humans are in line with the properties of callosal connections between the striate and prestriate areas in cat and monkey, showing that the callosal neural messages elicited by sinusoidal patterns are considerably attenuated in comparison with the direct geniculo-cortical inputs to the striate cortex (Fig. 6), especially at the high spatial and temporal frequencies (Berardi et al. 1987; Berardi et al. 1988a). It is tempting therefore to assume that the interactions between sinusoidal stimuli observed in human subjects reflect primarily the filtering properties of callosal connections at the striate and prestriate level. This is also the level where the patternreversal VEPs (or most of their power) are generated (Maier et al., 1987). In the cat, the spatial frequency limits in the electrophysiological properties of callosal striatal connections seem to be reflected also in
66
Berardi and Fiorentini SPLIT
CHIASM
eye
MONKEY
9 ipsilateral o contralateral
50-
20-
10-
0 0
v
w r7
5_
F__I r~ < &. w
25Hz
'~
1-
0.5-
O.2-
t 0.3
i
i lfri 0.6
1. . . . . 1
=
l 3
'
'1
SPATIAL FREQUENCY (c/deg)
-
6
I 1
1 2
I
1
~
5
,t
I,!
10
I 20
TEMPORAL FREQUENCY (Hz)
Figure 6: Cortical VEP responses in split-chiasm monkeys as function of callosal versus direct geniculo-cortical input.
behavioural tasks implying discrimination of orientation of sinusoidal gratings (Berardi et al. 1988a). It is possible that the visual interhemispheric transfer for other types of tasks is mediated by callosal connections between visual areas beyond V 1 and V2. A recent report suggests that, in man, the interhemispheric transfer of visual responses to localized light flashes as revealed by latencies of evoked potential components in a visuomotor (RT) task is likely to be mediated by callosal connections between visual areas anterior to the occipital lobe (Ipata et al., in press). Whether interhemispheric connections between higher cortical areas in non-human primates suffer from more or less severe limitations in the
Interhemispheric Transfer 67 spatial and temporal frequency domains remains to be investigated. In the cat, no such limitations have been found in area 19, where the neurons have lower spatial resolution and larger receptive-field size than in the primary visual cortex (Guillemot, personal communication). In conclusion, the results from VEP and RT experiments in human subjects agree with findings in animals at the V l-V2 border, indicating the attenuation in the interhemispheric transfer of visual information for sinusoidal stimuli of high spatial and temporal frequency. In man, no visual field preference was observed for the sinusoidal stimuli employed in either the psychophysical or the electrophysiological experiments. In the cat, it is the property of the interhemispheric connection at the level of areas 17-18 to be reflected in a simple behavioural task.
II. Discrimination of spatial phase in complex gratings presented in the left or right visual field A. Limits in the lateralization of spatial phase discrimination Another way to investigate interhemispheric transfer of visual information in humans is by testing visual tasks that show a clear lateralization, with either a left or a right visual field advantage. For instance, the presence of lateralization in a visual task for short, but not for longer presentation times, is suggestive of an interhemispheric transmission time that has to be exceeded in order for the information to be transferred from the more competent hemisphere to the other. In a visual task that we have been studying (Fiorentini & Berardi, 1984; Berardi & Fiorentini, 1987), there is a left field advantage for stimuli presented at some distance from the vertical meridian, but the advantage may disappear close to the vertical meridian. The stimuli were complex gratings consisting of the sum of two harmonics (of spatial frequency f and 3fand contrast ratio 3 to 1, as the first two harmonics of a square wave), superimposed with variable relative phases. The task consisted in discriminating gratings differing for the spatial phase of their harmonics. The discrimination of these complex gratings is performed more accurately in the left than in the right visual field when the gratings are separated 5 deg from the vertical meridian (Fig. 7). This left visual field/ fight hemisphere advantage has been replicated in a number of subjects
68
Berardi and Fiorentini -1
90
c0---,
80 113
,._._
o L. ,._..
70
',oo I
o
O
o ~ 60
50
-
50
60
70
80
90
100
LF correct responses (%) Figure 7: Discrimination of complex gratings as a function of left versus right visual field. Diamonds - left-handed subjects; circles - right-handed subjects; open symbols - fight eye viewing; filled symbols - left eye viewing.
(Fiorentini & Berardi, 1984). At 1 deg from the vertical meridian, however, the performance with gratings of fundamental frequency 2 c/deg is as accurate in the fight as in the left visual field (Fig. 8). At a higher spatial frequency there is a left field advantage even for stimuli close to the meridian, unless the contrast of the gratings is very high (Berardi & Fiorentini, 1987). An explanation of these findings in terms of a bilateral representation of the central retina would require that the ipsilateral projection of either retina extends up to more than 1 deg from the vertical midline, which seems unlikely, on the basis of animal data (Stone et al., 1973; Bunt et al., 1977; Leventhal et al., 1988; Fukuda et al., 1989; Tootell et
Interhemispheric Transfer 0
0
NB
60"
w ,..I
69
0"
r
LF
RF
LF
RF
Z~ ~z n
--
ILl
~OLU 160"_ r. _~.
AF
u.. U..
120"_ LU
22) r.,O
: ~ ~<~ Z
:~
80"
~
40'_ m
O'_
LF
RF
LF
RF
Figure 8: Discrimination of complex gratings as a function of visual field and retinal eccentricity for two observers.
al., 1988). It has to be ruled out also on the basis of measurements of RT for the discrimination of complex gratings at various distances from the vertical meridian: the RTs are longer in the left than in the right visual field, even for stimuli of low spatial frequency at 1 deg from the midline, thus indicating a functional dissociation of the hemispheres for this task, even near the vertical meridian (Berardi, Fiorentini & Gravina, 1988b). Therefore, as in the experiments with sinusoidal gratings, the results with complex gratings are consistent with an interhemispheric transfer of visual information limited to a narrow region astride the
70
Berardi and Fiorentini
vertical midline. But here a~ain there is evidence suggesting that the transfer not only is restricted to a narrow strip along the midline, but is also limited in terms of spatial frequency and/or contrast of the stimuli. From this point of view, the analogy with the properties of callosal connections in animals is even more compelling, because of its dependence on the stimulus contrast. In callosal receiving neurons at the 17-18 border, the contrast sensitivity and the contrast gain of the callosal input is considerably lower than that of the direct geniculo-cortical input (Berardi et al. 1987). A contribution of the other important structure connecting the two hemispheres, the anterior commissure, to the interhemispheric transfer for the discrimination of complex gratings cannot be excluded, but seems unlikely because of the properties of the cells of the visual areas that in the monkeys are connected by the anterior commissure (see Introduction).
B. Interhemispheric transfer of perceptual learning for spatial phase discrimination of complex gratings Another way to study the interhemispheric transfer of visual information is to test how visual tasks learned by one hemisphere transfer to the other hemisphere. This can be done in animals with longitudinal section of the optic chiasm, in which each hemisphere receives an input only from the ipsilateral eye. It has been shown, for instance, that in split-chiasm cats form discrimination learned with one eye transfers to the other eye, provided the corpus callosum is intact (see Berlucchi, 1990 for review). Interestingly in the cat the callosal fibers of areas 17 and 18 are not crucial for this transfer, which seems to be mediated by interhemispheric connections between more anterior visual cortical areas (Berlucchi et al. 1978; Berlucchi, 1990). In primates, pattern discrimination learned with one hemisphere also transfers to the other, and both the splenium of the corpus callosum and the anterior commissure contribute to the transfer. This was shown in split-chiasm animals (Sperry, 1961; Black & Myers, 1964) and in animals with intact chiasm (Eacott & Gaffan, 1989) in which one or both commissures had been sectioned. In monkeys, differently from the cat, the striate and prestriate visual areas are essential for form discrimination, but to ensure the interhemispheric transfer of form discrimination, the inferotemporal cortex must also be intact (Seacord, Gross & Mishkin, 1979). This area is thought to be crucial for object vision in primates.
Interhemispheric Transfer 71 While under normal conditions, visual memory (e.g. recognition of complex patterns) is not restricted to the area of the visual field where the stimuli have first been presented, visual learning implying a nondeclarative type of memory (perceptual learning) may have purely local effects. Among the visual tasks subject to perceptual learning, the forced-choice discrimination of spatial phase in complex gratings has some interesting properties, because: (i) learning may be achieved rapidly, even within a single experimental session, (ii) there is long-term retention of learned performance (for weeks and even months), (iii) the effects of learning transfer interocularly, but (iv) are selective for the parameters of the stimuli (orientation and spatial frequency) and (v) are restricted to the area stimulated during the learning trials (Fiorentini & Berardi, 1980, 1981). The latter three properties suggest that the learning process occurs cortically, probably at a relatively early level, where single neurons have small receptive fields, selective for stimulus orientation and spatial frequency. Since learning for spatial phase discrimination does not transfer from the trained area to an untrained area of the visual field (as for instance in the case of two adjacent areas of the same hemifield or two areas above and below the fixation point, respectively) one may expect that, for this type of learning, there is no interhemispheric transfer. And indeed, learning of a spatial phase discrimination with stimuli presented in one hemifield at 5 deg from the vertical midline is not retained if the stimuli are removed to the mirror symmetric position in the other hemifield (Fig. 9 A). However, transfer of learning can be found for stimuli in mirror symmetric positions close to the midline (Fig. 9 B) (Berardi and Fiorentini, 1987). That transfer of perceptual learning occurs only for these particular stimulus pairs is in agreement with findings obtained in experiments with sinusoidal gratings. Also in line with those findings, interhemispheric transfer of perceptual learning is conditional upon the parameters of the stimuli: it occurs only for gratings of the same orientation and of low spatial frequency (less than 4 c/deg). Under these conditions transfer also occurs interocularly (Berardi and Fiorentini, 1987). The fact that an interhemispheric transfer of information is limited to regions symmetrical with respect to the vertical midline, within a short distance from the midline and occurs between structures tuned to the same stimulus orientation and receiving input from both eyes, is reminiscent of the properties of callosal receptive fields in the cat visual
72
Berardi and Fiorentini
Left fietd O Right field |
A 100 [-
2 c/deg 5* ecc.
I-1.1.1 or" n." 0
z
50t 20
I 100
1 200
..
_I 300
100
2 c/6eg
1, ecc.
L
5o I 20
I
I
I
100 200 300 TOTAL NUMBER OF TRIALS
Figure 9: Transfer of learning of a spatial phase discrimination task as a function of retinal eccentricity. A: 5 ~eccentricity" B: 1~eccentricity.
cortex (Berlucchi and Rizzolatti, 1968). In man, however, the corpus callosum is not the only structure that assures an anatomical connection between visual areas of the two hemispheres: this role is shared by the anterior commissure. In order to test whether the interhemispheric
Interhemispheric Transfer 73 transfer of perceptual learning could occur in the absence of the corpus callosum, an experiment was run on a subject with callosal agenesis but intact anterior commissure (Fiorentini et al. 1992). It was found that in this subject there was transfer of perceptual learning between regions symmetric with respect to the vertical midline, and that the transfer was present even at 3 deg from the midline, that is, at a larger distance than in normal subjects. Thus the anterior commissure is sufficient to assure transfer of perceptual learning. It may not be necessary, however: it is possible that, in the acallosal subject, the anterior commissure has taken up a task that would normally be subserved by the corpus callosum. Transfer for other tasks thought to be mediated by the corpus callosum was not present in this subject (see section on VEP experiments). III. lnterhemispheric transfer of information on chromatic contrast
Split-chiasm non-human primates can transfer to the untrained eye visual discriminations based on colour (e.g red vs green, blue vs yellow) and learned with the other eye (Butler, 1968; Hamilton, Hillyard & Sperry, 1968). However, for the discrimination of such gross colour differences, the transfer can be mediated by subcortical mechanisms (Black and Myers, 1968). The neocortical commissures are necessary for an interhemispheric comparison of colour (red-green matching-tosample task) in split brain monkeys (Hamilton, Hillyard & Sperry, 1968) and probably for granting the perceptual colour constancy that is normally experienced across translations of an object in the visual field past the vertical meridian. To our knowledge, however, detailed information on this problem is still lacking. A somewhat different question is whether information about the spatial distribution of colour contrast in equiluminant patterns is transferred interhemispherically, and if so, to what degree. Recent experiments (Berardi, Fiorentini & Sartucci, in preparation) indicate that red-green equiluminant gratings differing for the spatial phase of their two harmonic components are better discriminated in the left than in the fight visual field if they are separated 2.5 deg from the vertical midline, while the visual field preference for red-green gratings disappears, at least in some subjects, at 1 deg from the midline. In these subjects the learning of spatial phase discrimination for the red-green patterns (fundamental spatial frequency 1 c/deg) is also transferred across the midline. This seems to replicate the findings described above for
74
Berardi and Fiorentini
black/white gratings differing for their luminance contrast, and suggests that information on pure chromatic red-green contrast can be transferred interhemispherically. Interestingly, for equiluminant blueyellow gratings, differing for the spatial phase of their harmonic components, some subjects show a visual field preference, and a lack of transfer of learning, even at a very low fundamental frequency (0.5 c/deg) and at I deg from the midline, while in the same subjects there is an interhemispheric transfer of learning for red-green patterns. This seems to indicate that, at least in some subjects, information on blueyellow spatial distributions of contrast is not transferred, or is transferred less efficiently, by interhemispheric connections. IV. Discussion
The use of sinusoidal gratings in psychophysical and electrophysiological experiments in humans allow researchers to not only show the presence of an interhemispheric cross-talk but also to study its limits in terms of spatio-temporal information content. Results clearly indicate that interhemispheric transfer is restricted to a range of relatively low spatial and temporal frequencies and that it is present only for stimuli located close to the vertical meridian. Both findings are consistent with the properties of callosal transfer of information at the V1/V2 border found in cat and monkeys. The experiments in animals have further shown that the callosal input is substantially attenuated with respect to the direct geniculo-cortical input, a finding that by itself might explain why the higher spatial and temporal frequencies are almost completely filtered out from the information transmitted interhemispherically. These limits seem to be specific for the callosal connections at the border between the striate and prestriate areas, being more severe on the 17 than 18 side of the border, and virtually absent at the level of area 19 in the cat. As far as data gathered in animals can be extended to explain human findings, one might speculate that the electrophysiological (VEP) and psychophysical interactions demonstrated with sinusoidal gratings symmetrically displaced with respect to the vertical meridian originate at the level of the striate and/or prestriate cortex. It is not surprising therefore that in the cat similar spatial frequency limits were found for those behavioural tasks for which the interhemispheric transfer depends on stimulus contrast (Berardi et al., 1988) and which imply processing at visual areas 17 and 18 (Sprague, De Weerd,
Interhemispheric Transfer 75 V andenbussche & Orban, 1993). Other tasks, like the discrimination of complex forms, do not require the integrity of the primary visual cortex to be performed (se Berlucchi & Sprague, 1981 for review), and it is quite expectable that the interhemispheric transfer of learning for these tasks will not reflect these limits. In discussing these findings in terms of interhemispheric transfer, we have implied a negligible contribution from the bilateral representation of the fovea due to a nasotemporal overlap of the retino-cortical projection. As explained in the Introduction, this assumption is justified both by anatomical data on the monkey and perceptual findings in human subjects. The discrimination of spatial phase in complex gratings shows left visual field preference, while no such visual field asymmetry is found for contrast sensitivity or spatial frequency discrimination with sinusoidal gratings in normal subjects (Fiorentini & Berardi, 1984; Kitterle & Kaye, 1985; Peterzell, Harvey & Hardyck, 1989). This suggests that spatial phase discrimination implies further stages of processing possibly occurring beyond the primary visual cortex. On the other hand, perceptual learning for spatial phase discrimination and its interhemispheric transfer are selective for the retinal location of the stimuli and for their orientation, and this seems to reflect the properties of visual areas where the cells have receptive fields of moderate size, tuned to the orientation of the stimuli. If so, the spatio-temporal limits found in the interhemispheric transfer of visual learning suggest that the properties of callosal connections between these visual areas in man are similar to those observed in the animal at the V l-V2 border. The selectivity for the stimulus parameters of complex grating discrimination contrasts with the size invariance and equivalence across retinal location that characterizes the discrimination of more complex forms. This requires further stages of visual processing involving visual areas beyond the striate and prestriate cortex and their interhemispheric commissures. Indeed, interhemispheric transfer of form discrimination in split-chiasm monkeys is abolished by ablation of inferotemporal cortex (Seacord, Gross & Mishkin, 1979). In view of the size and complex structure of the receptive fields in these areas, the question of their possible spatio-temporal filtering properties seems less appropriate. There is one last point we would like to make. Without plunging into a discussion of the unavoidable temporal and spatial filtering introduced by the sheer presence of additional synaptic connections established by
76
Berardi and Fiorentini
the callosal pathway, we would like to point out another practical reason for filtering out the higher temporal and spatial frequencies in information transferred interhemispherically, namely a reason of economy. In the monkey, input from the retino-geniculate pathway is relayed and subsequently distributed to the different cortical processing streams primarily via area V 1 that therefore receives the bulk of information available for conscious vision. Accepting that the information transfer capacity of the corpus callosum as a neural cable cannot be infinite (both for the intrinsic properties of the cable and for the relatively small number of fibers connecting homologous areas) it is conceivable that some information has to be filtered out. Why filter out the highest spatial and temporal frequencies? The higher the spatial frequency, the smaller the spatial extent of the visual field necessary to ensure its visibility, and vice versa. In other words, for a periodic stimulus (such as a grating or one of the spatial frequency components of a more complex pattern) to be correctly identified, the window through which the stimulus is seen has to be wide enough to encompass a minimum number of periods (5-6 cycles). For gratings near visual acuity (50 c/deg), a field of one tenth of a degree is sufficient. This extent might be granted by the dual foveal representation, which is present within 0.150.2 deg on either side of the fixation point (Fendrich & Gazzaniga, 1989). For lower spatial frequencies, say 1 c/deg, a field of bilateral representation of at least + 2.5 deg from the vertical meridian is needed, and here is where callosal connections come into play.
References
Berardi N. & Fiorentini A. (1987) Interhemispheric transfer of visual information in humans: spatial characteristics. Journal of Physiology, 384, 633-647. Berardi N., Bisti S. & Maffei L. (1987) The transfer of visual information across the corpus callosum: spatial and temporal properties in the cat. Journal of Physiology, 384, 619-632. Berardl, N., Bisti, S., Fiorentini, A. & Maffei, L. (1988a) The transfer of visual information across the corpus callosum in cats, monkeys and humans: spatial and temporal properties. In Hicks, T.P. & Benedek, G. (Eds.), Progress in Brain Research, Vol. 75. Amsterdam: Elsevier Science B.V., pp. 181-185. Berardi, N., Fiorentini, A., Gravina, A. (1988b) Functional dissociation of the hemispheres in the discrimination of complex gratings. Vision Research, 2 8, 491-496.
Interhemispheric Transfer 77 Berardi, N., Bodis-Wollner, I., Fiorentini, A., Giuffre', G. & Morelli, M. (1989) Electrophysiological evidence for interhemispheric transmission of visual information in man. J. of Physiology, 411, 207-225. Berlucchi, G. (1990) Commissurotomy studies in animals. In Boiler, F. & Grafman, J. (Eds.), Handbook of Neuropsychology, Vol. 4. Amsterdam. Elsevier Science B.V, pp. %47. Berlucchi, G. & Antonini, A. (1989) Ttie role of the corpus callosum in the representation of the visual field in cortical areas. In Trevarthen, C.B. (Ed.), Brain Circuits and Functions of_ the Mind. Cambridge, U.K.: Cambridge University Press, pp. 129-139. Berlucchi, G., Antonini, A., Mascetti, G.G. & Tassinari, G. (1987) Role of callosal connections in the representation of the visual field in the primary visual cortex of the cat. In Ottoson, D. (Ed.), Duality and Unity of the Brain. London: MacMillan, pp. 349-366. Berlucchi, G. & Rizzolatti, G. (1968) Binocularly driven neurons in the visual cortex of split chiasm cats. Science, 159, 308-310. Berlucchi, G. & Sprague, J.M. (1981) The cerebral cortex in visual learning and memory, and in interhemispheric transfer in the cat. In Schmitt F.O., Worden F.G., Adelman G. & Deniss J.G. (Eds.), The Organization of the Cerebral Cortex. Cambridge, MA: The MIT Press, pp. 415-440. Berlucchi, G., Sprague, J.M., Lepore, F. & Mascetti, G.G. (1978) Effects of lesions of areas 17, 18 and 19 on interocular transfer of pattern discriminations in split-chiasm cats. Experimental Brain Research, 31, 275-297. Berlucchi, G., Tassinari, G. & Antonini, A. (1986) The organization of the callosal connections according to Sperry's princil~le of supplemental complementarity. In Lepore, F., Ptlto, M & . Jasper, H.H. (Eds.), Two Hemispheres- One Brain: Functions of the Corpus Callosum. Alan R. Liss. New York, pp. 171-188. Black, P. & My.ers, R.E. (1964) Visual functions of the forebrain commlssures m the chimpanzee. Science, 146, 799-780 Bunt, A.H., Minckler, D.S. & Johanson, G.W. (1977) Demonstration of bilateral projection of the central retina of the monkey with horseradish peroxidase neuronography. Journal of Comparative Neurology, 171, 619-630. Butler, C.R. (1968) A memory record for visual discrimination habits produced in both cerebral hemispheres of monkeys when only one hemisphere has received direct visual information. Brain Research, 10, 152-167. Davidoff, J. (1982) Studies with non-verbal stimuli. In Beaumont, J.G. (ed.), Divided Visual Field Studies of Cerebral Organization. London: Academic Press, pp. 29-55. Eacott, M.J. & Gaffan, D. (1989) Interhemispheric transfer of visual learning in monkeys with intact optic chiasm. Experimental Brain Research, 74, 348-352. Fendrich, R. & Gazzaniga, M.S. (1989) Evidence of foveal splitting in a commissurotomy patient. Neuropsychologia, 27, 273-281 Fiorentini, A. & Berardi, N. (1980) Perceptual learning specific for orientation and spatial frequency. Nature, 287, 43-44.
78
Berardi and Fiorentini
Fiorentini, A. & Berardi, N. (1981) Learning in grating waveform discrimination: specificity for orientation and spatial frequency. Vision Research, 21, 1149-1158. Fiorentini, A. & Berardi, N. (1984) Right-hemisphere superiority in the discrimination of spatial phase. Perception, 13, 695-708. Fiorentini, A., Berardi, N., Falsini, B. & Porciatti, V. (1992) Interhemispheric transfer of visual perceptual learning in callosal agenesis. Clinical Vision Sciences, 7, 133-141. Fiorentini A., Pirchio M. & Spinelli D. (1983) Electrophysiological evidence for spatial frequency selective mechanisms in adults and infants. Vision Research, 2 3, ll 9-127. Fujita, I., Tanaka, K., Ito, M. & Chen~, K. (1992) Columns for visual features of objects in monkey inf~rotemporal cortex. Nature, 360, 343 -346. Fukuda, Y., Sawai, H., Watanabe, M. Wakakuwa, K. & Morigiwa, K. (1989) Nasotemporal overlap of crossed and uncrossed retinal ganglion cell projections in the Japanese monkey (Macaca Fuscata). Journal of Neuroscience, 9, 2353-2373. Gazzaniga, M.S., Bogen, J.E. & Sperry, R.W. (1965) Observations on visual perception after disconnection of the cerebral hemispheres in man. Brain, 88, 221-236. Gross, C.G., Bender,D.B. & Mishkin, M. (1977) Contributions of the corpus callosum and the anterior commissure to visual activation of inferior temporal neurons. Brain Research, 131,227-239. Gross, C.G. & Mishkin, M. (1977) The neural basis of stimulus equivalence across retinal translations. In Harnad S., Doty R.W., Goldstein L., Krauthamer G. (eds.), Lateralization of the Nervous System. New York: Academic Press, pp. 109-122. Hamilton, C.R., Hillyard, S.A. & Sperry, R.W. (1968) Interhemispheric comparison of colour in split-brain monkeys. Experimental Neurology, 21, 486-494. Harvey, L.O. (1978)Single representation of the visual midline in humans. Neuropsychologia, 16, 601-610. Haun, F. (1978) Functional dissociation of the hemispheres using foveal mput. Neuro~sychologia, 16, 725-733 Ipata, A., Girelli, M., Miniussi, C. & Marzi, C.A. (1997) Interhemispheric transfer of visual information in humans: The role of different callosal channels. Archives Italiennes de Biologie, in press. Kennedy, H., Dehay, C. & Bullier, J. (1986). Organization of the callosal connections of visual areas V1 and V2 in the macaque monkey. Journal of Comparative Neurology, 247, 398-415. Kitterle, F.L., Chnstman, S. & Conesa, J.S. (1995). Spatial-frequency selectivity in hemispheric transfer. In Kitterle, F.L. (ed.), Hemispheric Communication: Mechanisms and Models. Hillsdale, N.J.: L. Erlbaum Assoc., pp. 319-346. Kitterle. F.L. & Kaye, R.S. (1985) Hemispheric symmetry in contrast and orientation sensitivity. Perception & Psychophysics, 3 7, 391-396. Koerner, F. & Teuber, H.L. (1973) Visual field defects after missile injuries to the geniculo-striate pathway in man. Experimental Brain Research, 18, 88-113.
Interhemispheric Transfer 79 Leventhal. A.G., Ault, S.J. & Vitek, D.J. (1988) The nasotemporal division in primate retina: the neural bases of macular sparing and splitting. Science, 240, 66-67. Lines, C.R. (1984) Nasotemporal overlap investigated in a case of agenesis of the corpus callosum. Neuropsychologia, 22, 85-90. Maier J., Dagnelie G., Spekreijse H. & van Dijk B.W. (1987) Principal component analysis for source localization of VEPs in man. Vision Research, 2 7, 165 - 177. Payne, B.R. (1990)Representation of the ipsilateral visual field in the transition zone between areas 17 and 18 of the cat's cerebral cortex. Visual Neuroscience, 4, 445-474. Peterzell, D.H., Harvey, L.O. & Hardyck, C.D. (1989) Spatial frequencies and the cerebral hemispheres: Contrast sensitivity, visible persistence and letter classification. Perception & Psychophysics, 46, 443 -455. Regan D. (1983) Spatial frequency mechanisms in human vision investigated by evoked potential recording. Vis. Research, 23, 1401-1407. Seacord, L., Gross, C.G. & Mishkin, M. (1979) Role of inferior temporal cortex in interhemispheric transfer. Brain Research, 167, 259-272. Sperry, R.W. (1961) Cerebral organization and behaviour. Science, 133, 1747-1757. Sperry, R.W. (1962) Orderly function with disordered structure. In Foerster, H.V., Zopt, G.W. (eds.), Principles of self-organization. New York: Pergamon Press, pp. 279-290. Sprague, J.MS., De Weerd, P., Vandenbussche, E. & Orban, G.A. (1993) Orientation discrimination in the cat and its cortical loci. Progress in Brain Research, 95, 381-400. Stone, J., Leicester, J. & Sherman, S.M. (1973) The naso-temporal division of the monkey's retina. Journal of Comparative Neurology, 150, 333-348. Sugishita, M., Hamilton, C.R., Sakuma, I. & Hemmi, I. (1994) Hemispheric representation of the central retina of commissurotomized subjects. Neuropsychologia, 32, 1994. Tanaka, K. (1993) Neuronal mechan-isms of object recognition. Science, 262, 685-688. Tootell, R.B.H., Switkes, E., Silverman, M.S. & Hamilton, S.L. (1988) Functional anatomy of macaque striate cortex. II. Retinotopic organization. Journal of Neuroscience, 8, 1531-1568. Van Essen, D.C. & Zeki, S.M. (1978) The topographic organization of rhesus monkey striate cortex. Journal of Physt~ology, 277, 193-226. Wang, G., Tanaka, K. & Tanifuji, M. (1996)Optical imaging of functional organization in the monkey inferotemporal cortex. Science, 272, 1665-1668.
This Page Intentionally Left Blank
SECTION 11: O B J E C T AND SPATIAL REPRESENTATIONS
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
83
Chapter 4
Hemispheric Asymmetry for Components of Spatial Processing Joseph B. Hellige University of Southern California The two cerebral hemispheres of the human brain do not have identical ability to localize stimuli in space. A predominant view has been that the right hemisphere is superior to the left for identifying spatial relations among objects. During the last decade, however, it has become apparent that such right-hemisphere superiority does not extend to the identification of all aspects of spatial relations. Instead, it has been hypothesized that each hemisphere is dominant for identifying different and, to some extent, complementary aspects of spatial relations. In the present chapter, I outline a distinction that has been made between "categorical" and "coordinate" spatial relations and review investigations of hemispheric asymmetry for identifying these different types of spatial relations, with a view toward identifying mechanisms that might underlie such asymmetry. Accordingly, the chapter begins with a brief review of the categorical/coordinate distinction and relevant studies of hemispheric asymmetry. This is followed by a review of recent experiments that have tested various hypotheses about the underlying mechanisms. I then turn to experiments that have attempted to extend the categorical/coordinate distinction beyond the domain of spatial relationships. The chapter ends with a reconsideration of underlying mechanisms and a discussion of directions for future studies of hemispheric asymmetry for spatial processing.
84 Hellige
The Categorical/Coordinate Distinction During the last 10 years or so, a considerable amount of theoretical and empirical work has beenconducted within a framework proposed by Kosslyn and colleagues in which the brain is hypothesized to compute two kinds of spatial-relation representations (e.g., Kosslyn, 1987). One type of representation ("categorical") is used to assign a spatial relation between two stimuli to a category such as "touching" versus "not touching" and "above" versus "below." The other type of representation ("coordinate") preserves location information using something like a metric coordinate system in which distances are specified effectively. Recent experiments suggest that the right hemisphere makes more effective use of this coordinate or metric distance information about spatial relationships whereas there is either no hemispheric asymmetry or there is left-hemisphere superiority for processing information about categorical spatial relationships (for reviews, see Hellige, 1993, 1995, 1996; Kosslyn, Koenig, Barrett, Cave, Tang & Gabrieli, 1989; Kosslyn & Koenig, 1992). Results consistent with the general conclusions just outlined have been reported for a variety of visual half-field experiments from a number of different laboratories. In such experiments, visual stimuli on each trial are presented briefly to either the left visual field and right hemisphere (LVF/RH trials) or to the right visual field and left hemisphere (RVF/LH trials) and performance is examined as a function of which hemisphere receives the stimulus information directly. For example, in an experiment with neurologically normal right-handed observers, Chikashi Michimata and I (Hellige & Michimata, 1989) had observers indicate whether a dot was above or below a line (a categorical Above/Below task) or, on a different block of trials, indicate whether the dot was within 2 cm of the line (a coordinate, Near/Far task). Figure 1 illustrates the possible positions of the dot relative to the line. The results are shown in Figure 2a. Consistent with the hypothesis outlined above, for the Above/Below task a RVF/LH advantage approached statistical significance and for the Near/Far task there was a highly significant LVF/RH advantage--producing a highly significant Task by Hemisphere interaction. We (Hellige, Bloch, Cowin, Eng, Eviatar & V. Sergent, 1994) have replicated this effect for a new group of right-handed observers (see Figure 2b), and also found the Task by Hemisphere interaction to be absent in left-handed observers. For additional examples of
Spatial Processing
85
0 0 0 0 0 0 0 0 0 0 0 0 Figure 1. Illustration of the line and dot positions used by Hellige and Michimata (1989). On each triM, observers saw the line and a single one of the 12 dots.
this type of Task by Hemisphere interaction for fight-handed observers, see Kosslyn (1987), Kosslyn et al. (1989), J. Sergent (1991; but only for a low level of luminance); Koenig et a1.(1992), Rybash and Hoyer (1992); Cowin and Hellige (1994, 1995), and Crebolder and Bryden (1996); see also Laeng, Peters and McCabe (1996). In addition, in some studies, this Task by Hemisphere interaction disappears with practice, possibly because observers learn to perform the distance judgment in a more categorical way as they become familiar with the stimuli (e.g., Kosslyn et al., 1989, Cowin & Hellige, 1994). Note from Figures 2a and 2b that, for the tasks developed by Hellige and Michimata (1989), the categorical task was easier than the coordinate task. If this were always the case, an alternative interpretation of the Task by Hemisphere interaction could be formulated in terms of task difficulty. However, in several experiments reported by Kosslyn and colleagues (e.g., Kosslyn, 1987; Kosslyn et al., 1989), the specific categorization tasks used (e.g., is a dot touching an irregular blob or not--an "on/off" task) were more difficult than the specific coordinate distance judgment tasks (e.g., is the dot within 2 mm of the blob).
86
Hellige
760 740
Hellige & Michimata
(1989)
720 700 i 680 ~ RT 660 640 620 LVF/RH RVF/LH
600 580 560
I
Above/Below
I
Task
Near/Far
Figure 2a. Reaction time for a categorical spatial task (Above/Below) and for a coordinate spatial task (Near/Far) for LVF/RH and RVF/LH trials. Results come from Hellige and Michimata (1989).
Nevertheless, there was a RVF/LH advantage for the on/off task and a LVF/RH advantage for the distance judgment task. Thus, the Task by Hemisphere interaction seems to be related to the categorical versus coordinate nature of the task and not simply to the difficulty of the spatial discrimination, at least as difficulty is measured by overall performance. The categorical/coordinate distinction has also been extended to tasks that require the generation and processing of visual images. For example, Kosslyn and his colleagues have suggested that the left
Spatial Processing 87 hemisphere is dominant for generating and processing visual images that require the correct categorical arrangement of parts or that are 760
740
,,
Hellige et al. ( 1 9 9 4 )
720
700 680 RT 660 640 620 --
LVF/RH F/LH
600
580 560
I
Above/Below
I
Task
Near/Far
Figure 2b. Reaction time for a categorical spatial task (Above/Below) and for a coordinate spatial task (Near/Far) for LVF/RH and RVF/LH trials. Results come from Hellige et al. (1994).
generated in a categorical, part-by-part manner; whereas the right hemisphere is dominant for generating and processing visual images in terms of precise metric coordinates. Thus, Kosslyn, Holtzman, Farah and Gazzaniga (1985) report that the right hemisphere of split-brain patient J. W. could not perform tasks that required him to make categorical decisions about parts of imaged objects (e.g., Do a hog's ears protrude above the top of the skull?), though the left hemisphere's performance was nearly perfect. In addition, J. W.'s right hemisphere could perform perfectly tasks that required a decision about the overall shape or size of
88
Hellige
an imaged object (e.g., Is a book higher than it is wide?). In a series of visual half-field experiments with neurologically intact individuals, Kosslyn (1988) has also reported a RVF/LH advantage for a task that is believed to use categorical relations to arrange letter segments within a grid and a LVF/RH advantage for a task that requires the arrangement of letter segments without benefit of a grid and believed to use coordinate relations to arrange the segments (see also, Kosslyn, Maljkovic, Hamilton, Horwitz, & Thompson, 1995). Michimata (1997) required right-handed observers to make categorical and coordinate decisions about the long and short hands of a clock face. For the categorical task, observers indicated whether both of the hands were above or below the horizontal midline of the clock face. For the coordinate task, observers indicated whether the angle formed by the two hands was greater or smaller than 60 degrees. In a perceptual condition, analog clock faces were presented to either the LVF/RH or RVF/LH on each trial. In an imagery condition, time was shown on a digital clock face (e.g., 3:25) in either the LVF/RH or RVF/LH, so that observers would have to generate images of analog clock faces in order to perform the tasks. In both the perceptual and the imagery conditions, there was a significant LVF/RH advantage for the coordinate task and a nonsignificant trend in the opposite direction for the categorical task. Thus, the Task by Hemisphere interaction does not require external stimulation of visual pathways. The similarity of perceptual and imagery results is consistent with hypothesized isomorphism between the processing of visual stimuli and the processing of visual images and indicates that the categorical/coordinate distinction is relevant for both. As this brief review suggests, the Spatial Task by Hemisphere interaction is sufficiently robust to merit further empirical and theoretical investigation. It is particularly important to consider what sort of computational and neural mechanisms might underlie these aspects of hemispheric asymmetry. With this in mind, I turn to the search for such mechanisms.
The Search for Underlying Mechanisms of Hemispheric Asymmetry for Spatial Processing In this section, I review the evolution of hypotheses regarding the computational and neural mechanisms that underlie hemispheric asymmetry for processing categorical and coordinate spatial relation-
Spatial Processing 89 ships. This section begins with a discussion of the speech/attention-shift hypothesis introduced by Kosslyn (1987). One issue that arises in considering possible mechanisms is the extent to which the two types of spatial relationships are processed independently of each other or in neural subsystems that do not interact with each other. Accordingly, in the second part of this section I consider this issue of independence. Hemispheric asymmetry for many aspects of visual information processing has been shown to be influenced by the nature of visual input and by which aspects of the visual input are most relevant for efficient performance of a task. In the third part of this section, I consider theoretical and empirical work that illustrates how the nature of task-relevant visual information is also important for spatial processing. The
Speech/Attention-Shift Hypothesis
Kosslyn's original hypothesis about hemispheric asymmetry for processing spatial relations was based on the assumption that the left hemisphere is specialized for the control of speech and the right hemisphere is specialized for the control of rapid shifts of attention across space (Kosslyn, 1987). He proposed that these initial specializations provided a "seed" function for each hemisphere, which would operate in the following way. As new skills are added during the course of phylogenetic or ontogenetic development, those skills become lateralized to one hemisphere or the other to the extent that they can be performed better by the neurological substrata laid down in one hemisphere compared to the other (see Hellige, 1993, for additional discussion of this type of developmental "seeding" idea). For a variety of reasons, Kosslyn argued that the neurological substrata for speech control and for rapid shifts of attention would be well-adapted for categorical and coordinate spatial processing, respectively. While not e x p l i c i t l y abandoning this speech/attention-shift hypothesis, more recently Kosslyn and colleagues have emphasized differences in the nature of the visual information that is most useful for computing categorical versus coordinate spatial information (e.g., Kosslyn. Chabris, Marsolek & Koenig, 1992). I will discuss this newer hypothesis in some detail later. Before doing so, however, it is useful to consider certain experimental results that point to limitations of the speech/attention-shift hypothesis and that must be considered in the evaluation of alternative hypotheses.
90
Hellige
Based on the idea of individual differences in seeding of the two hemispheres, Kosslyn (1987) suggested that individuals who show a relatively large LVF/RH advantage for coordinate spatial processing should also show a relatively large RVF/LH advantage for categorical spatial processing. This assertion received some tentative support from a study in which larger laterality effects of both sorts were obtained for strongly right-handed individuals than for ambidextrous individuals, with the assumption being that different seeding of the two hemispheres is more likely in the strongly right-handed group (Kosslyn, 1987; Kosslyn et al., 1989). However, other results suggest that, in fact, the two types of asymmetry are not correlated with each other. For example, Hellige and Michimata (1989) and Hellige et al. (1994) had the same individuals perform both a categorical task (is a dot above or below a line) and a coordinate task (is a dot within 2 cm of a line), so that they could examine the relationship between the perceptual asymmetries for the two tasks. In addition, Hellige et al. (1994) included left-handed as well as fight-handed individuals. To the extent that individuals who show a relatively large LVF/RH advantage for the coordinate task also show a relatively large RVF/LH advantage for the categorical task, the L V F RVF difference scores for the two tasks should correlate negatively. This was clearly not the case, as none of the correlations even approached statistical significance. In fact, Hellige and Michimata reported a nonsignificant positive correlation of .16 (the dependent variable was reaction time of correct responses). For right- and left-handed individuals combined, Hellige et al. (1994) reported a correlation o f - . 0 1 (with the correlations for fight- and left-handed groups being .19 and -. 15, respectively). Thus, laterality for the two types of spatial relations tasks seem to be uncorrelated (see also Laeng & Peters, 1995). While this does not necessarily invalidate the concept of seeding of the two hemispheres in different ways, it does suggest that the seeds that create a bias toward efficient categorical versus coordinate processing are sown independently of each other. Multi-task investigations of individual differences in hemispheric asymmetry have also provided information about the extent to which either of these two aspects of laterality for spatial processing arise from a more primitive left-hemisphere superiority for speech. If, for example, a common seed underlies hemispheric asymmetry for speech processing and for processing categorical spatial relationships, then we might expect an appropriate correlation between hemispheric asymmetry for
Spatial Processing 91 speech perception and hemispheric asymmetry for categorical spatial processing. In a multi-task study of individual variation in hemispheric asymmetry, Hellige et al. (1994) included the categorical and coordinate line and dot tasks used by Hellige and Michimata (1989), as well as a dichotic listening task requiring the identification of stop-consonants and a visual half-field task that required the identification of nonword trigrams. For present purposes, the interesting finding was that hemispheric asymmetry for both categorical and coordinate spatial processing was unrelated to ear asymmetry for the verbal dichotic listening task or to the visual field asymmetry for identifying nonword trigrams (though the latter two types of asymmetry were significantly correlated). Studies by Boles (1991, 1992) also suggest that laterality for spatial processing is independent of laterality for verbal processing. Thus, hemispheric asymmetry for processing spatial relationships does not seem to be based on a more primitive asymmetry for speech processing.
Are Categorical and Coordinate Spatial Relationships Processed Independently? In considering mechanisms that might contribute to the processing of categorical and coordinate spatial relationships, it is useful to consider the extent to which the two types of spatial relationships are processed independently of each other. To be sure, the existence of Task by Hemisphere interactions suggests that somewhat different neural mechanisms underlie the two types of spatial processing or that different aspects of visual information are important for different tasks. However, this does not necessarily mean that these mechanisms or different aspects of visual information are completely independent of each other in the sense that they do not interact. For example, Justine Sergent (1991) reported that absolute distance between two stimuli sometimes influences performance in categorical tasks that require individuals to make decisions about their relative location. Issues of independence have also been addressed by Niebauer (1996). In one experiment, he presented line-and-dot stimuli similar to those described earlier to the center of the visual field (CVF) on each trial. With the CVF presentation, Niebauer found that the time to perform a coordinate task (near/far judgment) was influenced by a prime stimulus that informed the observer whether the upcoming dot would be above or below the line (a categorical prime). That is, a valid
92
Hellige
prime decreased reaction time relative to an invalid prime. Interestingly, a coordinate prime (indicating whether the upcoming dot was near the line or far from the line) had no effect on making a categorical decision (Is the dot above or below the line?). Based on these priming effects, Niebauer suggests that computation of categorical information precedes, and serves as input to, the computation of coordinate spatial information--at least on CVF trials on which both hemispheres receive the stimulus information. This is an interesting idea that merits additional investigation. As has been the case in many previous experiments, reaction times in Niebauer's studies were faster for the categorical task than for the coordinate task. Among other things, it is important to determine whether this same asymmetry in priming would be found when the difficulty of categorical and coordinate tasks is reversed. The priming effects found by Niebauer (1996) on CVF trials disappeared when the line-and-dot stimuli were presented to the LVF/RH or RVF/LH on each trial. This may indicate that the priming effects on CVF trials reflect interactions between the two, equivalently stimulated, hemispheres. It is instructive, however, to consider alternative possibilities. For example, suppose that priming is obtained to the extent that the prime allows the observer to focus attention on a single critical "boundary" when a prime is presented (especially when compared with the no-prime condition). For the Above/Below task, the boundary is defined by the line that is actually presented on the screen. For the Near/Far task, there are two "invisible" boundaries that separate near from far stimuli, one above the line and one below the line. If either the prime does not permit attention to be restricted to a single, critical boundary or such a restriction is already possible in the no-prime condition, then no priming is obtained. For the Above/Below task on CVF trials, attention can be directed to the boundary line regardless of whether there is a prime. Thus, a Near/Far cue produces no additional information about the boundary and no priming is predicted. For the Near/Far task, in the no-prime condition the observer must split attention between two near/far boundaries, one above the line and one below the line. In this case, an Above/Below cue allows attention to be restricted to a single one of these boundaries, and priming results. By way of contrast, when the stimuli are directed randomly to the LVF/RH or RVF/LH on each trial, attention can never be restricted to a single critical boundary because there is always uncertainty about which visual field will be stimulated. Hence, all priming disappears. Although this
Spatial Processing
93
explanation is admittedly speculative, it could be tested in an experiment that randomly intermixed CVF and lateralized presentations, with the prediction being that priming would disappear on CVF trials. Even though additional studies are needed to choose among specific interpretations of the priming effects, the existence of such effects is generally consistent with hypothesized interaction between the two types of spatial processing. However, Kosslyn et al. (1992) found that a neural-network model that codes only categorical information (Is a dot above or below a line?) is sensitive to the distance between the dot and the line. Specifically, the further the dot from the line, the better the performance of the model. This demonstrates that a network devoted to the processing of categorical spatial relationships need not be completely immune to the effects of distance (a type of coordinate spatial information). Thus, the presence of distance effects in a categorical task (e.g., J. Sergent, 1991) does not necessarily mean that a network devoted to categorical processing interacts with a network devoted to coordinate processing. In view of Niebauer's (1996) priming results, it would be interesting to use similar neural-network models to simulate the various types of priming (categorical-on-coordinate and coordinate-on-categorical). For example, suppose that a network devoted to coordinate processing performs more efficiently with a valid categorical prime than with an invalid categorical prime. This would suggest that the existence of priming in and of itself does not necessarily indicate any interaction between computational mechanisms devoted to the two types of spatial processing. The Nature of Task-Relevant Visual Information
Hemispheric asymmetry for the identification of visual stimuli is sensitive to the precise nature of the visual input and to the aspects of visual information that are relevant for efficient performance of the task that the observer is required to perform. It is useful to consider how the nature of task-relevant visual information might also influence hemispheric asymmetry for determining the location of visual stimuli in space. Detailed review of the stimulus identification studies is beyond the scope of the present chapter. However, a brief review of the relevant stimulus identification findings provides an important context for considering potential mechanisms that underlie hemispheric asymmetry for stimulus localization.
94 Hellige
J J J JJJJJJJJ J J J
J J J J J J
Figure 3. Example of an hierarchical visual pattern composed of small letters (the local level) arranged into the shape of a large letter (the global level).
Visual stimuli can contain many levels of embedded structure, with smaller local patterns or parts contained within larger global patterns (see Figure 3). With respect to stimulus identification, there is a great deal of evidence that the fight hemisphere is superior for the processing of global levels of visual information (e.g., the large H in Figure 3) whereas the left hemisphere is superior for the processing of local levels of visual information (e.g., the small J's in Figure 3) [e.g., Delis, Robertson & Efron, 1986; Ivry & Robertson, 1997; Lamb, Robertson & Knight, 1989, 1990; Martin, 1979; Robertson, 1995; Robertson, Lamb & Knight, 1988, 1991; J. Sergent, 1982; van Kleeck, 1989]. There is a clear relationship between global and local aspects of a visual stimulus and what is referred to as low versus high spatial frequency. A single spatial frequency consists of a regular sinusoidal variation of luminance across space and looks somewhat like alternating dark and light bars with fuzzy borders. Spatial frequency refers to the number of dark-light cycles per unit of space--the more cycles per unit of space, the higher the spatial frequency. The concept of spatial frequency has generated considerable interest because it is possible, in principle, to represent any complex image as a set of spatial frequencies in specific orientations, phase relationships and so forth. Under normal viewing conditions, information about the larger, global aspects of a visual stimulus is carried by a lower range of spatial frequencies than is
Spatial Processing 95 information about the smaller, local aspects of the same stimulus. Thus, it may not come as a surprise that, at some level of processing beyond the sensory cortex, the fight and left hemispheres are biased toward efficient use of lower and higher spatial frequencies, respectively. In fact, a variety of evidence suggests that at least three aspects of spatial frequency influence hemispheric asymmetry for stimulus identification: (1) the absolute range of spatial frequencies contained in the stimulus; (2) the range of spatial frequencies that is relevant for the task being performed or that is attended to by the observer; and (3) whether the attended frequencies are high or low relative to other frequencies contained in the stimulus (for empirical examples and reviews, see Christman, 1989, 1990, Christman, Kitterle & Hellige, 1991, Hellige, 1993, 1995, 1996; Ivry & Robertson, 1997; Kitterle & Christman, 1991; Kitterle, Christman & Conesa, 1993; Kitterle, Christman & Hellige, 1990; Kitterle & Selig, 1991; J. Sergent, 1983, 1987; J. Sergent & Hellige, 1986). In view of the fact that the nature of presented and attended visual information exerts powerful influences on hemispheric asymmetry for stimulus identification, it is important to consider how the nature of visual information might contribute to hemispheric asymmetry for localizing stimuli in space. As noted earlier, Kosslyn et al. (1992; see also Koenig & Kosslyn, 1992) have provided a conceptualization of the mechanisms that underlie hemispheric asymmetry for spatial processing that emphasizes the nature of the visual information that is most useful for computing categorical versus coordinate information. In a set of neural-network simulations, Kosslyn et al. (1992) found that networks receiving input that had been filtered through units with relatively large, overlapping "receptive fields" computed coordinate spatial information better than networks that received input that had been filtered through units with relatively small, nonoverlapping "receptive fields." Exactly the reverse was found for the computation of categorical spatial information (for critique and additional discussion of these neural network simulations, see Cook, Fruh & Landis, 1995 and Kosslyn, Chabris, Marsolek, Jacobs & Koenig, 1995). To account for hemispheric differences in categorical versus coordinate processing, Kosslyn and colleagues hypothesize that the left hemisphere is predisposed toward efficient use of information from visual channels with small, nonoverlapping receptive fields whereas the right hemisphere is predisposed toward efficient use of information from visual channels
96 Hellige with large, overlapping receptive fields. In support, Kosslyn et al. (1992) suggest that magnocellular ganglia (which tend to have relatively large receptive fields) may project preferentially to the right hemisphere. In addition, they note the evidence mentioned earlier that, at some level of processing beyond the sensory cortex, the left and fight hemispheres are dominant for processing visual information carried by channels tuned to relatively high and low spatial frequencies, respectively. Although these neural network simulations have been criticized (Cook et al., 1995), and alternative interpretations of the simulations are possible, the hypotheses that were generated by the simulations can be subjected to empirical test. There are two types of predictions about the effects of experimental manipulations that serve to accentuate or attenuate processing in visual streams that have characteristics similar to those attributed to the input filtering units of the neural network models. One type of prediction has to do with the effect of experimental manipulations on categorical versus coordinate processing tasks--and this type of prediction does not directly involve hemispheric asymmetry. For example, an experimental manipulation that attenuated processing along the magnocellular visual pathway would be expected to disrupt coordinate spatial processing more than categorical spatial processing. The other type of prediction has to do with the Task by Hemisphere interaction. For example, to the extent that the two hemispheres utilize different aspects of the visual information to perform efficiently, the attenuation of one type of information should have a greater detrimental effect on the hemisphere that is more dependent on that type of information (for examples of this logic in the domain of visual identification, see Hellige, 1993; Jonsson & Hellige, 1986). With these types of predictions in mind, it is instructive to consider recent examinations of the effects of perceptual manipulations on categorical and coordinate spatial processing. It has been hypothesized that the processing of visual information in primates is accomplished by two parallel visual pathways with different spatial and temporal characteristics (for discussion, see Breitmeyer & Williams, 1990; Breitmeyer, May & Heller, 1991; Livingstone & Hubel, 1984, 1987, 1988; Schiller & Malpeli, 1978; Shapley, 1994; Van Essen, 1985). In general, the magnocellular system is most sensitive to low spatial frequencies, has high temporal resolution and responds quickly and transiently to moving targets. This system is thought to be involved in such things as brightness discrimination, the perception of motion
Spatial Processing 97 and depth, the localization of visual stimuli in coordinate space and in the global analysis of visual scenes. By way of contrast, the parvocellular system is most sensitive to high spatial frequencies, has a long response persistence and responds in a sustained fashion to stationary targets. This system is thought to be involved in such things as the identification of visual patterns, especially small local details, and in color perception. Given the characteristics attributed to these two visual pathways, one possible interpretation of the neural network results described earlier is that categorical and coordinate spatial processing depend relatively more on information carried by the parvocellular and magnocellular pathways, respectively. Elizabeth Cowin Roth and I (Cowin and Hellige, 1994) examined the effects of dioptric blurting on categorical (above/below) and coordinate (near/far) spatial processing tasks using line and dot stimuli similar to those described earlier. Dioptric blurting selectively impairs processing of relatively high visual spatial frequencies and, according to the hypotheses outlined, such blurting should be particularly disruptive of categorical spatial processing. In fact, we found that dioptric blurring consistently increased reaction time and error rate for a categorical task that required observers to indicate whether a dot was above or below a line. However, the amount of dioptric blurring that we used had no consistent effect on either reaction time or error rate for a coordinate task that required observers to indicate whether the dot was within 3 mm of the line. On an initial block of trials, we found significantly fewer errors on LVF/RH than on RVF/LH trials for the coordinate processing task and this LVF/RH advantage was independent of whether the stimuli were clear or blurred. While we did not design this experiment with parvocellular and magnocellular pathways in mind, a dioptric blurring manipulation might be expected to differentially attenuate processing along the parvocellular pathway. Thus, our results suggest that processing along this pathway is more critical for categorical than for coordinate spatial processing. More recently, Roth and I (Cowin & Hellige, 1995; Roth & Hellige, 1997) have attempted to examine the effects of attenuating processing along the magnocellular visual pathway. We did so in the following way. Breitmeyer and his colleagues (e.g., Breitmeyer & Williams, 1990; Breitmeyer et al., 1991; Williams, Breitmeyer, Lovegrove & Guitierrez, 1991) have reported that both metacontrast masking and the perception of stroboscopic motion are considerably weaker when stimuli are
98
Hellige
presented on a red background than on an isoluminant green background. Neurophysiologically based models of metacontrast and stroboscopic motion indicate that both phenomena result from the interaction of two visual subsystems: the sustained subsystem and the transient subsystem. As Breitmeyer and colleagues note, the parvocellular and magnocellular pathways in monkey may be neural analogs of the more functionally defined sustained and transient channels, respectively. Within the context of a sustained-transient channel approach, metacontrast masking is produced when the faster responding transient channels activated by the subsequent mask inhibit the slower responding sustained channels activated by the target. The perception of stroboscopic motion is thought be the result of response integration within the transient visual system. Thus, the fact that a red background reduces both metacontrast masking and the perception of stroboscopic motion indicates that the activity of transient visual channels is attenuated by red relative to green backgrounds. From this perspective, it is interesting that a subpopulation of magnocellular neurons have receptive fields that are characterized by a red-dominant surround mechanism (e.g., Wiesel & Hubel, 1966; DeMonasterio, 1978; DeMonasterio & Schein, 1980; Livingstone & Hubel, 1984; Marrocco, McClurkin & Young, 1988). As noted by Breitmeyer and Williams (1990), this may be the reason why diffuse red light has been found to provide tonic suppression of activity in certain neurons contained in the magnocellular pathway (Dreher, Fukuda & Rodieck, 1976; Livingstone & Hubel, 1984; Van Essen, 1985). In view of the results reported by Breitmeyer and colleagues, we reasoned that the use of green stimuli on an isoluminant red background would attenuate processing along the magnocellular pathway relative to the parvocellular pathway. To the extent that processing along the magnocellular pathway is more important for coordinate spatial processing than for categorical spatial processing, coordinate processing should be more disrupted by the use of green-on-red compared to red-on-green stimulus conditions. In our experiments with colored stimuli, the stimuli on each trial consisted of a horizontal line and two dots, with the dots being on the same horizontal level as each other (see Figure 4). The line varied in length from trial to trial as did the horizontal distance between the two dots. The categorical task required observers to indicate whether the dots were above or below the line whereas the coordinate task required observers to indicate whether or not the line on that trial could fit
Spatial Processing 99 In these stimuli, the dots are above and the line fits between the gap: m
m
mm
m
m
m
m
~
mm mm
In these stimuli, the dots are below and the line fits between the gap: m
m
In these stimuli, the dots are above and the line does not fit between the gap:
m
m
m
m
m
m
In these stimuli, the dots are below and the line does not fit between the gap: m
m
mm ~
mm
~
mm
mm
m
Figure 4. Illustration of line and dot patterns used by Cowin and Hellige (1995). between the two dots. These stimuli and tasks were patterned after those used by Rybash and Hoyer (1992) and shown in previous experiments to produce a robust Task X Hemisphere interaction. For some observers, green lines and dots were presented on an isoluminant red background. For other observers, red lines and dots were presented on an isoluminant green background. The stimuli on each trial were presented briefly to either the LVF/RH or RVF/LH. Figure 5 shows reaction time for categorical and coordinate tasks as a function of background color. As shown in the figure, there was a robust Task by Color interaction. For the coordinate task, reaction time was significantly longer in the green-on-red condition than in the redon-green condition. For the categorical task, there was a nonsignificant trend in the opposite direction. Note that this interaction is consistent with the hypothesis that the coordinate task is more dependent on magnocellular processing than is the categorical task. Although Figure 5 presents the results in terms of background color, it is not clear from this experiment how much the color-condition effect is actually attributable to the background color and how much might be attributable to the color of the stimuli. In view of the fact that metacontrast masking and perception of stroboscopic motion are both reduced when black stimuli are presented against a red background, we suspected that much of the effect is likely produced by background color. In
100
Hellige 900
800 I I
700 RT
600
m
--
500
CAT
COOR
400
Green
Red Background
Color
Figure 5. Reaction time for categorical (CAT) and coordinate (COOR) spatial processing tasks as a function of background color in the red-on-green versus greenon-red experiment. order to separate effects of background color from effects of stimulus color, we have recently completed an additional experiment in which color is presented only in the background or only in the stimulus. That is, there are four color conditions: black lines and dots presented on a red background; black lines and dots presented on a green background; red lines and dots presented on a black background; and green lines and dots presented on a black background. Although Figure 5 presents the results in terms of background color, it is not clear how much the color-condition effect is actually attributable to the background color and how much might be attributable to the color of the stimuli. In view of the fact that metacontrast masking and perception of stroboscopic motion are reduced when black stimuli
Spatial Processing
101
900
800
700 RT
B'-"""
"
600 .
I
CAT
.
500
400
,_,'-I
,
I
COOR
!
Green
Red Background
Color
Figure 6. Reaction time for categorical (CAT) and coordinate (COO) spatial processing tasks as a function of background color in the black-on-green versus black-on-red experiment.
are presented against red backgrounds, we suspected that much of the effect is likely produced by background color. In order to separate the effects of background versus stimulus color, we have recently completed an additional study in which color is presented only in the background or only in the stimulus (i.e., there are four color conditions: black lines & dots presented on a red background; black lines & dots presented on a green background; red lines & dots presented on a black background; and green lines & dots presented on a black background). Figure 6 shows reaction time for categorical and coordinate tasks as a function of background color in the case where the line and dot stimuli were black. Note that the Task by Color interaction is very similar to (and not significantly different from) that shown in Figure 5.
102
Hellige 900
800
700
RT 600 1-----'-'m
1 .... El
500
400
I
CAT COOR
I
Green
Red
Background Color Figure 7. Reaction time for categorical (CAT) and coordinate (COOR) spatial processing tasks as a function of stimulus color in the green-on-black versus redon-black experiment.
There were also differential effects of stimulus color on the two spatial processing tasks, though the direction of the effect is opposite what would be expected if stimulus color contributed to the Task by Color interaction shown in Figure 5. Figure 7 shows reaction time for both tasks as a function of stimulus color for black backgrounds. Note that, for the coordinate task, reaction time was longer to red stimuli than to green stimuli, whereas stimulus color had no effect on performance of the categorical task. Thus, performance of the coordinate (but not the categorical) task is disrupted when the only color in the display is red, regardless of whether the color is restricted to the background or restricted to the stimulus. When different colors are present in the
Spatial Processing
103
Figure 8a. Illustration of a low-pass version of a stimulus from Figure 4. Although this figure provides a reasonable illustration of a band-pass filtered stimulus, the illustration is not completely identical to the stimuli as they appem~ on the viewing screen.
stimulus and in the background, the background color seems to be far more important--possibly because the continuously present background color is far more important for determining the relative efficiency of processing along magnocellular and parvocellular visual pathways. Elizabeth Roth has also presented observers with spatially filtered versions of the stimuli shown in Figure 4. Of particular importance is the distinction between low-pass stimuli, which only contain spatial frequencies lower than or equal to 2 cycles per degree (cpd) of visual angle, and high-pass stimuli, which only contain spatial frequencies equal to or higher than 8 cpd (see Figures 8a [low-pass] and 8b [highpass]). The high-pass condition (low spatial frequencies are eliminated) is similar to the use of a red background, in the sense that it is likely to attenuate processing along the magnocellular pathway--at least relative
104
Hellige
Figure 8b. Illustration of a high-pass version of a stimulus from Figure 4. Although this figure provides a reasonable illustration of a band-pass filtered stimulus, the illustration is not completely identical to the stimuli as they appemed on the viewing screen.
to the low-pass condition. From this perspective, it is interesting that reaction time is significantly longer in the high-pass condition than in the low-pass condition for the coordinate task, but not for the categorical task (see Figure 9). However, the different effects for categorical and coordinate tasks must be treated with caution, as the Task by Filtering Condition interaction was not statistically significant. The results from the foregoing set of experiments can be summarized in the following way. For the categorical processing tasks, only dioptric blurring (which is likely to be more disruptive of processing along the parvocellular pathway than of processing along the magnocellular pathway) had a significant detrimental effect. There was no such effect of dioptric blurring for a coordinate processing task. However, a red background (and red stimuli, when the background was
Spatial Processing 105 800
700
RT
600it ~
--
CAT
m
I"!
,
~
C O O R .
500
.
.
I
.
.
,,,
,,,
I
High-Pass
Low-Pass Filtering
Condition
Figure 9. Reaction time for categorical (CAT) and coordinate (COOR) spatial processing tasks using low-pass versus high-pass versions of the stimuli.
black) was significantly disruptive relative to a green background (and green stimuli, when the background was black). In addition, the use of high-pass stimuli was significantly disruptive relative to the use of lowpass stimuli. Although other interpretations of these effects may be possible, it is interesting that the stimulus conditions that produced significant disruption for the coordinate task were chosen to attenuate processing along the magnocellular visual pathway. With respect to hemispheric asymmetry for processing different types of spatial information, it is important to note that all of these experiments produced a Task by Hemisphere interaction similar to that reported in previous experiments. That is, for the coordinate task,
106 Hellige reaction time was faster on LVF/RH trials than on RVF/LH trials whereas the reverse pattern was found for the categorical task. However, none of these perceptual manipulations influenced the magnitude of the Task by Hemisphere interaction or the magnitude of the hemispheric differences for either the categorical or coordinate tasks considered individually. That is, none of the Task by Hemisphere by Color condition interactions and none of the Task by Hemisphere by Filtering Condition interactions even approached statistical significance 1. With respect to the hypotheses derived from Kosslyn's computational model (Kosslyn et al., 1992), the results are somewhat mixed. The pattern of perceptual effects on the two tasks is consistent with the hypothesis that the magnocellular pathway is more important for coordinate than for categorical spatial processing. In view of what is known about receptive field sizes, this is consistent with the simulation models indicating that coordinate spatial information is computed more efficiently when input is filtered through units with relatively large receptive fields than through units with relatively small receptive fields, whereas the opposite may be the case for computing categorical spatial information. At the same time, the perceptual manipulations do not change the hemispheric asymmetries. Or, to put it in a slightly different way, the perceptual manipulations seem to have the same effect on both hemispheres. This suggests that, for both categorical and coordinate tasks, the two hemispheres rely on the same visual information and on the same computational mechanisms as each other--though they do not always use that information with equal efficiency. For example, a red background interferes with coordinate spatial processing equally in both hemispheres. This suggests that coordinate processing depends on the magnocellular visual pathway to the same extent in both hemispheres. Nevertheless, there is a clear LVF/RH advantage for the coordinate processing tasks. The fact that this fight-hemisphere advantage does not change with the perceptual manipulations that we have used suggests that it arises at a level of processing beyond the early sensory level, a general conclusion that is consistent with what is known about a variety of other types of hemispheric asymmetry (for discussion, see Bradshaw, 1989; Hellige, 1993; Moscovitch, 1986). To be sure, we must be cautious in speculating about the anatomical locus of these hemispheric asymmetries for components of spatial processing. It is worth noting, however, that diffuse red light has been found to suppress activity of certain neurons (so-called Type IV
Spatial Processing
107
neurons) not only in the retinal ganglia and in cells within the lateral geniculate nucleus but also in the magnocellular layers of Area 17 in the primate visual cortex (Livingstone & Hubel, 1984). In view of the fact that the color manipulations had similar effects on LVF/RH and RVF/LH trials, this suggests that the hemispheric asymmetries of the sort we have studied arise at some level subsequent to Area 17. The pattern of results just described is also generally consistent with other recent findings that question low-level, wired-in interpretations of visual laterality. For example, based on visual half-field experiments in which observers indicated whether two successive line segments had the same orientation, Kosslyn, Anderson, Hillger and Hamilton (1994) reject the idea of hemispheric differences in strength of the projection of the magnocellular ganglion cells or that there are wired-in hemispheric differences in the sizes of receptive fields. Instead, they argue that, while there appears to be a bias for the right hemisphere to attend to larger patterns and the left hemisphere to attend to smaller patterns, this is an attentional bias that can be overcome by specific task demands. With all of the foregoing results in mind, it is useful to consider an interesting alternative account of hemispheric asymmetry for processing visual information, including information about categorical and coordinate spatial relationships. Ivry and Robertson (1997) have proposed a general model of perceptual asymmetry referred to as Double Frequency by Filtering or DFF. While a detailed account of the DFF model is beyond the scope of the present chapter, I will summarize briefly its critical elements as applied to vision and its application to hemispheric asymmetry for spatial processing. An important feature of the DFF model is the proposal that hemispheric asymmetries are sensitive to relative rather than absolute spatial frequencies. In addition, the emphasis on attentional selection is especially interesting in view of the conclusions reached by Kosslyn et al. (1994). According to the DFF model, initial sensory processing is identical in the two cerebral hemispheres, so that asymmetries arise at a postsensory level. An attentional mechanism is proposed to select taskrelevant information by choosing the region of the sensory spectrum to be enhanced for further processing. In vision, this selection is proposed to be in terms of spatial frequency and in audition this selection is proposed to be in terms of temporal frequency. This first frequencyfiltering stage is hypothesized to be identical for the two hemispheres. Hemispheric asymmetries are hypothesized to arise in a subsequent,
108 Hellige second frequency-filtering stage. During this second filtering stage, low spatial and temporal frequencies are enhanced in the right hemisphere and high spatial and temporal frequencies are enhanced in the left hemisphere. Because this second filtering stage operates only on the information selected by the first filtering stage, the DFF model predicts hemispheric asymmetry in terms of relative rather than absolute spatial and temporal frequencies; specifically, left hemisphere superiority for processing relatively high spatial and temporal frequencies and right hemisphere superiority for processing relatively low spatial and temporal frequencies. Ivry and Robertson review several examples of hemispheric asymmetry that are consistent with the DFF model. For present purposes, it is particularly important to consider a DFF account of hemispheric asymmetry for categorical and coordinate spatial processing. As Ivry and Robertson (1997) note, the DFF model emphasizes the spatial frequency information required to perform the different spatial relationship tasks without any particular regard for whether the task would be described as categorical or coordinate. To the extent that one task requires higher spatial frequencies than does the other task, a RVF/LH advantage is predicted for the first task and a LVF/RH advantage is predicted for the second task. As an example of how this would operate, they consider two tasks employed by Kosslyn et al. (1989) presenting a dot and an amorphous blob on each trial. As discussed earlier, the categorical task required observers to indicate whether the dot was touching the blob and the coordinate task required observers to indicate whether the dot was within 2 mm of the blob. As Ivry and Robertson note, in the Kosslyn et al. experiment, the spacing between the dot and the blob was 0, 1 or 10 mm. Thus, the 0 mm and 10 mm stimuli were assigned to different response categories for both tasks, but the assignment of the 1 mm stimulus changed across tasks. For the categorical task, the 1 mm stimulus was assigned to the same category as the 10 mm stimulus, so that the most difficult discrimination was between the 0 mm and 1 mm stimuli. For the coordinate task, the 1 mm stimulus was assigned to the same category as the 0 mm stimulus, so that the most difficult discrimination was between the 1 mm and 10 mm stimuli. Note that, in this case, a much finer perceptual discrimination was required for the categorical task than for the coordinate task. Perhaps for this reason, overall performance was worse for the categorical task than for the coordinate task. If we assume that a finer
Spatial Processing
109
discrimination depends on higher spatial frequencies than does a less fine discrimination, then the DFF model provides an alternative account of the Task by Hemisphere interaction without regard to the categorical versus coordinate nature of the tasks. In support of the DFF account, Ivry and Robertson (1997) describe an experiment by Ivry, Prinzmetal and Hazeltine that is similar to the one just described, but that included two different spacing conditions. The first condition was similar to that used by Kosslyn et al. (1989), with the spacing between the dot and the blob being 0, 1 or l0 mm. As was the case in the Kosslyn et al. experiment, the categorical task required observers to make one response to the 0 mm stimuli and a different response to the 1 and l0 mm stimuli and the coordinate task required observers to make one response to the 0 and 1 mm stimuli and a different response to the l0 mm stimulus. In the second spacing condition, the spacing between the dot and the blob was 0, 8 and l0 mm. A categorical task required observers to make one response to the 0 mm stimuli and a different response to the 8 and l0 mm stimuli. A coordinate task required observers to make one response to the 0 and 8 mm stimuli and a different response to the 10 mm stimuli. Note that, in this second spacing condition, the coordinate task required a finer discrimination than the categorical task, so the DFF model predicts a reversal of the hemispheric asymmetries. In fact, Ivry and colleagues found exactly the sort of Spacing Condition by Task by Hemisphere interaction that would be expected--though this experiment alone does not indicate whether the relevant variable is absolute or relative spatial frequency and all laterality effects were restricted to the first trial block (cf., Kosslyn et al., 1989). If overall performance is taken as a measure of the difficulty of the perceptual discrimination, then the DFF model would seem to predict a RVF/LH advantage for whichever task leads to worse performance and a LVF/RH advantage for whichever task leads to better performance. As noted earlier, this is clearly not the case. In fact, for most of the studies cited earlier, coordinate tasks have led to considerably worse overall performance than categorical tasks (e.g., Cowin & Hellige, 1994; Hellige & Michimata, 1989; Kosslyn et al., 1989, line-and-dot tasks; Michimata, 1997). Despite this, the coordinate tasks have produced a significant LVF/RH advantage and the categorical tasks have produced a trend toward a RVF/LH advantage. However, while overall performance may be a reasonable measure of task difficulty, it is not necessarily the best
1 10 Hellige measure of the extent to which a task depends on lower versus higher spatial frequencies--and it is the frequencies that are more relevant for generating predictions from the DFF model. For example, although Cowin and Hellige found that their Near/Far task led to more errors and longer reaction times than their Above/Below task, dioptric blurring (which impairs processing of relatively high spatial frequencies) interfered with performance only for the Above/Below (categorical) task. Thus, there is a dissociation of overall performance and whether one task is more or less dependent than another task on high spatial frequencies. Consequently, differences in the overall level of perfor, mance of two tasks is of limited value in generating predictions from the DFF model. Ivry and Robertson (1997) offer additional support for the plausibility of a DFF account of hemispheric differences in spatial processing in the form of a set of neural network simulations. The simulations are based on the sort of line-and-dot stimuli described earlier and, in fact, use dot spacings patterned after those used by Hellige and Michimata (1989) and illustrated in Figure 1. Among other things, these simulations demonstrate that hemispheric differences in spatial processing tasks can emerge as a byproduct of hemispheric differences in processing relatively high versus relatively low spatial frequencies. In addition, these simulations demonstrate that such things as the actual spacing of stimuli in a set can be expected to influence the hemispheric asymmetry that is observed. Presentation of the details of the neural network simulations is beyond the scope of the present chapter. However, a critical portion of the neural network architecture can be summarized in the following way. Input is filtered through spatial frequency channels by using input layers whose units differ in the size of their receptive fields--with smaller receptive field sizes being associated with higher spatial frequencies than larger receptive field sizes. Although this is accomplished in different ways by the DFF simulation and by the neural network simulations presented by Kosslyn et al. (1992), both types of simulation rely on input units with different size receptive fields. A critical difference is the extent to which these two models emphasize hemispheric differences in processing absolute versus relative ranges of spatial frequency. By suggesting that the hemispheres may differ in receptive field size, the model presented by Kosslyn et al. emphasizes absolute spatial frequency differences in the processing properties of the two hemispheres. In contrast, in the DFF model, the hemispheres are
Spatial Processing 111 identical in their processing of different ranges of absolute spatial frequency, but selective attention operates in a way that makes the hemispheres differ in the processing of relative spatial frequencies. Although more theoretical and empirical work is needed to separate the DFF model from other computational possibilities, it is instructive to reconsider the effects of perceptual manipulations (e.g., blurring, background color, spatial frequency filtering) from the perspective provided by the DFF model. From this perspective, it is interesting that manipulations which should be more disruptive of low spatial frequency information (e.g., red background, high-pass stimuli) selectively disrupted a coordinate spatial processing task whereas a manipulation (dioptric blurring) that should be more disruptive of high spatial frequency information selectively disrupted a categorical spatial processing task. It is clear that these perceptual manipulations would be expected to influence the absolute spatial frequencies that are available for processing. The influence on relative frequency is more complicated. To be sure, such things as filtering operations affect relative frequencies within a stimulus in the following way. Consider a complex stimulus containing spatial frequencies between 0.1 cpd and 32 cpd. If such a stimulus is filtered to remove all frequencies below 8 cpd, the 8 cpd component becomes a relatively low frequency but if the stimulus is filtered to remove all frequencies above 8 cpd, the 8 cpd component becomes a relatively high frequency. However, in view of the fact that all of these stimuli contain a range of spatial frequencies, the filtering operation does not necessarily change whether the spatial frequencies required by two tasks are high or low relative to each other. With this in mind, it is interesting that the Task by Hemisphere interaction was not influenced by the perceptual manipulations outlined earlier. While these results can be accommodated by the DFF account, it must be noted that these experiments were not designed to be a test of the DFF model or to discriminate between the DFF model and other possibilities. The new results and simulations summarized in this section suggest that hemispheric asymmetry for categorical versus coordinate spatial processing is related to more general differences between the two hemispheres in aspects of visual information processing. Further consideration of these mechanisms and others that might underlie hemispheric asymmetry for processing spatial relationships can benefit from recent extensions of the categorical/coordinate distinction beyond the domain
112
Hellige
of explicit spatial localization. Accordingly, I turn now to a review of several relevant studies.
Extensions of the Categorical/Coordinate Distinction In the studies reviewed so far, observers have been required to make explicit judgments of categorical or coordinate spatial relationships. In this section, I review evidence for Categorical/Coordinate by Hemisphere interactions that have arisen in studies in which the spatial judgments are either implicit or completely absent. Following this brief review, I will consider the extent to which these various results can be treated as different manifestations of the same underlying hemispheric asymmetry. Laeng (1994) examined categorical and coordinate processing in patients with unilateral stroke and in a control group with no known neurological damage. In one experiment observers were first shown, on each trial, a drawing of one or more objects (e.g., a large cat to the left of a small cat). Following a delay of approximately 5 s, the observers were shown the original drawing and a drawing in which the objects were transformed in either a categorical way (e.g., a large cat to the fight of a small cat) or a coordinate way (e.g., a large cat to the left of a small cat, with the distance between them being larger than in the original drawing). The observer was instructed to choose the drawing that was identical to the original. In a second experiment, the same observers were shown an original drawing followed by two choices, neither of which was identical to the original. One choice contained a categorical change (e.g., left-right reversal) and the other contained a coordinate change (e.g., distance between the objects). The task of the observer was to indicate which version looked more similar to the original drawing. Note that neither of these tasks requires an explicit identification of spatial relations. In the first experiment, patients with left-hemisphere stroke tended to mistake the categorical transformation for the original whereas patients with right-hemisphere stroke tended to mistake the coordinate transformation for the original. In the second experiment, patients with left-hemisphere stroke judged the categorical transformation to be more similar than the coordinate transformation to the original stimulus whereas patients with right-hemisphere stroke showed the opposite bias.
Spatial Processing
1 13
Similar stimuli have been used in visual half-field studies with neurologically intact right-handed and left-handed observers (e.g., Laeng & Peters, 1995). Specifically, observers indicated whether a brief, lateralized drawing was identical to a drawing seen immediately before in free vision. When the two drawings were not identical, the new drawing contained the same objects as the original with either a categorical or a coordinate aspect of the drawing changed. In general, right-handed observers were faster to identify categorical changes on RVF/LH trials and coordinate changes on LVF/RH trials. There were no visual field differences for left-handed observers. Christman (1997) has provided preliminary evidence that the categorical/coordinate distinction extends to the processing of dynamic visual input. On each trial, observers were presented with a dot in one visual field that appeared to either "shrink" or "grow" and to change size either quickly or slowly. That is, over a 167 ms time period, the diameter of the dot changed by either 1.0 degrees of visual angle ("quickly") or 0.5 degrees of visual angle ("slowly"). Thus, the perceived rate of change was perfectly correlated with the spatial magnitude of the change. Christman hypothesized that indicating whether the dot shrunk or grew involved categorical processing, whereas indicating whether the rate of size change was fast or slow involved coordinate processing. The percentage of errors was significantly smaller for the categorical task than for the coordinate task and, more importantly, there was a significant Task by Hemisphere interaction. For the coordinate task there was a significant LVF/RH advantage and for the categorical task there was a nonsignificant trend in the opposite direction. The Task by Hemisphere interaction was not significant for reaction time of correct responses, perhaps because the error rates in some conditions were quite high (e.g., 40% or so). Niebauer and Christman (1996) have attempted to extended the categorical/coordinate distinction to a task that requires judgments of spatial frequency relations in sinusoidal gratings. Observers indicated whether the second of two successively presented gratings was higher or lower in spatial frequency than the first (after Kitterle & Selig, 1991). In one condition, the two gratings differed by 1.0 octave. Because a difference of 1.0 octave could be easily characterized (e.g., each grating could be easily categorized as "thick" o r "thin"), Niebauer and Christman consider this to be a categorical condition. In a second condition, the two gratings differed by only 0.125 octave. Because this
114 Hellige involves a much finer discrimination and would require processing of precise spatial frequencies, Niebauer and Christman consider this to be a coordinate condition. From this perspective, it is interesting that there was a significant LVF/RH advantage for the coordinate task and no significant hemispheric asymmetry for the categorical task. Weiner and Christman (1994) have also examined hemispheric asymmetry for the processing of what they refer to as categorical and coordinate auditory pitch relations. The results of this study are especially interesting because the tasks do not involve visual stimuli and do not demand the processing of spatial relationships (either explicitly or implicitly). Listeners were presented with a baseline tone (800 Hz) binaurally followed by a comparison tone to one ear or the other (the ear not receiving the comparison tone received white noise). The comparison tone was either 450, 600, 1067 or 1356 Hz. Note that two of the comparison tones (1067 and 1356 Hz) were higher in pitch than the baseline tone and two of the comparison tones (450 and 600 Hz) were lower in pitch than the baseline tone. In addition, two of the comparison tones (600 and 1067 Hz) were "near" in pitch to the comparison tone and two of the comparison tones (450 and 1356 Hz) were "far" in pitch from the comparison tone. For the categorical task, listeners indicated whether the pitch of the comparison tone was higher or lower than the pitch of the baseline tone. For the coordinate task, listeners indicated whether the pitch of the comparison tone was near to or far from the pitch of the baseline tone. Weiner and Christman reported a highly significant Task by Hemisphere (Ear) interaction, with there being a significant right-ear/left-hemisphere advantage for the Above/ Below task and a significant left-ear/fight-hemisphere advantage for the Near/Far task. Based on these pitch processing results, Weiner and Christman suggest that modes of categorical versus coordinate processing reflect general aspects of hemispheric asymmetry that are not specifically tied to visual stimuli. To be sure, it might be possible for listeners to use visual imagery to represent the pitch of the tones to themselves as points on a unidimensional vertical continuum, with higher pitches represented by points higher on the continuum. But the use of an imagery analog to the line-and-dot stimuli is not required. The DFF model could account for the Weiner and Christman results to the extent that the categorical task required processing of relatively higher temporal frequencies and the coordinate task required processing of
Spatial Processing
1 15
relatively lower temporal frequencies. However, it is not obvious (at least to me) whether this is the case. It is also useful to consider hemispheric asymmetry for visual shape processing tasks that differ in ways that seem analogous to the categorical/coordinate distinction with respect to spatial relationships. For example, Marsolek (1995) had observers learn to categorize novel visual stimuli by presenting them with specific exemplars that were variations on a set of unseen prototypes. After an initial learning phase, a single item was presented to the LVF/RH or RVF/LH on each trial and the observer was required to place it into the correct category. Marsolek found a LVF/RH advantage for the classification of the "old" exemplars (those used during training) and a RVF/LH advantage for the classification of the previously unseen prototypes that had been used to define the categories. In addition, on LVF/RH trials observers responded faster to old exemplars than to new, previously unseen, exemplars, with the results for prototypes being intermediate. On RVF/LH trials, observers responded faster to the previously unseen prototypes than to either type of exemplar, with there being no difference between old and new exemplars. Based on this pattern of results and on a variety of other findings, Marsolek suggests that the right hemisphere is superior for processing specific shape information of the sort that would be needed to distinguish among the exemplars within a single category whereas the left hemisphere is superior for extracting and identifying the categorydefining prototype. Jacobs and Kosslyn (1994) have extended the neural network models of Kosslyn et al. (1992) to examine the possible importance of receptive field sizes in the encoding of shape information. They found that networks that received input from units with large, overlapping receptive fields coded the identity of specific shapes better than networks that received input from units with small, nonoverlapping receptive fields. Exactly the opposite was found for the assignment of shapes to categories (see also Brown & Kosslyn, 1995). Thus, there would seem to be a connection between categorical spatial processing and the assignment of shapes to categories without regard for distinctions among the exemplars within a category and a connection between coordinate spatial processing and the ability to distinguish among the specific exemplars. In addition, Marsolek, Kosslyn and Squire (1992) have reported that, at least for words, a highly formspecific type of visual priming is restricted to the fight hemisphere (see
1 16 Hellige also Marsolek, Schacter & Nicholas, 1996). Specifically, in word-stem completion priming tasks there is more priming if the case of the letters remains the same across all phases of the experiment than if the letter case changes--but only on LVF/RH and not RVF/LH trials. Jacobs and Kosslyn relate this effect to their neural network models by suggesting that form-specific priming depends on information filtered through units with large, overlapping receptive fields. What common thread, if any, connects the experiments reviewed in this section to each other and to studies that require explicit judgments about spatial relationships? It is not difficult to see a connection between studies in which the spatial judgments are explicit and studies in which the spatial judgments are implicit (e.g., Laeng, 1994; Laeng & Peters, 1995; Christman, 1997). It is not so clear, however, what thread might connect these studies to studies of pitch relations in audition (e.g., Wiener & Christman, 1994) and to certain types of shape processing (e.g., Marsolek, 1995). One possibility is the distinction between representations that treat a wide range of stimuli as an equivalence class (a "categorical" representation) and representations that preserve discriminable differences among stimuli within an equivalence class ("coordinate" representations). While this distinction applies to the two types of spatial relationships that we have discussed, it also applies to the other dimensions illustrated in the present section. Indeed, it is this more general distinction, or something close to it, that motivated much of the research reviewed in this section and that leads the various authors to connect their work to the categorical/coordinate distinction in spatial processing. In fact, this general conceptual distinction contributed to Kosslyn's (1987) original formulation of the categorical/coordinate distinction in spatial processing. In view of this thread that runs through the various studies, it is useful to conclude the present chapter by considering what type of general mechanism could underlie all of these results and to consider promising directions for future research. Concluding Comments: More on Mechanisms and Future Directions
In this final section, I reconsider the mechanisms that may underlie hemispheric asymmetry for spatial processing in view of the additional findings just reviewed. In doing so, I also point out some of the important questions that need to be resolved and what would seem to be fruitful directions for future research.
Spatial Processing
1 17
As noted at the outset, a considerable amount of theoretical and empirical work has dealt with the distinction between categorical and coordinate spatial relations. I believe that much of this work has been fruitful, even though specific interpretations continue to evolve and significant theoretical issues remain. Several years ago, Justine Sergent and I (J. Sergent & Hellige, 1986) argued that, in building theories of hemispheric asymmetry, it is important to specify the characteristics of the input on which the brain operates and to move toward descriptions of hemispheric asymmetry in terms of the operations on that input that are actually performed by the brain. From this perspective, I believe that we have made progress in understanding hemispheric differences in spatial processing by acknowledging that different spatial tasks require different aspects of perceptual input to be performed with maximum efficiency. Thus, hemispheric asymmetry for the different types of spatial processing can be viewed as manifestations of underlying mechanisms that have the potential to account for other aspects of hemispheric asymmetry as well. It has been suggested that the type of asymmetries reviewed in the immediately preceding section are manifestations of the same underlying mechanisms that produce Task by Hemisphere interactions when explicit judgments about spatial location are required. To the extent that this is the case, the various perceptual manipulations (e.g., blurring, background colors, frequency-filtering) would be expected to have similar effects across these various experimental paradigms--including those in which judgments about spatial location are implicit rather than explicit (e.g., Laeng & Peters, 1995; Niebauer & Christman, 1996) and those in which there may be no spatial judgments at all (e.g., Marsolek, 1995). In designing such studies, it will be important to include conditions that have the potential to discriminate between the effects of absolute and relative spatial frequency (cf., Kosslyn et al., 1992; Ivry & Robertson, 1997). It is also important to know whether the distinction between categorical and coordinate spatial relationships is important for determining hemispheric asymmetry outside the visual modality. To be sure, Weiner and Christman (1994) report an analog of sorts for auditory processing, but their task involved pitch relations rather than spatial relations. It would be useful to have studies that make explicit, spatial processing demands. For example, within both auditory and tactile domains it is possible to present lateralized stimuli that come from
1 18 Hellige different locations in space and manipulate whether the required spatial judgment is categorical (e.g., in front of versus in back of) or coordinate (within a specified distance). Finding similar Task by Hemisphere interactions across stimulus modalities would motivate the search for mechanisms that transcend any single modality. It is difficult, of course, to speculate about the sort of mechanisms that might transcend specific modalities or that might connect the various findings reviewed in the present chapter. Nevertheless, I think there are plausible candidates that deserve careful consideration. For example, Kosslyn et al. (1992) were motivated to vary receptive field size in the input units of their neural network simulations (and the extent to which the receptive fields of different units overlap) by considering whether there might be hemispheric differences in what has been termed coarse coding. In general, coarse coding refers to a situation in which coarse processing at one level results in very precise processing at another level. A prototypical example is color vision, in which precise coding of color emerges from only three types of cones in the retina with overlapping distributions of sensitivity centered at different wavelengths of light. There are many examples of this sort of coarse coding within the sensory domains (e.g., Hinton, McClelland & Rumelhart, 1986). In view of the fact that coarse coding of various sorts is used widely as a computational strategy by the brain, any systematic hemispheric differences in the use of coarse coding could account for the fact that subtle, complementary hemispheric asymmetries occur for a broad range of domains including motor performance, language, spatial processing and emotion (see Hellige, 1993). And if it is the case that coarse coding provides an effective computational mechanism for preserving discriminable differences among stimuli within an equivalence class, as some neural network simulations have suggested (e.g., Jacobs & Kosslyn, 1994; Kosslyn et al., 1992), then hemispheric differences in the use of coarse coding could underlie the sort of results reviewed in the immediately preceding section of the present chapter. In fact, it has even been hypothesized that hemispheric differences in semantic processing come about because of relatively coarse coding of semantic information in the right hemisphere and relatively fine coding of semantic information in the left hemisphere (Beeman, Friedman, Grafman, Perez, Diamond & Lindsay, 1994). I began the present chapter by observing that the two hemispheres do not have identical ability to localize stimuli in space. Research in the
Spatial Processing
119
last 10 years has done a great deal to refine our conceptualization of hemispheric asymmetry for spatial processing and to focus attention on m e c h a n i s m s that might underlie the asymmetries that have been documented. There are, at the present time, several promising leads that need to be explored. We are fortunate to have at our disposal an increasing range of converging operations to assist in that exploration; techniques for the study of behavioral asymmetries in neurologically intact individuals and in patients, neural-network modeling techniques, a variety of structural and functional brain-imaging techniques, and so forth. Each of these techniques is useful for differentiating among the hypotheses considered here and together they provide a powerful means of exploring the leads that have been identified. Although some of the leads are likely to take us to dead ends and others will lead us to more complexity than we imagine, we will undoubtedly learn a great deal about spatial processing and hemispheric asymmetry along the way. Notes 1. Cowin and Hellige (1995) reported that, for the green-on-red versus red-on-green experiment, there was a Background Color by Hemisphere interaction on the very first 24-trial block with reaction time as the dependent variable. Specifically, averaged across the two spatial tasks, there tended to be a LVF/RH advantage when the background was red and a RVF/LH advantage when the background was green. We are not inclined to treat this effect as reliable for the following reasons. In an overall analysis from the same experiment that included trial block as an independent variable, there was no hint whatsoever of either a Background Color by Hemisphere interaction or a Trial Block by Background Color by Hemisphere interaction. That is, the omnibus analysis provided no support for the claim that there is a Background Color by Hemisphere interaction at all on any block of trials. Furthermore, when the background was red, an analysis restricted to the first trial block indicated that there were significantly fewer errors on RVF/LH than on LVF/RH trials--a visual field difference exactly opposite that found with reaction time as the dependent variable. In addition, even when analysis was restricted to the first trial block, there was no hint of any Color Condition by Hemisphere interaction (or Filtering Condition by Hemisphere interaction) for any of our subsequent experiments reported in the present section--and this was the case for both reaction time and percentage of errors. That is, we have not replicated this particular effect. Finally, in none of the experiments was there ever any indication that the theoretically relevant Task by Hemisphere interaction changed with perceptual manipulations.
120 Hellige References
Beeman, M., Friedman, R. B., Grafman, J., Perez, E., Diamond, S., & Lindsay, M. B. (1994) Summation priming and coarse coding in the fight hemisphere. Journal of Cognitive Neuroscience, 6, 26-45. Boles, D. B. (1991) Factor analysis and the cerebral hemispheres: Pilot studyand parietal functions. Neuropsychologia, 29, 59-92. Boles, D. B. (1992) Factor analysis and the cerebral hemispheres: Temporal, occipital and frontal functions. Neuropsychologia, 30, 963 -988. Bradshaw, J. L. (1989) Hemispheric specialization and psychological function. Chichester, England: Wiley. Breitmeyer, B. G., & Williams, M. C. (1990) Effects of isoluminantbackground color on metacontrast and stroboscopic motion: Interactions between sustained (P) and transient (M) channels. Vision Research, 30, 1069-1075. Breitmeyer, B. G., May, J. G., & Heller, S. S. (1991) Metacontrast reveals asymmetries at red-green isoluminance. Journal of the Optical Society of America, 8, 1324-1329. Brown, H. D., & Kosslyn, S. M. (1995) Hemispheric differences in visual object processing: Structural versus allocation theories. In R. J. Davidson & K. Hugdahl (Eds.), Brain Asymmetry (pp. 77-122). Cambridge MA: MITPress. Christman S. (1989) Perceptual characteristics in visual laterality research. Brain and Cognition, 11,238-257. Christman S. (1990) Effects of luminance and blur on hemispheric a~yammetries in temporal integration. Neuropsychologia, 28, 3619, J
I
-'1"o
Christman, S. (1997) Hemispheric asymmetry in categorical versus coordinate processing of dynamic visual input. Paper presented at the Annual Meeting of the International Neuropsychological Society, Orlando, FL. Christman, S., Kitterle, F. L., & Hellige, J. B. (1991) Hemispheric asymmetry in the processing of absolute versus relative spatial frequency. Brain and Cognition, 16, 62-73. Cook,N. D., Fruh H., & Landis T. (1995) The cerebral hemispheres and neural network simulations: Design considerations. Journal of
Experimental Psychology: Human Perception and Performance, 21,
410-422. Cowin E. L., Hellige J. B. (1995) Spatial processing and hemispheric asymmetry: Contribution of the magnocellular system. Paper presented at the Annual Meeting of the Psychonomic Society, Los Angeles, CA. Cowin, E. L., & Hellige, J. B. (1994) Categorical versus coordinate spatial processing: Effects of blurring and hemispheric asymmetry. Journal of Cognitive Neuroscience, 6, 156-164. Crebolder, J. M., & Bryden, M. P. (1996) Assessing verbal and spatial hemispheric performance with identical material. Brain and Cognition, 30, 383-385.
Spatial Processing
121
Delis D. C., Robertson L. C., & Efron R. (1986) Hemispheric ecialization of memory for visual hierarchical stimuli. uropsychologia, 24, 205-214. DeMonasterio, F. M., & Schein, S. J. (1980) Protan-like spectral sensitivity of foveal Y ganglion cells of the retina of macaque monkeys. Journal of Physiology, London, 299, 385-396. DeMonasterio, F. M. (1978) Center and surround mechanisms of opponent-color X and Y ganglion cells of retina of macaques. Journal of Neurophysiology, 41, 1418-1434. Dreher, B., Fukuda, Y., & Rodieck, R. W. (1976) Identification, classification and anatomical segregation of cells with X-like and Ylike properties in the lateral genlculate nucleus of old-world primates. Journal-of Physiology, London, 258, 433-452. Hellige, J. B. (1993) Hemispheric Asymmetry: What's Right and What's Left.. Cambridge: Harvard University Press. Hellige, J. B. (1995) Hemispheric asymmetry for components of visual information processing. In R. J. Davidson & K. Hugdahl (Eds.), Brain Asymmetry (pp. 99-121). Cambridge, MA: MIT Press. Hellige, J. B. (1996) Hemispheric asymmetry for visual information processing. Acta Neurobiologiae Experimentalis, 56, 485-497. Hellige, J. B., & Michimata, C. (1989)Categorization versus distance: Hemispheric differences for processing spatial information. Memory & Cog~nition, 17, 770-776. Hellige, o. B., Bloch, M. I., Cowin, E. L., Eng, T. L., Eviatar, Z., & Sergent, V. (1994) Individual variation in hemispheric asymmetry: Multitask study of effects related to handedness and sex. Journal of Experimental Psychology: General, 123, 235-256. Hellige J. B., & Sergent Ji (1986) Role of task factors in visual field asymmetries. Brain and Cognition, 5, 200-222. Hinton, G. E., McClelland, J. L., & Rumelhart, D. E. (1986) Distributed representations. In D. E. Rumelhart & J. L. McClelland (Eds.),
Parallel distributed processing: Explorations in the microstructure of cognition, Vol. 1 (Pp. 77-109). Cambridge, MA: MIT Press. Ivry, R. & Robertson, L. C. (1997) The two sides of perception.
Cambridge, MA: MIT Press. Jacobs, R. A., & Kosslyn, S. M. (1994) Encoding shape and spatial relations: The role of receptive field size in coordinating complementary representations. Cognitive Science, 18, 361-386. Jonsson, J. E., & Hellige, J. B. (1986)Lateralized effects of blurring: A test of the visual spatial frequency model of cerebral hemisphere asymmetry. Neuropsychologia, 24, 351-362. Kitterle F . L . , & Sellg L. M. (1991) Visual field effects in the discrimination of sine wave gratings. Perception & Psychophysics, 50, 15-18. Kitterle F. L., Christman S., & Conesa J. (1993) Hemispheric differences in the interference among components of compound gratings. Perception & Psychophysics, 54, 785-793. Kitterle, F. L., Christman, S., & Hellige, J. B . (1990) Hemispheric differences are found in the identification but not the detection of low vs. high spatial frequencies. Perception & Psychophysics, 48, 297-306.
122 Hellige Kitterle, F. L., Hellige, J. B., & Christman, S. (1992) Visual hemispheric asymmetries depend on which spatial frequencies are task relevant. Brain and Cognition, 20, 246-258. Koenig, O., Reiss, L. P., & Kosslyn, S. M. (1990) The development of spatial relation representations: Evidence from studies o f cerebral lateralization. Journal of Experimental Child Psychology, 50, 119130. Kosslyn S. M., Chabris C. F., Marsolek C. J., Jacobs R. A., & Koenig O. (1995) On computational evidence for different types of spatial relations encoding: Reply to Cook et al. (1995). Journal of
Experimental Psychology: Human Perception and Performance, 21,
423-431. Kosslyn, S. M. (1987)Seeing and imagining in the cerebral hemispheres: A computational approach. Psychological Review, 94, 148-175. Kosslyn, S. M. (1988) Aspects of a cognitive neuroscience of mental imagery. Science, 240, 1621-1626. Kosslyn, S. M., & Koenig, O. (1992) Wet mind: The new cognitive neuroscience. New York: Free Press. Kosslyn, S. M., Anderson, A. K., Hillger, L. A., & Hamilton, S. E. (1994) Hemispheric differences in sizes of receptive fields or attentional biases? Neurops~chology, 8, 139-147. Kosslyn, S. M., Chabris, C. r., Marsolek, C. J., & Koenig, O. (1992) Categorical versus coordinate spatial relations: Computational analyses and computer simulation. Journal of Experimental Psychology: Human Perception and Performance, 18, 562-577. Kosslyn, S. M., Holtzman, J., Farah, M. J., & Gazzaniga, M. S. (1985) A computational analysis of mental image generation: Evidence from functional dissociations in split-brain patients. Journal of Experimental Psychology_: General, 114, 311-341. Kosslyn, S. M., Koenig, O., Barrett, A., Cave, C. B., Tang, J., & Gabrieli, J. D. E. (1989) Evidence for two types of spatial representations: Hemispheric specialization for categorical and coordinate relations.
Journal of Experimental Psychology: Human Perception and Performance, 15, 723-735.
Kosslyn, S. M., Maljkovic, V., Hamilton, S. E., Horwitz, G., & Thompson, W. L. (1995) Two types of image generation: Evidence for left and fight hemisphere processes. Neuropsychologia, 33, 14851510. Laeng, B. (1994) Lateralization of categorical and coordinate spatial functions: A study of unilateral stroke patients. Journal of Cognitive Neuroscience, 6, 189-203. Laeng, B., & Peters, M. (1995) Cerebral lateralization for the processing o f spatial coordinates and categories in left- and right-handers. Neuropsychologia, 33, 421-439. Laeng, B., Peters, M., & McCabe, B. (1996) A left hemisphere s bias for encoding visuo-spatial categories? Submitted for publication. Lamb M. R., Robertson L. C., & Knight R. T. (1989) Attention and interference in the processing of hierarchical patterns: Inferences from patients with right and left temporal-parietal lesions. Neuropsychologia, 27, 471-483.
Spatial Processing 123 Lamb, M. R., Robertson, L. C., & Kn!ght, R. T. (1990) Component mechanisms underlying the processing of hierarchically organized patterns: Inferences from patients with unilateral cortical lesions.
Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 471-483.
Livingstone M. S., & Hubel D. H. (1988) Segregation of form, color, movement, and depth: Anatomy, physiology, and perception. Science, 240, 740-749. Livingstone M. S., & Hubel, D. H. (1987) Psychophysical evidence for separate channels for the perception of form, color, motion, and depth. Journal of Neuroscience, 7, 3416-3468. Livingstone, M. S., & Hubel, D. H. (1984) Anatomy and physiology of a color system in the primate visual cortex. Journal of Neurosci-ence, 4, 309-356. Marrocco, R. T., McClurkin, J. W., & Young, R. A. (1988) Spatial summation and conduction latency classification of cells in the lateral geniculate nucleus of macaques. Journal of Neuroscience, 2, 12751291. Marsolek, C. J. (1995) Abstract-visual-form representations in the left cerebral hemisphere. Journal of Experimental Psychology: Human
Perception and Performance, 21,375-386.
Marsolel~, C. J,, Kosslyn, S. M., & Squire, L. R. (1992) Form-specific visual priming in the right cerebral hemisphere. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 18,
492-508. Marsolek, C. J., Schacter, D. L., & Nicholas, C. D. (1996) Form-specific yisual priming for new associations in the fight cerebral hemisphere. Memory & Cbgnition, 24, 539-556. Martin M. (1979) Hemispheric specialization for local and global processing. Neuropsychologia, 17, 33-40. Michimata, C. (1997)Hemispheric processing of categorical and coordinate spatial relations in vision and visual imagery. Brain and
Cognition.
Moscovitch, M. (1986) Afferent and efferent models of visual asymmetries: Theoretical and empirical implications. uropsychologia, 24, 91-114. Niebauer, C. L. (1996) On the nature of categorical and coordinate spatial relation representations. Unpublished Ph.D. Dissertation, The University of Toledo, Toledo, OH. Niebauer, C., & Christman, S. (1996) Categorical versus coordinate processing of spatial frequency relations. Submitted for publication. Robertson L. C. (1995) Hemispheric specialization and cooperation in processing complex visual patterns. In F. L. Kitterle (Ed.), Hemispheric communication: Mechanisms and models, (Pp. 301318). Hillsdale, N.J.: Erlbaum Press. Robertson L. C., Lamb M. R., Knight R. T. (1988) Effects of lesions of temporal-parietal junction on perceptual and attentional processing in humans. Journal of Neurosdence, 8, 3757-3769. Robertson L. C., Lamb M. R., Knight R. T. (1991) Normal global-local analysis in patients with dorsolateral frontal lobe lesions. Neuropsychologia, 29, 959-968.
124 Hellige Roth, E. & Hellige, J. B. (1997) Unpublished manuscript. Rybash, J. M., & Hoyer, W. J. (1992) Hemispheric specialization for categorical and coordinate spatial representations: A reappraisal. Memory & Cognition, 20, 271-276. Schiller, P. H., & Malpeli, J. G. (1978) Functional specificity of lateral geniculate nucleus laminae of the rhesus monkey. Journal of Neurophysiology, 41,788-797. Sergent J. (1982) The cerebral balance of power: Confrontation or cooperation? Journal of Experimental Psychology: Human Perception and Performance, 8, 253-272. Sergent J. (1983) The role of the input in visual hemispheric asymmetries. Psychological Bulletin, 93, 481-514. Sergent J. (1987) Failures to confirm the spatial-frequency hypothesis: Fatal blow or healthy complication? Canadian Journal of Psychology, 41,412-428. Sergent, J. (1991) Judgments of relative position and distance o n representations of spatial relations. Journal of Experimental Psychology: Human Perception and Performance, 17, 762-780. Sergent, J., & Hellige, J. B. (1986) Role ot" input factors in visual-field asymmetries. Brain and Cognition, 5, 174-199. Shapley R. (1994) Parallel neural pathways and visual function. In M. S. Gazzaniga (Ed.), The cognitive neurosciences, (Pp. 315-324). Cambridge, MA: MIT Press. Van Essen, D. C. (1985) Functional organization of primate visual cortex. In E. G. Jones & A. A. Peters (Eds.), Cerebral cortex, (Vol. 3, Pp. 259-329). New York: Plenum Press. van Kleeck M. H. (1989) Hemispheric differences in global versus local processing of hierarchical visual stimuli by normal subjects: New data and a meta-analysis of previous data. Neuropsychologia, 27, 11651178. Weiner, R., & Christman, S. (1994) Hemispheric asymmetry in the processing of categorical versus coordinate auditory pitch relations. Paper presented at the Annual Meeting of the Psychonomic Society, St. Louis, MO. Wiesel, T. N., & Hubel, D. H. (1966) Laminar and columnar distribution of geniculocortical fibers in the macaque monkey. Journal of Neurophysiology, 29, 1115-1156. Williams, M. C., Breitmeyer, B. G., Lovegrove, W. J., & Guitierrez, C. (1991) Metacontrast with masks varying in spatial frequency and wavelength. Vision Research, 31 , 2017-2023.
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
125
Chapter 5
Computational Analyses and Hemispheric Asymmetries in Visual-Form Recognition Chad J. Marsolek E. Darcy Burgund University of Minnesota
Much of contemporary theory in cognitive neuroscience adheres to a theoretical framework in which behavioral abilities are understood as arising from the operations of and interactions between relatively independent processing subsystems of the brain. Although these subsystems may be only weakly modular (e.g., Farah, 1994), a primary goal for the field is to discover the broad architecture of interconnected subsystems and to pinpoint the particular functions performed by the various subsystems. In other words, the goal is to understand how the brain is "carved at its joints" (e.g., Kosslyn & Koenig, 1992) or, perhaps more appropriately given weak modularity, "stretched at its interconnections" into relatively independent functional entities. This quest for component subsystems is especially important in understanding functional hemispheric asymmetries. Theories that are cast in terms of how component subsystems operate more or less effectively in different hemispheres are providing more powerful and compelling explanations than those that rely on general principles or fundamental dichotomies (see Hellige, 1993a).
126
Marsolek and Burgund
In order to theorize about the function of any system, one must characterize how inputs are mapped to outputs. Inputs, outputs, and the couplings between the two are needed to describe any function; if one of these three elements is missing or vague, no matter which one, a function per se cannot be described. Hence, if one's goal is to delineate and understand the function of a neural processing subsystem, one must theorize not only about what it produces as output, but also about what it accepts as input and which inputs will be used to signal the production of which outputs. This somewhat obvious point helps to illuminate some of the important questions to consider when doing the difficult job of stretching the brain at its interconnections, functionally speaking. An effective strategy to follow when hypothesizing about neural processing subsystems is to consider clues about inputs, outputs, and their mappings from the perspectives of different levels of explanation (Marr, 1982). Useful computational clues come from considerations of the goals that should be satisfied by the relevant subsystems, what information is available to the relevant subsystems to help achieve these goals, and what sort of strategy would be useful for achieving the appropriate goals given the available information. Useful implementational clues include aspects of the underlying physical substrate that suggest how inputs may be represented, where the inputs come from, how outputs may be represented, where the outputs are sent to, and how the mechanism may operate to map inputs to outputs. In this chapter, we consider such clues to hypothesize about the subsystems involved in visual-form recognition. First, we theorize that two relatively independent subsystems underlie different aspects of visual-form recognition and that each subsystem operates more effectively in one cerebral hemisphere than in the other. Then, we summarize the results from behavioral studies that support the separate subsystems theory. Next, we offer computational analyses and describe results from a computational modeling study that illuminate the contradictory natures of the internal processing strategies that these subsystems may use. Finally, we summarize the results from additional behavioral studies that support the computational theory. VISUAL FORM S U B S Y S T E M S
Visual-form recognition is an essential human ability. By most accounts, it entails the activation of a previously stored visual-form
Visual Form Recognition 127 representation that corresponds best to the currently processed input form. Generally, the neural mechanisms involved in this ability appear to operate in occipital-temporal and inferior-temporal cortex of the brain (e.g., Buckner et al., 1995; Petersen & Fiez, 1993; Schacter et al., 1995; Sergent, Ohta, & MacDonald, 1992; Squire et al., 1992), in a region that may be a homologue to the occipital-temporal " w h a t " pathway in nonhuman primate vision (as opposed to the occipitalparietal "where" or "action" pathway; Felleman & Van Essen, 1991; Goodale & Milner, 1992; Haxby et al., 1991; Ungerleider & Mishkin, 1982). These areas accept retinotopically coded inputs from primary visual cortex (Fox, Miezin, Allman, Van Essen, & Raichle, 1987; Kosslyn et al., 1993; Tootell, Silverman, Switkes, & De Valois, 1982), and they appear to send output representations that signal a recognized form to non-visual subsystems (e.g., phonological, conceptual/ associative, motoric, etc.) as well as to other visual subsystems (e.g., the occipital-parietal "where" or "action" pathway).
Computational Constraints Careful considerations of our visual abilities, in terms of goals and strategies, suggest that these visual-form areas do not perform a single or simple process. Given retinotopically coded images of forms that appear in the world, visual-form subsystems accomplish at least two essential goals. First, they underlie the ability to recognize abstract categories of forms. For example, when reading a book, a reader usually categorizes word forms at only the coarse level of classification needed to access the appropriate meanings associated with the words. More concretely, one categorizes forms at the coarse level in which all of the forms in Figure 1 belong to the same category and hence produce the same output. However, visual subsystems also underlie another important ability, that of recognizing specific instances within the same abstract category of form. For example, to recognize a signature, one usually categorizes a letter string at the fine-grained level needed to distinguish a restricted set of the many possible ways in which the same letter string can appear. More concretely, one categorizes forms at the fine-grained level in which the forms in Figure 1 produce different outputs. According to many theories, these two abilities should be accomplished in a single, undifferentiated processing subsystem (for general computational theories, see Knapp & Anderson, 1984; McClelland &
128
Marsolek and Burgund
BEAR BEAR BEAR
BEAR
bear bear
bear bear
Figure 1. Different specific instances that belong to the same abstract category of visual form.
Rumelhart, 1985; for object recognition theories, see Hummel & Biederman, 1992; Tarr, 1995; see also most theories of word recognition). However, the two abilities seem to place contradictory demands on relevant processing mechanisms. To recognize abstract categories, it should be useful for a subsystem to ignore the visually distinctive information that differentiates specific instances in a category and to focus on the information that is relatively invariant across instances. For example, only the shared information between the two input forms in Figure 1 is very useful for accomplishing the goal of recognizing the common abstract category. Note that the relatively invariant information across the forms in Figure 1 is only found in parts of the larger wholes (see Fig. 2a). In contrast, to recognize specific instances, it is necessary for a subsystem to focus on just the sort of information that may be effectively ignored when recognizing abstract categories. For example, the information that distinguishes the inputs in Figure 1 must be processed to accomplish the goal of discriminating the forms. It is important to note that the visually distinctive information that differentiates the forms in Figure 1 in addition to the specific instances of other words is found in the wholes of those forms (see Fig. 2b). Implementational Constraints In addition, considerations of findings in the neuropsychological literature suggest that there are important differences in how visual-form subsystems operate across the different physical substrates of the left
Visual Form Recognition 129
Relatively In variant Information
BEAR BEAR BEAR
BEAR
bear bear
bear bear Figure 2a. The relatively invariant information (fight) that is common to different specific instances in the same abstract category of visual form (left). In the lower display, the visually distinctive information (fight) that distinguishes specific instances in the same abstract category of visual form (left).
hemisphere (LH) and fight hemisphere (RH). First, subsystems in the LH and RH play important roles in word recognition and face recognition, respectively (e.g., Damasio & Damasio, 1983; Geffen, Bradshaw, & Wallace, 1971; Petersen & Fiez, 1993; Rhodes, 1985; Sergent et al., 1992; Sergent & Bindra, 1981). Second, recognition of word, face, and object forms appears to be performed through two, not one or three, independent capacities. In a review of the visual associative agnosia literature, Farah (1990, 1991) noted that only a subset of the logically possible combinations of impaired abilities exhibited by individual brain-damaged patients are found frequently. The combinations indicate that two kinds of visual recognition capacities are susceptible to damage, one involving word and sometimes object
130
Marsolek and Burgund
Visually Distinctive Information
BEAR BEAR BEAR BEAR
z ~ ,~ .~
BEAR BEAR BEAR BEAR
bear bear
bear ~
bear bear
bear
bear bear
Figure 2b. The visually distinctive information (fight) that distinguishes specific instances in the same abstract category of visual form (left).
recognition (and affected by LH damage), and the other involving face and sometimes object recognition (and affected by RH damage). These results have direct implications for abstract-category and specificinstance recognition, to the extent that word recognition usually relies on abstract-category processing, face recognition usually relies on specific-instance processing, and object recognition likely relies on different processes (abstract versus specific) in different circumstances.
Relatively Independent Visual-Form Subsystems Although these considerations do not rule out the possibility that a single subsystem underlies both abilities, they do lead to the theory that relatively independent subsystems underlie visual-form recognition. We
Visual Form Recognition 131 have hypothesized that an abstract visual-form (AVF) subsystem underlies recognition of abstract categories of forms, processes relatively-invariant input information through the use of a parts-based internal processing strategy, and operates more effectively in the LH than in the RH. In contrast, a specific visual-form (SVF) subsystem underlies recognition of specific instances of forms, processes visuallydistinctive input information through the use of a holistic internal processing strategy, and operates more effectively in the RH than in the LH. These subsystems may operate relatively independently in large part because they rely on contradictory internal processing strategies. In the next section, we summarize initial behavioral evidence in support of this theory. Afterward, we use computational analyses and modeling to suggest a concrete specification of these internal processing strategies and to illuminate their contradictory natures. BEHAVIORAL EVIDENCE FOR RELATIVELY INDEPENDENT SUBSYSTEMS
We have tested the AVF/SVF subsystems theory in the following studies. Initial behavioral tests utilized divided-visual-field presentations of visual forms during the test phases of various memory experiments. Rationale for Divided-Visual-Field Studies
We use the divided-visual-field technique as a tool to help test whether a single visual-form subsystem operates in a fairly unitary manner in the brain or two subsystems operate in a relatively independent manner. Our rationale is similar to the rationale that interactions between task and field of visual presentation are needed to adequately study laterality effects in this paradigm (Hellige, 1983). Of course, in a divided-visual-field presentation, the information presented directly to one hemisphere must cross brain commissures to be processed by the other. The important implication is that mechanisms in the first hemisphere obtain higher quality information (e.g., Dimond, Gibson, & Gazzaniga, 1972; Gross, Rocha-Miranda, & Bender, 1972) and obtain it more quickly than mechanisms in the other hemisphere. Thus, if the characteristic processing of one hypothesized subsystem (e.g., AVF processing) is performed more effectively when high quality visual input is processed initially in one hemisphere (e.g., left) than in
132
Marsolek and Burgund
the other, whereas the characteristic processing of a different hypothesized subsystem (e.g., SVF processing) is performed more effectively when high quality visual input is processed initially in the other hemisphere (e.g., fight) compared with the one yielding the first advantage, then two sorts of processing must rely on two sets of neural circuitry that operate at least relatively independently. Note that, taken alone, such behavioral results do not suffice to indicate the extent to which the subsystems are lateralized or the degree to which they are modular. They only indicate that at least relatively independent subsystems are involved and that they are at least weakly lateralized. Further investigations utilizing additional methodologies are needed to clarify the unresolved issues. For example, further methods are needed to determine whether one subsystem operates in only one hemisphere, whether both subsystems operate with asymmetric efficiency in each hemisphere, etc.
Visual Repetition Priming In one study, we examined repetition priming for visual word forms (Marsolek, Kosslyn, & Squire, 1992). Subjects were asked to read lists of common words presented in the central visual field during an initial encoding phase. Half of the words were presented in all lowercase letters (e.g., "convince"), and half were presented in all uppercase letters (e.g., "PRIMARY"). Afterward, in a presumably unrelated second phase of the experiment, subjects completed word stems (three-letter beginnings of words that can be completed to form many common words; e.g., "con") to form the first words that came to mind. Each stem was presented in the left or right visual field. Repetition priming was exhibited when they produced word completions that corresponded to words that were viewed earlier in the experiment with a greater-thanchance tendency. The results were that such priming was greater when stems were presented in the same letter case as previously presented words, compared with the different letter case. More important, this letter-case-specific priming effect was found when the stems were presented directly to the RH (briefly in the left visual field) but not when they were presented directly to the LH (briefly in the right visual field). In a related study, Marsolek, Squire, Kosslyn, and Lulenski (1994) discovered that, in certain experimental conditions, the same pattern of
Visual Form Recognition 133 results is obtained when subjects use the stems as cues to help them explicitly recall previously seen words. These results suggest that the structural changes that underlie visual memory effects may be qualitatively different across subsystems that operate asymmetrically in the two hemispheres. Structural changes that underlie storage of the visually distinctive information that differentiates specific instances in an abstract category of form (e.g., lower- vs. uppercase versions of the same word) are instantiated more effectively in the RH than in the LH. Given that this sort of information storage should be characteristic of an SVF subsystem, the results indicate that an SVF subsystem, but not an AVF subsystem, operates more effectively in the RH than in the LH.
Task Demands in Visual Repetition Priming In a similar study, we examined how task demands may influence priming in visual-form subsystems (Burgund & Marsolek, 1996). All subjects read lists of common words and pronounceable nonwords (half of each in all lowercase letters, and half of each in uppercase) that were intermixed and presented in the central visual field during initial encoding. Then, in the test phase of one experiment, subjects performed a standard perceptual identification task by identifying briefly presented letter strings and writing them down, without following any particular instructions on how to write the strings. In the test phase of a different experiment, subjects performed a form-specific perceptual identification task by identifying briefly presented letter strings and writing them down in the s a m e l e t t e r c a s e as they had appeared on the computer monitor. Repetition priming was measured as the tendency to identify and report letter strings that had been viewed earlier in the experiment more accurately than letter stings that had not been processed previously. The results were that letter-case-specific priming was greater when test items were presented directly to the RH compared to the LH (as observed in word-stem completion studies; Marsolek et al., 1992, 1994), but this was true only when the form-specific perceptual identification task was performed and not when the standard perceptual identification task was performed (cf. Koivisto, 1995). These results suggest that task demands directly affect which subsystems are recruited in different priming tests. Performance in the standard perceptual identification task may be influenced by processing
134
Marsolek and Burgund
in an AVF subsystem to a greater degree than by processing in other subsystems, because letter-case-specific information does not need to be processed for accurate performance (unlike in the form-specific perceptual identification task). In addition, there is only one correct response per trial in this task, which may have the effect that the most efficient subsystem for the job (i.e., an AVF subsystem) dominates the production of a response (unlike in the word-stem completion task, in which the Iarge number of "correct" responses per stem may have the effect that various subsystems contribute to the production of a response). Of course, performance in the form-specific perceptual identification task should be influenced highly by an SVF subsystem, because letter-case-specific information must be processed for accurate performance. In light of these task demands, the results further support the AVF/SVF subsystems theory. Visual Classification
In another experiment, we investigated classification of novel visual forms (Marsolek, 1995). Subjects first learned to associate labels to categories of unfamiliar letterlike forms, and each form was presented in the central visual field during learning. Afterward, subjects were asked to classify test forms using the newly-learned categories. The results were that they classified the previously unseen prototypes of the newly learned categories (each prototype was the central tendency of the instances in one category) more effectively when they were presented directly to the LH than to the RH. In contrast, subjects classified the previously seen specific instances more effectively when they were presented directly to the RH than to the LH (and they did not classify previously unseen non-prototype instances differently depending on hemisphere of presentation). Generally, the prototypes of these visual-form categories possessed a large amount of the visual information that remained relatively invariant across the different instances in one category. In fact, each prototype contained a larger amount of this relatively invariant information than did any of the other specific instances in its category. Hence, the results indicate that information that is useful for an AVF subsystem (relatively invariant information) is stored more effectively in the LH than in the RH. In contrast, the previously seen specific instances contained visually distinctive information. Hence, the results indicate that the information
Visual Form Recognition 13 5 that is useful for an SVF subsystem is stored more effectively in the RH than in the LH. Note that a RH advantage for processing the previously seen specific instances was obtained even though the demands of the categorization test task may have favored processing in an AVF subsystem more than processing in an SVF subsystem. We suggest that an SVF subsystem contributed to performance in this task nonetheless, because all of the forms were unfamiliar preexperimentally, and an SVF subsystem should store novel visual forms especially well. This reasoning for this hypothesis (see below) stems from our analyses of the internal processing strategies for AVF and SVF subsystems. CONTRADICTORY INTERNAL PROCESSING STRATEGIES
Many previous theories of functional asymmetries have emphasized distinctions that can be understood in terms of analytic/parts-based processing versus holistic processing in the left and right hemispheres, respectively (e.g., Bever, 1980; Bradshaw & Nettleton, 1981; Corballis, 1989; Diamond & Carey, 1986; Farah, 1990, 1991; Levine & Calvanio, 1989). Unfortunately, little attention has been given to explicating precisely the differences between these two kinds of processes (see Marshall, 1981, for important problems with analytic vs. holistic distinctions). In addition, little attention has been given to clarifying the conditions under which each kind of processing should be recruited or to explaining why the two kinds of processing appear to operate in different parts of the brain. In this section, we offer computational analyses suggesting that one visual-form subsystem (an AVF subsystem) performs parts-based processing, because that sort of processing is required for effective recognition of abstract categories per se. Another visual-form subsystem (an SVF subsystem) performs holistic processing, because that sort of processing is required for effective recognition of specific instances per se. This analysis illuminates a concrete specification of the distinction between--and the contradictory natures of--parts-based versus holistic internal processing. In addition, we discuss how results from a neural network modeling study supplement and directly test this computational reasoning.
136
Marsolek and Burgund
Neural Network Models of Visual-Form Recognition
We have examined relatively simple feedforward neural network models that receive "retinotopically-coded" bit-mapped inputs, in an effort to investigate how parallel distributed processing systems might accomplish the task of recognizing such input forms (Marsolek, 1994). Each network was composed of three layers of processing units: an input layer, a hidden layer, and an output layer. Input units were connected to hidden units via weighted interconnections, and in turn hidden units were connected to output units via weighted interconnections. In each processing trial, an input form was presented to the network as a pattern of activation across the two-dimensional array of input units. These input forms were just smaller than the array of input units, and in one trial an input form was presented in either the upper-right, lower-right, lower-left, or upper-left region of the array. After an input form was presented, activation flowed across the first set of weighted connections to activate the hidden units, and in turn activation flowed across the second set of weighted connections to activate the output units. Weights on the internal connections modulated the flow of activation between layers. Each network was trained to perform input-to-output mappings through the use of an error-correction procedure (backpropagation-oferror; Rumelhart, Hinton, & Williams, 1986). This training was used to guide each network to discover a set of weights across its internal connections that allowed the intended input-output mappings to take place. When such networks are trained to accomplish input-output mappings that apparently take place in certain areas of visual cortex, they tend to discover mapping solutions that the relevant neural subsystems appear to use (e.g., Churchland & Sejnowski, 1992; Lehky & Sejnowski, 1988; O'Reilly, Kosslyn, Marsolek, & Chabris, 1990; Zipser & Andersen, 1988). Note that any neural implausibility of the specific backpropagation training process used in this study was not important, in part because networks that use training algorithms that are very similar to backpropagation, yet are biologically plausible, tend to produce results that are highly similar to those found with backpropagation (see, e.g., Mazzoni, Andersen, & Jordan, 1991). The input patterns and the target output patterns in our networks were random-dot patterns that adhered to statistical similarity constraints. However, in the following discussion we describe the models
Visual Form Recognition 13 7 as though word forms like "BEAR," "bear," "bear," served as input patterns, simply for clarity.
"bear," etc.,
AVF, SVF, and Intermediate Input-Output Mappings Each network was trained to solve two of the following three inputto-output mapping problems, with the three mapping problems examined in different pairs across different networks. In AVF mappings (see Figure 3), the uncorrelated input patterns "BEAR" and "bear" (note the visual dissimilarities between these forms), in addition to the correlated input patterns "bear" and " b e a r " (note the visual similarities between these forms), were all mapped to the same output representation. However, in SVF mappings (see Figure 4), each specific instance (e.g., "BEAR," "bear," "bear," and "bear," e t c . ) w a s mapped to a different output representation. Furthermore, in an intermediate-level mapping problem (see Figure 5), the correlated input patterns "BEAR" and "BEAR" were mapped to the same output representation, yet the input patterns "bear" and "bear," which were correlated with each other but not with "BEAR" and "BEAR," were mapped to a different output representation. Note that the input patterns were the same for all three mapping problems; only the mappings to output representations differed across the three mappings. We examined whether the internal architectures of the networks affected their abilities to perform different pairs of mapping problems. One set of networks was trained to perform both AVF and SVF mappings of input patterns through different subsets of output units, half of the output units allocated to the AVF task and the other half to the SVF task. Some of these networks were unified or "unsplit" models, in which all hidden units were connected to all output units. However, other networks were "split" (e.g., Rueckl, Cave, & Kosslyn, 1989), in that one subset of the hidden units was connected only to the AVF output units and the other subset was connected only to the SVF output units. In this way, split networks had separate subcomponents devoted to the different mapping tasks, but unsplit networks accomplished the two tasks through a unified model. Results were that, after an arbitrary number of training trials, models with separate subcomponents performed the AVF and SVF mappings more efficiently than unified models. (These split networks also
138
Marsolek and Burgund
A VF Mappings
BEAR BEAR BEAR BEAR Output 1
bear bear
bear bear GANG G/klqG GANG
GANG ggang
Output 2
ang
gang Figure 3. AVF mappings of input forms (left) to output representations (fight) simulated in neural network models.
Visual Form Recognition 139
S VF Mappings
BEAR BEAR BEAR BEAR
,
# ~ ~ ~
Output Output Output Output
-~
Output 5
~ .~
Output 6
-9 ~
Output 7 Output 8
GANG
)
Output 9
GANG
~
Output 10
,,,
bear bear bear
,
bear
GANG
,,,
.~
1 2 3 4
Output 11
GANG
,-,._ --
Output 12
gang gang
~
Output 13
gang
_~ Output 15
Output 14
Figure 4. SVF mappings of input forms (left) to output representations (right) simulated in neural network models.
140
Marsolek and Burgund
Intermediate-Level Mappings
BEAR BEAR BEAR BEAR
Output 1
bear bear
Output 2
bear bear GANG GANG GANG
Output 3
~G an
Output 4
gang Figure 5. Intermediate-level mappings of input forms (left) to output representations (fight) simulated in neural network models.
Visual Form Recognition 141 outperformed networks with the same number of hidden-to-output unit connections as the split networks but with no systematic splitting of the hidden units into separate pools allocated to the different mappings.) It is important to note that this separate-subnetwork advantage did not extend to models that performed both AVF and intermediate-level mappings. Furthermore, the separate-subnetwork advantage did not extend to models that performed both SVF and intermediate-level mappings (see also Knapp & Anderson, 1984; McClelland & Rumelhart, 1985). Hence, split networks did not always outperform unsplit networks; they did so in this study only when AVF and SVF mappings, in particular, were performed in the same networks. The split networks described above were strongly split networks in that they contained a strong modularity in their internal architectures. Strong modularity was not necessary, however, for the separatesubnetwork advantage. AVF and SVF networks that were weakly split (such that one subset of hidden units was dedicated to the AVF task, another to the SVF task, and a third to both tasks) also outperformed unsplit networks, even though weakly-split AVF and intermediate networks and weakly split SVF and intermediate networks did not outperform their unsplit counterparts. The idea that SVF and intermediate mappings are compatible with one another in this study (Marsolek, 1994), as well as in other computational studies (see Knapp & Anderson, 1984; McClelland & Rumelhart, 1985), is one aspect of the present theory and models that is different from Kosslyn's (1987; Kosslyn, Chabris, Marsolek, & Koenig, 1992) theoretical distinction between coordinate and categorical spatial relations encoding and from Jacobs and Kosslyn's (1994) neural network models of "coordinate shape" and "categorical shape" processing. Their coordinate-categorical mapping distinction is very similar to our SVF-intermediate mapping distinction, in that one of the tasks (SVF or coordinate) involves mapping visually similar inputs to different output representations and the other task (intermediate or categorical) involves mapping visually similar inputs, and only inputs that are similar to them, to the same output representation. Yet, we find that SVF and intermediate mappings are compatible, whereas Jacobs and Kosslyn (1994) conclude that different mapping solutions are useful for coordinate-shape and categorical-shape processing in their neural network models.
142
Marsolek and Burgund
The apparently inconsistent results across modeling studies may be resolved with the help of additional analyses (Marsolek, 1994). I n our study, projective fields of hidden units in all split networks were held to a constant size across the pools of AVF, SVF, and intermediate-level hidden units. The projective field of a hidden unit is determined by its connections to the output units (Lehky & Sejnowski, 1988), thus split networks with larger numbers of output units devoted to the SVF mapping than to the intermediate mapping would have SVF hidden units with larger projective fields than those of intermediate-mapping hidden units. This is important because methodological difficulties arise when the two subsets of hidden units in a split network do not have the same-sized projective fields, which was the procedure used by Jacobs and Kosslyn (1994). When SVF hidden units have larger projective fields than intermediate-level hidden units, we found that split networks outperform their unsplit counterparts, in restricted conditions, but the result is equivocal (Marsolek, 1994). The split advantage in this situation could be due to facilitation from splitting networks into large and small projective-field subnetworks (which may have to do with optimal learning rates being proportional to fan-in and fan-out of processing units in networks like these; cf. Plaut & Hinton, 1987), regardless of whether contradictory mapping tasks are performed in the separate subnetworks. Therefore, we conclude that AVF and SVF mappings, but not the other pairs of mappings, are performed more efficiently through separate subnetworks than through unified networks. These results indicate that the internal processing strategies that are useful for parallel distributed processing systems to perform AVF mappings and the internal processing strategies that are useful for such systems to perform SVF mappings may interfere with one another or are contradictory in some way. In what ways are the strategies contradictory? Our computational argument is that different internal processing strategies are useful for subsystems that perform AVF and SVF recognition. Partsbased processing should be useful for an AVF subsystem because the relatively invariant information of the forms in one category tend to be found in their parts; in contrast, holistic processing should be useful for an SVF subsystem because the visually distinctive information that differentiates specific instances of forms is found in the holistic structures of the forms (see Figure 2). A single, undifferentiated mechanism cannot perform both parts-based and holistic processing effectively. For proper assessment, this hypothesis requires the following explication.
Visual Form Recognition 143
Assumptions Two important assumptions are needed for the computational analysis offered here. First, visual-form subsystems (and artificial networks that simulate them) receive the retinotopically-mapped representations of input forms that are currently being captured by selective attention mechanisms. That is, visual selective attention serves to filter out extraneous inputs and acts to "surround" the form that currently is being processed, regardless of the location or size of the form as it appears on the retina (e.g., LaBerge, 1995). Such a selectiveattention filter may operate at least in part before processing in visualform subsystems of the inferior temporal cortex, because cells in the inferior temporal region do not change their response selectivities appreciably when visual forms change locations and sizes yet remain within the receptive fields of the cells (Schwartz, Desimone, Albright, & Gross, 1983). This assumption is needed to account for effects of size and location invariance in visual-form priming; changes in size and location of forms between initial encoding and subsequent test do not greatly influence priming effects (Biederman & Cooper, 1992; Cooper, Schacter, Ballesteros, & Moore, 1992). The analogy in the network models is that input forms are conceptualized to be scaled so that they are just smaller than the input grid in any one trial, regardless of their sizes or locations on the "retina" in some earlier stage of processing. The second important assumption is that visual-form priming is produced by structural changes in visual-form subsystems (and in artificial networks that simulate them). This assumption is needed to account for relatively long-term priming effects, which can last for several days in the case of object priming, even for amnesic patients (Cave & Squire, 1992). Long-term priming must be supported through some kind of physical change that serves to store information for a period of time (possibly through local synaptic changes such as in longterm potentiation in visual cortex; Artola & Singer, 1987; Komatsu, Fujii, Maeda, Sakaguchi, & Toyama, 1988). The analogy in the network models is that priming is supported by small changes in the previously established weights on the internal connections of the networks due to recent processing of a prime stimulus (and the accompanying backpropagation-of-error). Different kinds of weight changes can be considered in these network models, depending on the kind of information they represent.
144
Marsolek and Burgund
Parts-Based Internal Processing Imagine the receptive fields that would develop in these network models during training (the receptive field of a hidden unit is determined by its connections to the input units). If each hidden unit in a network were to have a relatively large absolute weight value on only a few of the connections feeding into it and very small weight values (perhaps 0) on the other connections, then each hidden unit would be connected functionally to only a few input units. If different hidden units were sensitive to different subsections of the input array, such that any one hidden unit were to be sensitive to only a part of any one input form, this would be an example of a network with a parts-based internal representation strategy. Such a network may utilize what is called a "sparse" or "local" coding strategy (see Churchland & Sejnowski, 1992). In this case, different hidden units necessarily represent different parts of any one input form. Note that this kind of internal processing strategy should be useful for an AVF subsystem. For example, an efficient internal representation for the abstract category in Figure 1 may be activation of the hidden units that are sensitive to that category's relatively invariant information (see Figure 2), coupled with little or no activation of the hidden units that are sensitive to the parts that are found in the visually distinctive information that differentiates specific instances in that abstract category or to the parts that are found in other abstract categories. Indeed, examinations of the trained networks described above (Marsolek, 1994) indicate that this sort of internal processing strategy is used in the AVF portions of split networks. The hidden units in these AVF subnetworks develop receptive fields that utilize a relatively parts-based strategy. The priming in such a network necessarily would be parts-based. Because the activation of any one hidden unit represents only one part of an input form, small weight changes on the connections feeding into different hidden units would yield priming for relatively independent information about differe~ parts of an input form. Hence, different parts of the same input form should be primed relatively independently in such a subsystem. Indeed, behavioral results summarized below support this hypothesis.
Visual Form Recognition 145
Holistic Internal Processing Now imagine another possibility in these network models. If each hidden unit in a network were to have relatively large (but varying across units) absolute weight values on perhaps all of the connections feeding into it, then each hidden unit would be connected functionally to the whole (or at least a very large portion) of the input array. If different units were differentially sensitive to different parts of the whole input array, such that different hidden units were sensitive to slightly different aspects of the whole of any one input form, this would be an example of a network with a holistic internal representation strategy. (For advantages in using "coarsely" coded representations like these, see Ballard, 1986; Hinton, McClelland, & Rumelhart, 1986.) In this case, the hidden units would not represent different parts of an input form explicitly as such, only implicitly as portions of the whole form to which perhaps every unit is sensitive. This kind of internal processing strategy should be useful for an SVF subsystem. For example, an efficient internal representation for the specific instance "BEAR" may be a distinct pattern of activation across perhaps all of the hidden units, each of which may be slightly differentially sensitive to all of the information in that particular input form. Indeed, examinations of the trained networks described above (Marsolek, 1994) indicate that this sort of internal processing strategy is used in the SVF portions of split networks. The hidden units in these SVF subnetworks develop receptive fields that utilize a relatively holistic strategy. The priming supported by such a network would b e holistic. Because the activation of any one hidden unit represents holistic structure, small weight changes on the connections feeding into any one hidden unit would yield priming for the wholes of forms. Hence, different parts ~of the same input f o r m could n o t be primed independently in such a subsystem. Indeed, behavioral results summarized below support this hypothesis. In addition, it is worthwhile to note t h a t the holistic priming supported by such a network should be useful for priming of unfamiliar forms. At some level, every unfamiliar form has parts (in the limit, various sorts of edges) that are familiar; it is the holistic structure of any unfamiliar form that contains the information that makes it unfamiliar. Hence, the novel aspects of unfamiliar forms should be primed more
146
Marsolek and Burgund
effectively in a holistic processing network than in a parts-based processing network. This hypothesis is also supported by behavioral results summarized below.
Contradictory Strategies It should be clear that a single, undifferentiated network could not implement both a parts-based and a holistic processing strategy. The same mechanism could not represent parts explicitly as such and n o t represent parts explicitly as such. Although the parts-based and holistic processing strategies described above are relatively extreme versions of both, less extreme versions of the two strategies also may be contradictory. The relatively invariant information associated with an abstract category like in Figure 1 is present in some subset ~of the information in any one input form, a subset that necessarily is smaller than the amount of visually distinctive information that differentiates any one specific instance in that category (i.e., the holistic structure of that instance). Likewise, the visually distinctive information that differentiates specific instances in Figure 1 is present in the holistic information in these inputs, a set of information that necessarily is larger than the relatively invariant information for that abstract category (i.e., parts of the forms). Hence, efficient internal representations for AVF categorizations should involve parts-based coding of a relatively small amount of information per form, whereas efficient internal representations for SVF categorizations should involve holistic coding of a relatively large amount of information per form. It would be difficult for a single mechanism to store both kinds of information effectively (only the relatively invariant information for the abstract category as well as the visually distinctive information that differentiates specific instances in that category). In other words, it would be difficult for a single mechanism to accomplish both AVF and SVF recognition effectively. Given that contradictory strategies (parts-based vs. holistic) are useful for accomplishing different goals (AVF vs. SVF categorizations), and both goals are vitally important for survival, selective pressures may have led to the evolution of relatively independent processing subsystems with the different subsystems utilizing different computational strategies. Note that the claim here is not that it would be impossible for a single subsystem to accomplish both AVF/parts-based
Visual Form Recognition 147 and SVF/holistic processes or that these processes are incompatible per se. Instead, the reasoning involves considerations of relative efficiency and of course should be tested. Next, we summarize behavioral evidence in support of our computational theory. BEHAVIORAL EVIDENCE FOR PARTS-BASED VERSUS HOLISTIC PROCESSING In further behavioral studies, we have tested directly whether an AVF subsystem utilizes a parts-based internal processing strategy whereas an SVF subsystem utilizes a holistic processing strategy. Recent behavioral studies using divided-visual-field presentations of memory test items and studies using an interhemispheric communication paradigm support these hypotheses. Visual Repetition Priming The computational theory is supported by a recent study of visual priming using the word-stem completion task (Marsolek et al., 1995). In this study, some of the word stems (and hence the beginnings of their corresponding words) were composed of letters with visually dissimilar lower- and uppercase structures (e.g., "bea" / "BEA"), whereas the other items were composed of letters with visually similar lower- and uppercase structures (e.g., "sco" / "SCO"). Like the priming studies described above (Marsolek et al., 1992; 1994), subjects read lists of words presented in the central visual field during an initial encoding phase. Half of the words were presented in all lowercase letters, and half were presented in all uppercase letters. Then, subjects completed word stems to form the first words that came to mind during a presumably unrelated second phase of the experiment. Results indicate that the parts-based information that is common to "BEAR" and "bear" is stored in an AVF subsystem that operates more effectively in the LH than in the RH. Same and different letter-case priming did not differ when the dissimilar stems were presented directly to the LH (even though same-case priming was greater than differentcase priming when these stems were presented directly to the RH). For example, "BEAR" primed "bea" as well as "bear" primed "bea" in LH stem presentations. (Note that this effect apparently is supported by visual subsystems per se, because "BEAR" primes "bear" to a greater
148
Marsolek and Burgund
degree than hearing the word bear primes the visual form "bear" [Bowers, 1996].) In contrast, the holistic information that differentiates even structurally similar forms, like "SCOOP" and "scoop," is stored in an SVF subsystem that operates more effectively in the RH than in the LH, albeit through interactions with the hippocampal formation (see several chapters in Schacter & Tulving, 1994; Cohen & Eichenbaum, 1993; McClelland, McNaughton, & O'Reilly, 1995; Squire, 1992). This qualification is needed because only explicit memory as measured in word-stem cued recall, but not repetition priming as measured in wordstem completion priming, produces greater same-case than differentcase memory for the similar-case items in RH presentations. For example, when "scoop" had been presented earlier in the experiment and "sco" was the test cue, that word was recalled more readily than when "SCOOP" had been presented earlier and "sco" was the test cue, in RH but not in LH presentations of the cues. Visual Priming for New Associations
In further experiments, we examined visual priming for new associations between previously unrelated words (Marsolek, Schacter, & Nicholas, 1996). During the encoding phase, subjects read lists of word pairs, one word presented above the other in the central visual field for each pair. Half of the pairs were presented in all lowercase letters, and half were presented in all uppercase letters. Afterward, subjects completed word stems that were presented beneath complete context words in a presumably unrelated test phase. Letter-case-specific priming in stem completion was found only when the context words were the same words that had appeared previously above the primed completion words during initial encoding and when the two items in a test pair (context word and word stem) were presented directly to the RH. When the context words were different words from those which had appeared above the primed completion words during initial encoding, no lettercase-specific priming was obtained in stem completion. These results indicate that priming for novel holistic information (i.e., one word form as it appears above another word form) is supported by a subsystem that distinguishes lower- versus uppercase versions of the same word and operates more effectively in the RH than in the LH, as predicted from the computational theory described above.
Visual Form Recognition 149
Interhemispheric Communication of Visual-Form Information In another behavioral study, we examined interhemispheric transfer of visual-form information (Nicholas & Marsolek, 1996). Subjects were asked to compare two visually presented letters in each trial. The two letters appeared in the same visual field in half of the trials, but in different visual fields in the other half of the trials, similar to a task used by Banich and Belger (1990) who presented three letters per trial and asked subjects to compare the bottom-most letter to the other two in each trial. With such a procedure, interhemispheric transfer of information m u s t have taken place when the comparison letters were presented across hemispheres (briefly displayed in different visual fields) but not necessarily when they were presented within a hemisphere (briefly displayed in the same visual field). We took advantage of this circumstance in experiments investigating two visual comparison tasks. Results indicate that AVF and SVF subsystems are differentially affected by interhemispheric transfer of visual-form information in ways that are predictable from the computational theory described above. In an AVF comparison task, subjects decided whether the pairs corresponded to the same letter of the alphabet (e.g., "a" and "A," "s" and "S") or not (e.g., "a" and "Q", "s" and "P"). In this task, they performed more accurately in across-hemisphere trials than in within-hemisphere trials. However, this across-hemisphere advantage was found for similar-case letters (e.g., s/S), but not for dissimilar-case letters (e.g., a/A) which yielded no within- or across-hemisphere advantage. Current theories of interhemispheric communication (Banich & Belger, 1990; Belger & Banich, 1992; Hellige, 1993b) would not predict this finding, yet the AVF/SVF subsystems theory may account for it. The effects of noise produced by interhemispheric transfer of similar-case letters may not be as detrimental as the effects of noise produced by interhemispheric transfer of dissimilar-case letters, in the AVF task. That is, noisy versions of the relatively invariant information in similar-case letters may be processed by an AVF subsystem more effectively than noisy versions of the relatively invariant information in dissimilar-case letters, because there is more relatively invariant information per letter for similar-case items than for dissimilar-case items to help overcome the noise. Hence, the relatively invariant information needed to make AVF categorizations of similar-case letters, but not dissimilar-case letters, may cross brain commissures effectively enough to take advantage of
150
Marsolek and Burgund
the benefits of distributing the processing of the two letters across the two hemispheres (cf. Banich, 1995; Dimond & Beaumont, 1971). In an SVF comparison task, subjects decided whether the letters in a pair were physically the same (e.g., "a" and "a," "S" and "S") or not (e.g., "a" and "A," "s" and "S"). In this task, they performed more accurately in within-hemisphere trials than in across-hemisphere trials. Hence, SVF processing apparently cannot take advantage of the benefits of distributing the processing of the two letters across the two hemispheres. Interestingly, this within-hemisphere advantage was found for similar-case letters (e.g., s/S), but not for dissimilar-case letters (e.g., a/A) which yielded no within- or across-hemisphere advantage. Current theories of interhemispheric communication (Banich & Belger, 1990; Belger & Banich, 1992; Hellige, 1993b) would not predict this result, but the AVF/SVF subsystems theory would. The visually distinctive information needed to make SVF categorizations of dissimilar-case letters may not cross brain commissures (to take advantage of neural distribution) effectively enough to produce an across-hemisphere advantage. Moreover, the visually distinctive information needed to make SVF categorizations of similar-case letters may be so fine-grained that it crosses brain commissures so ineffectively that a withinhemisphere advantage is produced. Indeed, by hypothesis, an SVF subsystem should be fairly sensitive to the noise produced by interhemispheric transfer of information, because it may treat noise as the kind of "visually distinctive" information that it is tuned to process when it is called upon to perform SVF categorizations. CONCLUSIONS AND IMPLICATIONS A fundamental aspect of the architecture of our neural processing subsystems appears to be that relatively independent subsystems underlie AVF versus SVF recognition. In this chapter, we have summarized research that supports this conclusion through an integration of behavioral studies and computational analyses and models. We have highlighted considerations of the inputs available to visual processing subsystems, the goals that these subsystems must satisfy to underlie important visual abilities, and plausible internal processing strategies that these subsystems could use to achieve the appropriate goals given the available input. All of these considerations are combined to characterize explicitly the functions--and contradictory
Visual Form Recognition 151 natures of--relatively independent AVF and SVF subsystems. We conclude that an AVF subsystem operates more effectively in the LH than in the RH, and it uses a parts-based internal processing strategy to focus on the relatively invariant information in visual-form inputs, which allows it to recognize abstract categories of visual forms effectively. In contrast, an SVF subsystem operates more effectively in the RH than in the LH, and it uses a holistic internal processing strategy to capture the visually distinctive information in visual-form inputs, which allows it to recognize specific instances of forms effectively. One set of results from previous research that we have not mentioned heretofore is consistent with our conclusions and may help to illuminate an important feature of our theoretical approach. In divided-visual-field studies, subjects identify and discriminate high spatial-frequency information more effectively when it is presented directly to the LH than to the RH, whereas they identify and discriminate low spatial-frequency information more effectively when it is presented directly to the RH than to the LH (e.g., Christman, Kitterle, & Hellige, 1991; Kitterle, Christman, & Hellige, 1990; Kitterle & Selig, 1991). Of course, these asymmetries may help to explain why LH advantages are found when subjects process the local parts of hierarchically arranged stimuli whereas RH advantages are found when they process the global forms of these stimuli (Robertson & Lamb, 1991; Van Kleeck, 1989), given that relatively high spatial-frequency information should be useful for processing the local forms and relatively low spatial-frequency information should be useful for processing the global forms (Sergent, 1982). These findings are consistent with our theory for the following reasons. A subsystem that has evolved to perform AVF recognition should perform parts-based processing effectively. An individual cell in such a subsystem should receive information from a relatively small portion of the input array, and hence the cell optimally should be tuned to a high spatial-frequency band. It is important to note that the cell also may be sensitive to even higher frequencies in some conditions, but it may not be very sensitive to substantially lower frequencies (e.g., Sergent, 1989). It is also important to note that a collection of such cells, each receiving information from different but overlapping portions of the input, should be very sensitive to even higher frequencies than an individual cell's optimal frequency, through the use of a distributed representation (of. Hinton et al., 1986). Furthermore, this collection of
152
Marsolek and Burgund
cells may be able to respond to relatively low frequencies as well, given that different cells in the distributed representation should respond differently due to the different levels of overall activation that are present in their different portions of a low spatial-frequency input. In contrast, a subsystem that has evolved to perform SVF recognition should perform holistic processing effectively. An individual cell in such a subsystem should receive information from a relatively large portion of the input array, and hence it optimally should be tuned to a low spatial-frequency band. Of course, by the same analysis as above, the cell also may be sensitive to higher frequencies in some conditions, but not to substantially lower frequencies (e.g., Sergent, 1989). Furthermore, a collection of such cells, each receiving slightly different information from large portions of the input, should be sensitive to higher frequency information than that to which an individual cell in the subsystem is sensitive (cf. Hinton et al., 1986), and the collection of cells should be sensitive to lower frequencies as well. Thus, both subsystems may process high and low spatial-frequency information, but not with the same efficacy. Assuming that the optimal frequency band for most individual cells in a subsystem determines the optimal range of frequencies for distributed representations in that collection of cells, and assuming roughly the same number of cells per subsystem, we conclude the following. Compared against one another, an AVF subsystem should process relatively high spatial-frequency information more effectively than an SVF subsystem, and an SVF subsystem should process relatively low spatial-frequency information more effectively than an AVF subsystem. Of course, the more general spatial-frequency hypothesis can account for these hemispheric asymmetries as well. The LH may be specialized for processing high spatial-frequency information whereas the RH may be specialized for processing low spatial-frequency information (Sergent, 1982). This hypothesis certainly has been useful for attracting attention to input factors and their roles in hemispheric asymmetries. However, concentration on input factors without their relations to other factors leaves an incomplete picture of hemispheric asymmetries. Just as one must consider how asymmetries in lower-level processes modulate asymmetries in higher-level processes when hypothesizing about the higher-level processes, one must also consider the goals that the lower-level processes accomplish, as well as the mappings between
Visual Form Recognition 153 inputs and outputs that they achieve, in order to explicate the functions of the neural subsystems that underlie the lower-level processes. The functions of relatively independent AVF and SVF subsystems may provide an explanation for the observed spatial-frequency asymmetries, one that is capable of explaining why these asymmetries evolved in the first place, but only after inputs, outputs, and their couplings are examined together. Indeed, consideration of all of the available constraints may be essential when doing the very difficult job of stretching the brain at its interconnections.
Acknowledgments Preparation of this chapter was supported by the National Institute of Mental Health, Grant MH53959-01; by the McDonnell-Pew Cognitive Neuroscience Center and the Arizona Cognitive Science Program of the University of Arizona; and by the Center for Research in Learning, Perception, & Cognition in conjunction with the National Science Foundation (GER 9454163), the Office of the Vice President for Research, and Dean of the Graduate School of the University of Minnesota. Correspondence may be sent to C. J. Marsolek, Dept. of Psychology, University of Minnesota, 75 East River Road, Minneapolis, MN, 55455. Electronic mail may be sent to:
[email protected].
References Artola, A., & Singer, W. (1987). Long-term potentiation and NMDA receptors in rat visual cortex. Nature, 330, 649-652. Ballard, D. H. (1986). Cortical connections and parallel processing: Structure and function. Behavioral and Brain Sciences, 9, 67-120. Banich, M. T. (1995). Interhemispheric processing: Theoretical considerations and empirical approaches. In R. J. Davidson & K. Hugdahl (Eds.), Brain asymmetry. Cambridge, MA: MIT Press. Banicli, M. T., & Belger, A. (1990). Interhemispheric interaction: How do the hemispheres divide and conquer a task.9 Cortex, 26, 77-94. Belger, A., & Banich, M, T. (1992). Interhemispheric interaction affected by computational complexity. Neuropsychologia, 30, 923929. Bever, T. G. (1980). Broca and Lashley were fight: Cerebral dominance is an accident of growth. In D. Kaplan & N. Chomsky (Eds.), Biology and language (pp. 186-230). Cambridge, MA: MITPress.
154
Marsolek and Burgund
Biederman, I., & Cooper, E. E. (1992). Size invariance in visual object priming. Journal Of Experimental Psychology: Human Perception and Performance, 18, 121-133. Bowers, J. S. (1996). Different perceptual codes support priming for words and pseudowords: Was Morton right all along.9- Journ'al of
Experimental Psychology: Learning, Memory, and Cognition, 22,
1336-1353. Bradshaw, J. L., & Nettleton, N. C. (1981). The nature of hemispheric specialization in man. Behavioral and Brain Sciences, 4, 51-91. BucKner, R. L., Petersen, S. E., Ojemann, J. G., Miezin, F. M., Squire, L, R., & Raichle, M. E. (1995). Functional anatomical studies of explicit and implicit memory retrieval tasks. Journal of Neuroscience, 15, 1229. Burgund, E. D., & Marsolek, C. J. (1996, April). Letter-case specific
priming in the right cerebral hemisphere on a modified perceptual identification task. Presented at the Cognitive Neuroscience Society
Annual Meeting, San Francisco, CA. Cave, C. B., & Squire, L. R. (1992). Intact and long-lasting repetition priming in amnesia. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 509-520. Christman, S., Kitterle, F. L., & Hellige, J. B. (1991). Hemispheric asymmetry in the processing of absolute versus relative spatial frequency. Brain and Cognition, 16, 62-73. Churchland, P. S., & Sejnowski, T. J. (1992). The computational brain. Cambridge, MA: MIT Press. Cohen, N. J., & Eichenbaum, H. (1993). Memory, amnesia, and the hippocampal System. Cambridge, MA: MIT Press. Cooper, L. A., Schacter, D. L., Ballesteros, S., & Moore, C. (1992). Priming and recognition of transformed three-dimensional objects: Effects of size and reflection. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 43-57. Corballis, M. C. (1989). Laterality and human evolution. Psychological Review, 96, 492-505. Damasio, A. R., & Damasio, H. (1983). The anatomic basis of pure alexia. Neurology, 33, 1573-1583. Diamond, R., & Carey, S. (1986). Why faces are and are not special: An effect of expertise. Journal of Experimental Psychology: General, 115, 107-117. Dimond, S. J., & Beaumont, G. (1971). Use of two cerebral hemispheres to increase brain capacitff. Nature, 232, 270-271. Dimond, S. J., Gibson, A. r~., & Gazzaniga, M. S. (1972). Cross field and within field integration of visual information. Neuropsychologia, 10, 379-381. Farah, M. J. (1990). Visual agnosia: Disorders of object recognition and what they tell us about normal vision. Cambridge: MIT Press. Farah, M. J. (1991). Patterns of co-occurrence among the associative nOSias: Implications for visual object representation. Cognitive
uropsychology, 8, 1-19.
Farah, M . L (1994). Neuropsychological inference with an interactive brain: A critique of the locality assumption. Behavioral and Brain Sciences, 17, 43-104.
Visual Form Recognition 155 Felleman, D. J., & Van Essen, D. C. (1991). Distributed hierarchical processing, in primate visual cortex. Cerebral Cortex, 1, 1-47. Fox, P. T., Miezin, F. M., Allman, J. M., Van Essen, D. C., & Raichle, M. E. (1987). Retinotopic organization of human visual cortex mapped with positron emission tomography. Journal of Neuroscience, 7, 913922. Geffen, G., Bradshaw, J. L., & Wallace, G. (1971). Interhemispheric effects on reaction time to verbal and nonverbal visual stimuli. Journal of Experimental Psychology, 85, 415-422. Goodale, M. A., & Milner, A. D. (1992). Separate visual pathways for perception and action. Trends in Neuroscience, 15, 20-25. Gross, C. G., Rocha-Miranda, C. E. & Bender, D. B. (1972) Visual rOperties of neurons in inferotemporal cortex of the macaque. urnal of Neurophysiology, 35, 96-111. Haxby, J. V., Grady, C. L., Horwitz, B., Ungerleider, L. G., Mishkin, M., Carson, R. E., Herscovitch, P., Schapiro, M. B., & Rapoport, S. I. (1991). Dissociation of object and spatial visual processing pathways in human extrastriate cortex. Proceedings of the National Academy of Sciences USA, 88, 1621-1625. Hellige, J. B. (1983). Hemisphere x task interaction and the study of laterality. In J. B. Hellige (Ed.), Cerebral hemispheric asymmetry: Method, theory, and research (pp. 411-443). New York: Praeger Hellige, J. B. (1993a). Hemispheric asymmetry: What's right and what's left. Cambridge, MA: Harvard University Press. Helllge, J. B. (1993b). Unity of thought and action: Varieties of interaction between the left and fight cerebral hemispheres. Current Directions in Psychological Science, 2, 21-25. Hinton, G. E., McClelland, J. L., and Rumelhart, D. E. (1986). Distributed representations. In D. E. Rumelhart & J. L. McClelland (Eds.), Parallel distributed processing: Explorations in the microstructure of cognition. Volume 1: Foundations (pp. 77-109). Cambridge, MA: MIT Press. Hummel, J. E., & Biederman, I. (1992). Dynamic binding in a neural network for shape recognition. Psychological Review, 99, 480-517. Jacobs, R. A., & Kosslyn, S. M. (1994). Encoding shape and spatial relations: The role of receptive field size in coordinating complementary representations. Cognitive Science, 18, 361-386. Kitterle, F. L., Christman, S., and Hellige, J. B. (1990). Hemispheric differences are found in the identification, but not the detection, of low versus high spatial frequencies. Perception & Psychophysics, 48, 297-306. Kitterle, F. L., & Selig, L. M. (1991). Visual field effects in the discrimination of sine-wave gratings. Perception & Psychophysics, 50, 15-18. Knapp, A., & Anderson, J. A. (1984). A signal averaging model for concept formation. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 616-637. Koivisto, M . (1995). On functional brain asymmetries in perceptual pnmmg. Brain and Cognition, 29, 36-53.
156
Marsolek and Burgund
Komatsu, Y., Fujii, K., Maeda, J., Sakaguchi, H., & Toyama, K. (1988). Long-term potentiation of synaptic transmission in kitten visual cortex. Journal of Neurophysiology, 59, 124-141. Kosslyn, S. M. (1987). Seeing and imagining in the cerebral hemispheres: A computational approach. Psychological Review, 94, 148-175. Kosslyn, S. M., Chabris, C. F., Marsolek, C. J., & Koenig, O. (1992). Categorical versus coordinate spatial relations: Computational analyses and computer simulations. Journal of Experimental Psychology: Human Perception and Performance, 18, 562-577. Kosslyn, S. M., & Koenig, O. (1992). Wet mind: The new cognitive neuroscience. New York: Free Press. Kosslyn, S. M., Alpert, N. M., Thompson, W. L., Maljkovic, V., Weise, S. B., Chabris, C. F., Hamilton, S. E., Rauch, S. L., Buonanno, F. S (1993). Visual mental imagery activates topographically organized visual cortex: PET investigations. Journal of-Cognitive Neuroscience, 5, 263-287. LaBerge, D. (1995). Computational and anatomical models of selective attention in object identification. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 649-663). Cambridge, MA: MIT Press. Lehky, S. R. & Sejnowski, T. J. (1988). Network model of shape-fromshading." Neural function arises from both receptive and projective fields. Nature, 333, 452-454. Levine, D., & Calvanio, R. (1989). Prosopagnosia: A defect in visual configural processing. Brain and Cognition, 10, 149-170. Marr, D. (1982). Vision. San Francisco, CA: W. H. Freeman. Marshall, J. C. (1981). Hemispheric spe_cialization: What, how, and why. Behavioral and Brain Sciences, 4, 72-73. Marsolek, C. J. (1994, March). Computational incompatibility for processing specific- vs. abstract-visual-form information. Presentedat the Cognitive Neuroscience Society Annual Meeting, San Francisco, CA. Marsolek, C. J. (1995). Abstract-visual-form representations in the left cerebral hemisphere. Journal of Experimental Psychology: Human Perception and Performance, 21,375-386. Marsolek, C. J., Ecker, B. C, Ewald, B. J., Stoltz, E. L., Wright, M. E., & Nicholas, C. D. (1995, November). Letter-case similarity and backward pattern masking modulate form-specific priming in th e cerebral hemispheres. Presented at the Psychonomic Society Annual Meeting, Los Angeles, CA. Marsolek, C. J., Kosslyn, S. M., & Squire L. R. (1992). Form-specific visual priming in the right cerebral hemisphere. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18, 492-508. Marsolek, C. J., Schacter, D. L., & Nicholas, C. D. (1996). Form-specific visual priming for new associations in the fight cerebral hemisphere. Memory & Cognition, 24, 539-556. Marsolek, C. J., Squire, L. R., Kosslyn, S. M., & Lulenski, M. E. (1994). Form-specific explicit and implicit memory in the right cerebral hemisphere. Neuropsychology, 8, 588-597.
Visual Form Recognition 157 Mazzoni, P., Andersen, R. A., & Jordan, M. I. (1991). A more biologically plausible learning rule for neural networks. Proceedings of the National Academy of Sciences USA, 88, 4433-4437. McClelland, J. L., McNaughton, B. L., & O'Reilly, R. C. (1995). Why there are complementary learning systems in the hippocampus and neocortex: Insights from the successes and failures ofconnectionist models of learning and memory. Psychological Review, 102, 419457. McClelland, J. L., & Rumelhart, D. E. (1985). Distributed memory and the representation of general and specific information. Journal of Experimental Psychology: General, 114, 159-188. Nicholas, C. D., & Marsolek, C. J. (1996, April). Differential
interhemispheric transfer for abstract and specific visual-form information. Presented at the Cognitive Neuroscience Society Annual
Meeting, San Francisco, CA. O'Reilly, R. C., Kosslyn, S. M., Marsolek, C. J. & Chabris, C. F. (1990). Receptive field characteristics that allow parietal lobe neurons to encode spatial properties of visual input: A computational analysis. Journal Of Cognitive Neuroscience, 2, 141-155. Petersen, S. E., & Fiez, J. A. (1993). The processing of single words studied with positron emission tomography. Annual Review of Neuroscience, 16, 509-530. Plaut, D. C., & Hinton, G.E. (1987). Learning sets of filters using backpropagation. Computer Speech and Language, 2, 35-61. Rhodes, G. (1985). Laterallzed processes m face recognition. British Journal of Psychology, 76, 249-271. Robertson, L. C,, & Lamb, M. R. (1991). Neuropsychological contributions to theories of part/whole organization. Cognitive Psychology, 23, 299-330. Rueckl, J. G., Cave, K. R., & Kosslyn, S. M. (1989). Why are "what" and "where' p r o c e s s e d by separate cortical systems? A computational investigation. Journal of Cognitive Neuroscience, 1, 171-186. Rumelhart, D. E., Hinton, G., & Williams, R. (1986). Learning internal representations by error propagation. In D. E. Rumelhart & J. L. McClelland (Eds.), Parallel distributed processing: Explorations in the microstructure of cognition. Volume 1: Foundations (pp. 318362). Cambridge, MA: MIT Press. Schacter, D. L., Reiman, E., Uecker, A., Polster, M. R., Y un, L. S., & Cooper, L. A. (1995). Brain regions associated with retrieval of structurally coherent visual information. Nature, 376, 587-590. Schacter, D. L., & Tulving, E. (Eds.). (1994). Memory systems 1994. Cambridge, MA: MIT Press. Schwartz, E. L., Desimone, R., Albright, T. D., & Gross, C. G. (1983). Shape recognition and inferior temporal neurons. Proceedings of the National Academy of Sciences USA, 80, 5776-5778. Sergent, J. (1982). The cerebral balance of power: Confrontation or cooperation? Journal of Experimental Psychology: Human Perception and Performance, 8, 253-272.
158
Marsolek and Burgund
Sergent, J. (1989). Structural processing of faces. In A. W. Young & H. D. Ellis (Eds.), Handbook of research on face processing (pp. 5791). Amsterdam: Elsevier. Sergent, J., & Bindra, D. (1981). Differential hemispheric processing of faces: Methodological considerations and reinterpretation. Psychological Bulletin, 89, 541-554. Sergent, J., Ohta, S., & MacDonald, B. (1992). Functional neuroanatomy of face and object processing. Brain, 115, 15-36. Squire, L. R. (1992). Memory and the hippocampus: A synthesis from findings with rats, monkeys, and humans. Psychological Review, 99, 195-231. Squire, L. R., Ojemann, J. G., Miezin, F. M., Petersen, S. E., Videen, T. O., & Raiclile, M. E. (1992). Activation of the hippocampus in humans: A functional anatomical study of memory. Pi"oceedings of the National Academy. of Sciences USA, 89, 1837-1841. Tarr, M. J. (1995). Rotating objects to recognize them: A case study on the role of viewpoint dependency in the recognition of threedimensional objects. Psychonomic Bulletin & Review, 2, 55-82. Tootell, R. B. H., Silverman, M. S., Switkes, E., & De Valois, R. L. (1982). Deoxy~glucose analysis of retinotopic organization in primate striate cortex. ~cience, 218, 902-904. Ungerleider, L. G. & Mishkin, M. (1982). Two cortical visual systems. In D. J. Ingle, M. A. Goodale, & R. J. W. Mansfield (Eds.), The analysis of visual behavior (pp. 549-586). Cambridge, MA: MIT Press. Van Kleeck, M. H. (1989). Hemispheric differences in global versus local processing of hierarchical visual stimuli by_ normal subjects: New data and a meta-analysis of previous studies. Neuropsychologia, 27, 1165-1178. Zipser, D. & Andersen, R. A. (1988). A back-propagation programmed network that simulates response properties of-a subset o! posterior parietal neurons. Nature, 331,679-684.
mm
mu
O0
mm me( )nil
me(
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 1997 Elsevier Science B.V.
161
Chapter 6
Amplification of Spatial Nonuniformities by Guided Search Mechanisms E. William Yund Neurophysiology-Biophysics
Research Laboratory
VA Medical Center, Martinez, CA and Department of Neurology University of California, Davis "I'he primary focus of this chapter is to describe and explain the c~:~nsistent location-specific performance differences seen in visual search, in the context of the Guided Search Model (GSM) as developed by Wolfe and his colleagues (Cave & Wolfe, 1990; Wolfe, 1994; Wolfe, 1996; Wolfe, Cave & Franzel, 1989). This topic is important to those interested in hemispheric asymmetries because GSM provides a mechanism that can transform minor, virtually insignificant, perceptual or attention-control right-left spatial nonuniformities into major performance asymmetries. The existence of such a mechanism indicates that even a major right-left visual field asymmetry for a particular stimulus type cannot be taken as strong evidence for hemispheric specialization in the processing of that type of stimulus in the intact
162 Yund human brain. Of course, if it is possible to dissociate the right-left visual field asymmetry from the GSM explanation in a particular case, then the hemispheric specialization conclusion becomes more tenable there. In the case of our own location-specific visual search results (Efron & Yund, 1996; Yund & Efron, 1996), however, the right-left asymmetries co-varied with the effects of other spatial nonuniformities of retinal origin as would be expected if all of these performance advantages occurred solely because of signal level nonuniformities in the bottomup stages of GSM. The chapter is divided into three major sections. The first section provides a brief introduction to visual search and GSM. The second section presents recent results from our laboratory concerning spatial nonuniformities in visual search and how these results can be understood within GSM. The implications of these results for the interpretation of visual-field/hemispheric asymmetries will be emphasized. The final section includes more general discussion of possible sources of right-left spatial nonuniformities. VISUAL S E A R C H AND THE GUIDED S E A R C H M O D E L
Visual search is a common task: We try to find a friend's face in a crowd, a particular book on a shelf, or an item we need in a cluttered drawer. How can we find such items visually? Considerable study has been devoted to various aspects of visual search and, as a result, it is well known that the speed and accuracy of target detection depends in large part on the stimulus parameters of the target and distractor items. When the target differs from otherwise identical distractors on only one major stimulus dimension, it may be very easy to detect. A red target, for example, is easy to find among green distractors. Similarly, a vertical line is easy to find among horizontal distractors, but more difficult to find among distractors with orientations closer to vertical, especially if the distractors' orientations differ from that of the target in both clockwise and counterclockwise directions. Fortunately, the details of the effects of stimulus parameters in visual search tasks are not critical in the present context, but it is important to remember that stimulus parameters are a major factor in visual search. Of particular interest here is the effect of the target's spatial location on its detection. Visual search experiments usually present target and distractor items at random locations throughout the visual field in order
Guided Search
163
to control for any location-specific effects that might be present. In experiments designed to measure location-specific effects, Robert Efron and I have demonstrated that the speed and accuracy of target detection in visual search depends on the location of the target as well as the locations of the distractors in the visual field (Buckles, Yund & Efron, 1991; Efron & Yund, 1996; Efron, Yund & Nichols, 1987, 1990a, 1990b, 1990c; Ostrosky-Solis, Efron & Yund, 1991; Yund & Efron, 1996; Yund, Efron & Nichols, 1990a, 1990b, 1990c). This dependence of detection performance on both target and distractor locations suggests some sort of competition among locations for a limited resource that is essential in identifying the target. GSM describes such a competitive mechanism for the allocation of the limited resource, attention. It is important to note, however, that GSM was not developed to account for location effects, but rather for differences in the slope of the function relating reaction time (RT) to the number of distractors: Depending on a variety of stimulus parameters, adding more distractors might leave the RT unchanged or cause it to increase by a small, intermediate or large amount. 1 GSM is described in detail elsewhere (Cave & Wolfe, 1990; Wolfe, 1994; Wolfe, 1996; Wolfe, Cave & Franzel, 1989) and will be summarized here only briefly; the reader is encouraged to consult the original sources and reviews (e.g., Bundesen, 1996) in order to appreciate the full power of this model as well as the context within which it was developed. The essence of GSM is the competition for attention among stimulus items presented simultaneously at different locations in the visual field. The mechanism of the competition is the signal level generated by each item at its location in the GSM activation map. In effect, the signal levels in the activation map determine a prioritized list of locations to be attended, with item locations having the highest signal levels being attended first. Top-down guidance (filtering and/or selective amplification processes that depend on the expected identity of the target item, as well as expected target-distractor differences) and bottom-up visual signal processing (including local contrast mechanisms and/or any spatial nonuniformities 2 that might affect signal level) combine to determine the overall relative signal levels and thus the order of locations on the prioritized list. The efficiency of GSM in putting the target location at the top of the list in a particular search task depends on both top-down and bottom-up processes. The greater the selectivity of the top-down guidance for the
164 Yund target, the closer to the top of the list the target location will be. Similarly, the greater the signal strength generated by the target in bottom-up processing, the closer to the top of the list the target location will be. Although top-down guidance and bottom-up processing are both contributing to signal level differences in the GSM activation map, the two are fundamentally different because only the top-down guidance mechanism can be programmed to use information concerning the identity of the target. The hardwired bottom-up processing is based on physiological and/or perceptual differences, yielding an enhanced signal level when an item differs from surrounding items in fundamental stimulus dimensions or features. Similarly, any spatial nonuniformity in the sensitivity of sensory/ perceptual processes to target and distractor features also would affect bottom-up signal levels. As a result of the fundamental difference between top-down guidance and bottom-up processing, the two may either reinforce each other in putting the target at the top of the list, or they may act in opposition. Consider two examples: (1) Top-down guidance and bottom-up processing would reinforce each other in the search for a red target among otherwise identical green distractors. The top-down guidance would select for the red color while the bottom-up processing would generate greater activation for the target than the distractors because the red target would differ more from surrounding green distractors than any green distractor would differ from surrounding items, which would be predominantly identical green distractors even if the red target were nearby. (2) Top-down guidance and bottom-up processing would act in opposition in a typical conjunction search, e.g. for a red vertical target among red horizontal and green vertical distractors. The top-down guidance would select for red and vertical as much as possible in order to bring the target to the top of the list3. The bottom-up activation at each target and distractor location, however, would depend on the spatial configuration of the target and distractors. A red horizontal distractor surrounded by green vertical distractors would generate more bottom-up activation than the target surrounded by any mix of distractors because the red horizontal distractor differs from each surrounding green vertical distractor in both color and orientation, while the target differs from each surrounding distractor in either color or orientation. In random arrangements of target and distractors, it should be common for some distractors to be surrounded
Guided Search
165
by a set of items in which the other distractor predominates and thus it should be common for some distractors to receive more bottom-up activation than the target 4. Of course, the ultimate position of the target on the list would depend on the relative strengths of the top-down and bottom-up activation differences. One other aspect of GSM is critical in the present context: GSM is capable of amplifying seemingly insignificant spatial nonuniformities into large RT or target detection differences. The key to this amplification effect is that the spatial nonuniformities exert their effects in the part of the system that determines the order of subsequent serial attentive processing. Under these circumstances, a small but consistent spatial nonuniformity, that could hardly be measured with a method employing single stimuli, could lead to a consistent order difference in GSM's list when multiple simultaneous stimuli are employed. Thus, GSM predicts that competing stimuli will reveal spatial (or other perceptual) nonuniformities that may not be apparent when only one stimulus is present. Remembering that top-down guidance and bottomup processing often act in opposition leads to the further expectation that the effects of spatial nonuniformities will decrease as the subject learns, from experience with the task, to increase the efficacy of topdown guidance. The reader may wonder, irrespective of the above discussion of GSM, how a visual search model may be relevant to an area of research (fight-left visual field asymmetries/hemispheric asymmetries) in which visual search for a specified target is not a common methodology. This question should be addressed directly, and the answer requires a slightly broader perspective on GSM. In visual search paradigms, GSM provides a way of allocating a limited serial resource (attention) to multiple simultaneously-presented visual objects. Although a given visual field asymmetry paradigm may not involve search for any particular predefined "target", it may involve the serial allocation of attention to multiple simultaneously- presented visual objects. The top-down guidance component of GSM cannot help us if no target features are defined, 5 but the bottom-up component of GSM and the activation map can provide a framework for understanding the attentional scan of the objects present. Consider, for example, an experiment in which the task is to identify some text (words or nonsense words) presented briefly in corresponding locations on both sides of the vertical meridian. Under such circumstances, spatial nonuniformities in the bottom-up processing
166 Yund should be manifest in the serial allocation of attention just as they are in our visual search experiments. Indeed, spatial nonuniformities may be seen even more clearly in experiments with no predefined target, because differences in bottom-up activation need not compete with target-specific top-down guidance in determining the order of the scan of attention. SPATIAL NONUNIFORMITIES IN VISUAL SEARCH Recent experiments in our laboratory concerning spatial nonuniformities in visual search suggest that a right-left visual field asymmetry is only one of several spatial nonuniformities in performance that may be mediated through bottom-up components of a guidance mechanism like that of GSM. Several experimental results seem to demonstrate that all of these spatial nonuniformities in performance vary in the same way with stimulus and task parameters. Only those results that seem particularly relevant to the nature of the right-left asymmetry will be discussed in this section; the full set of experiments are described and discussed in a more general context elsewhere (Efron & Yund, 1996; Yund & Efron, 1996). The reader also should consult those reports for complete descriptions of methods and statistical analyses, as needed. In all of our experiments to be discussed here, the target of the visual search was a 1.3" square filled with 4 cycles of a luminance square wave grating pattern. In different experiments, the target orientation was either vertical or horizontal. Distractor stimuli were the same size squares (always with horizontal and vertical sides) filled with square wave gratings of different orientations or with checkerboard patterns. Stimuli were presented on a background such that light parts of the patterns were above the background luminance level and dark parts were below the background luminance level. Distractor patterns were always all different from each other (heterogeneous) and were randomly assigned to locations independently for each trial of each experiment. The response was a single button, go-no go reaction time (RT) measured with millisecond resolution from the vertical sync of the first frame of the stimulus display. The display lasted for 3 frames (nominally 50 ms). Stimuli were viewed binocularly.
Guided Search
E
t
167
Horizontal Target
_
O
b-O iim
m
tr L
v
r---] 1L
2L
w
I
I
I
I
3L
4L
5L
6L
1R
2R
3R
4R
5R
8R
Target Location
Vertical Target o
mmm
4mr
m
(I)
n-"
=---r~!---I 1L
2L
I
I
I
I
3L
4L
5L
6L
I'-'-'-"
1R
1
1
I
I
!
2R
3R
4R
SR
SR
Target Location F i g u r e 1: The results of Experiment A for the horizontal and vertical targets. Target locations are labeled with a number (1 through 6) to indicate the vertical position on the circle and a letter (L or R) to indicate the visual halffield. Circle
168 Yund symbols plot simple RT; triangles plot choice RT; squares plot 12-pattern visualsearch RT. The fine line without symbols plots the display delay of the different locations because of the VGA raster display. The ordinate is a relative-RT scale with ticks at 10 ms intervals because the choice and visual-search curves are shifted down to facilitate shape comparisons.
Experiment A: Importance of Multiple Stimuli The purpose of Experiment A (called Experiment D in Efron & Yund, 1996) was to determine the importance of the presence of multiple stimuli on a trial in generating the RT differences we had found among different target locations when multiple stimuli were present. In Experiment A, the same subjects were tested in 6 different conditions, 3 with the horizontal and 3 with the vertical target. The first condition with each target measured simple RTs to the appearance of a pattern (always the target) at each of 12 locations on a circle with a 4.6 radius around the fixation point. The purpose of this condition was to measure any spatial nonuniformities in the simple detection if the target stimulus at the 12 locations. In the second condition, again, only one pattern appeared on each trial, but half of the patterns were the target and half were the orthogonally oriented distractor. Choice RTs for discriminating the target from the distractor were measured at each of the 12 locations. The purpose of this condition was to measure any spatial nonuniformity in target-distractor discrimination in the absence of patterns at other locations (i.e., in the absence of any competition). The third condition with each target measured the visual search RT for detecting the target when a pattern was present at each of the 12 locations at the same time. The p u r p o s e of this c o n d i t i o n was to m e a s u r e the spatial nonuniformities in visual search for comparison with the simple and choice RTs of the same subjects. The results for the horizontal and vertical targets are presented in the upper and lower panels of Figure 1, respectively. The ordinate is a relative-RT scale because the choice RT curves (triangle symbols) were shifted downward by 80 ms and the visual-search RT curves (square symbols) were shifted downward by 180 ms to facilitate shape comparisons among these curves. Target locations, on the abscissa, are labeled with a number (1 to 6 for locations from the top to the bottom of the circular configuration) and a letter (L or R for which side of the vertical meridian). The only apparent spatial nonuniformity seen in the simple RTs (circle symbols) to either target was a small increase in RT
Guided Search
169
from the top to the bottom of the display, and this nonuniformity disappeared when we corrected for the difference in display times between the different locations due to the raster display of the VGA monitor. The simple RT results are plotted in Figure 1 without correction for the raster delay, and the lines with no symbols plot the raster delay itself. The average RT was slightly faster in the fight than in the left visual field for both horizontal (0.87 ms) and vertical (1.13 ms) targets, but neither of these differences approached statistical significance (F < 1.0 for the field effect in both cases). In the single-choice RTs, the effect of the raster delay is no longer apparent, particularly in the case of the horizontal target for which locations near the top of the display seem to have longer rather than shorter RTs. Small spatial nonuniformities appeared in the choice RT results: both the right-left asymmetry and the position effect were significant for both the horizontal and the vertical target. The RVF RT advantages for the horizontal and vertical targets were 5.36 and 5.96 ms, respectively. Comparison of the choice- and simple-RT position effects indicates that these were not determined by the same processes: Although in the case of the vertical target, the two curves seem to have essentially the same linear trends in both visual half-fields (RT increasing from locations 1 to 6), in the case of the horizontal target the linear trends are in opposite directions. RTs decreased from locations 1 to 6 in both visual fields for the horizontal choice RT condition. Similarly, the horizontal- and vertical-target visual search data cannot be easily related to that of the simple RT and single-pattern choice RT conditions. RTs for the horizontal target revealed the strongest position effect seen in Experiment A, with slower RTs at the top and bottom of the circle of stimuli; this U-shaped function is particularly clear in the left field. RTs for the vertical target show a much weaker position effect which was not statistically significant in this experiment. A significant right-left asymmetry was present for both horizontal and vertical targets and was also larger in magnitude than that found in the single-pattern choice condition. The RVF advantages in visual search conditions were 52.31 ms for the horizontal target and 28.06 ms for the vertical target. In summary, we found no evidence in Experiment A that the spatial nonuniformities of visual search have precursors that can be measured with single-stimulus methods, indicating that the competition of multiple simultaneous stimuli is critical in revealing these spatial nonuniformities.
170 Yund Even in the case of the RVF advantage that was present in both single choice and visual search RTs, that RVF advantage was virtually identical for the horizontal and vertical targets in the single choice condition (5.36 versus 5.96 ms) and was very different in the visual search condition (52.31 versus 28.06 ms). Furthermore, we found a large difference in the visual search position effect measured with horizontal as opposed to vertical targets, suggesting a perceptual spatial nonuniformity underlying the performance differences in visual search for horizontal and vertical square wave grating stimuli. In addition to the well known loss of visual resolution as the point of stimulation is moved away from the fovea, there are two other spatial nonuniformities that might affect ability to detect and/or discriminate these square wave stimuli: (1) The falloff of cone receptor density (Curcio, Sloan, Packer, Hendrickson & Kalina, 1987) and visual resolution (Rijsdijk, Kroon & van der Wildt, 1980) along the horizontal meridian occurs more slowly than along the vertical meridian, yielding an elliptically-shaped equal-resolution contour called the "visual lobe". By itself, this visual lobe effect might account for a U-shaped RT position gradient like that seen for the horizontal target in Figure 1, but it would not account for the difference in position effects for the two target orientations. (2) Detection and discrimination of oriented stimuli like our square wave gratings also has been shown to be better when the orientation of the grating is parallel to a radial meridian passing through the fovea (Rovamo, V irsu, Laurinen & Hyv~irinen, 1982; Scobey & van Kan, 1991; Temme, Malcus & Noell, 1985; Toet & Levi, 1992). This "radial organization" effect also might account for a U-shaped position gradient for the horizontal target, because the orientation of the horizontal target is almost radial at positions 3 and 4 (near the horizontal meridian) and almost perpendicular to radial at positions 1 and 6 (near the vertical meridian). In the case of the vertical target, however, the radial orientation effect should produce an inverted U-shaped position gradient, because the orientation of the vertical target is almost perpendicular to radial at positions 3 and 4 (near the horizontal meridian) and almost radial at positions 1 and 6 (near the vertical meridian). Thus a combination of visual lobe and radial organization effects is needed to account for the different position effects seen for the horizontal and vertical target patterns in the visual search RTs of Experiment A. Of course, such small differences in visual resolution would not be expected to produce the observed large differences in RT
Guided Search
171
without an amplification mechanism such as that provided by GSM when multiple stimuli are present. The next experiment was designed to replicate the horizontal and vertical target effects in a different spatial configuration and a new group of subjects, as well as to study the nature of the right-left asymmetry.
Experiment B: Visual Lobe, Radial Organization, and Right-Left Asymmetry The patterns and pattern locations used in Experiment B (called Experiment F in Efron & Yund, 1996) are illustrated in Figure 2. There are 4 positions arranged in square configuration in each quadrant: The centers of positions 1 and 4 are on the 45 diagonal at a distance of 4.4 and 7.3 from the central fixation; positions 2 and 3 are 6.1 from fixation and position 2 is closer to the horizontal meridian while position 3 is closer to the vertical meridian. Positions 2 and 3 are numbered to be consistent with the expected effects of the visual lobe and radial organization factors. For both horizontal and vertical target patterns, the visual lobe factor should reduce the RT at position 2 relative to position 3. For the horizontal target, the radial organization factor also should reduce the RT at position 2 relative to position 3, but for the vertical target, the radial organization factor should reduce the RT at position 3 relative to position 2. Thus we should expect the horizontal target results to show faster RTs at position 2 compared to position 3 and we should expect the vertical target results to show a smaller position-2 advantage over position 3, although the actual relative values for the two positions would depend on the relative strengths of the visual lobe and radial organization factors for the vertical pattern at the two positions. We also should expect position 1 to yield the fastest RTs and position 4 to yield the slowest RTs for both target orientations because of the decrease in retinal resolution with distance from the fovea: It would be quite surprising if the visual lobe and/or radial organization factors were sufficient to counteract the fundamental distance factor of retinal resolution. The results for the position effect in Experiment B, presented in Figure 3, are entirely consistent with the expectations described above. RTs were faster at position 2 than at position 3 for the horizontal target and that difference was reduced, indeed reversed, for the vertical target.
172 Yund
13] i4 I
fill3
1
2 7.3 ~
2 -1.3s
4.4 ~
6.1 ~
2
i 2 4
Figure 2: The stimulus patterns (top) and the location configuration (bottom) used in Experiment B. In all of the experiments, the black-and-white patterns were presented on an intermediate-level (gray) background. Panel B is drawn to illustrate the size and positions of all possible locations used as well the numerical position labels. Only a small fixation cross and 8 patterns (those in the fight, left, upper, or lower half-field) were present during one triM.
Guided Search
173
The faster RT at position 3 than at position 2 for the vertical orientation suggests that the radial organization factor was stronger than the visual lobe factor under these conditions. Of course, the relative strengths of these two factors is not critical in the present context, but comparable strengths for the two factors would account for the virtually flat RT gradient seen for the vertical target and the very steep RT gradient seen for the horizontal target in the circular configuration of Experiment A. In an earlier experiment measuring target detection rather than RT (Yund, Efron & Nichols, 1990c), stimuli on both sides of the vertical meridian were necessary to produce a robust right-left detection asymmetry; multiple stimuli separately on either side alone produced almost symmetrical detection performance in the two visual fields. Because that experiment was vulnerable to the criticism that it included stimuli quite close to the vertical meridian, 6 the closest edge of any pattern was 2.46 from the meridian in the present experiment. To further examine the nature of the right-left asymmetry in Experiment B, only 8 patterns were presented on each trial, 4 in each quadrant of one visual half-field: right, left, upper, or lower. In this way, we could measure RTs for patterns in one quadrant when 4 distractors in a different quadrant were either on the same side or the opposite side of the vertical meridian. In order to account for our right-left asymmetry in visual search in the context of GSM, we (Efron & Yund, 1996) have suggested a right-left spatial nonuniformity in "signal strength", such that right field stimuli generate greater signal strength than otherwise equivalent left field stimuli at equivalent locations. Our initial assumption was that right-left differences in signal strength are the same as other differences in signal strength attributed to other spatial nonuniformities (e.g. visual lobe or radial organization). If a right-left spatial nonuniformity in the parallel processing stage of GSM is responsible for the right-left asymmetry in visual search, then the effect of the 4 nontarget-quadrant distractors in this experiment should be greater when they are in the right field as opposed to the left field. A model of the right-left asymmetry in which there is an independent ability to identify the target in the right and left visual fields would yield greater distractor effect when the 4 nontarget-quadrant distractors are on the same side of the vertical meridian as the target (right or left) than when the target and nontarget-quadrant distractors are on the opposite side of the vertical meridian. Of course, right-left nonuniformity in the parallel-processing stage of GSM is not the only model within which
174 Yund
Vertical Target
Horizontal Target
I 500
490
480
RT (ms) 470
460
450
440
1
2
3
4
1
2
3
4
Target Position Figure 3: RTs in Experiment B for the different target positions for the horizontal and vertical targets. Note that the reaction time differences between positions 2 and 3 were different for the two targets.
Guided Search
175
right-field distractors would have a greater effect than left-field distractors; any general type of model with a right-field bias prior to a stage where all stimuli must compete for some limited resource or capacity, would make the same predictions as GSM, in this case. It is difficult to predict whether such a general type of model would produce a right-left field asymmetry for multiple-pattern displays with all patterns on one or the other side of the vertical meridian. Within GSM, for example, the prediction would depend on how the spatial nonuniformity affected the signal levels in the activation map. If the bottom-up signal strength for the left field were reduced relative to the fight field, that should yield a lower performance level (higher RTs) in the left field alone relative to the right field alone, because the GSM selection mechanism compares activation peaks generated at target and distractor locations and would be operating at a lower signal-to-noise ratio in the left field alone than in the right field alone. Alternatively, if there were a fight field bias at a later stage in the GSM model, where it would not affect the quality of target selection in the left field alone, then only a small visual-search asymmetry would be expected without simultaneous stimuli on both sides of the vertical meridian. Under this alternative, the within-field search would be like the single-stimulus choice RT conditions of Experiment A, because none of the stimuli present on the same trial would receive any advantage or disadvantage due to a fight-left field bias. The RTs for each target quadrant when the nontarget-quadrant distractors were in the left and right fields are plotted as the first four bars on the left in Figure 4; the black section of each bar indicates the RT when the nontarget-quadrant distractors were in the left field and the grey section indicates the increase in RT that occurred when the nontarget-quadrant distractors were in the fight field. The presence of the grey section on the top of all four bars means that the quadrant-competition RTs were completely consistent with the GSM expectation: Nontarget-quadrant distractors increased RTs more when they were in the RVF than in the LVF, independent of the location of the target quadrant. (Statistically, distractor field showed a significant main effect and no interaction with target quadrant.) The two bars on the right in Figure 4 show the left and right field RTs when all of the stimuli were on one side of the vertical meridian. Thus, this experiment revealed a large RVF RT advantage for unilateral displays. As noted above, this right-left difference is more consistent with a relatively low level, perceptual source for the field advantage, than
176 Yund 510 500 490
RT 480
(ms) 470 460 450 44O LL
UL
UR
LR
Target Quadrant
All L
All R
Field
Figure 4: The results of Experiment B by target quadrant and field. The lower black sections of the quadrant bars represent the RTs when the competing, nontarget quadrant was in the LVF and the full bars represent the RTs when the competing, nontarget quadrant was in the RVF. The top, gray section of each bar shows the increase in RT when the competing, nontarget quadrant was in the RVF compared to the LVF. The "Field" RTs represent unilateral visual field performance, when all patterns were in the LVF or the RVF. There was no difference between umlateral upper- and lower-field results.
a relatively high level, cognitive/attentional source. It is also interesting that the RVF RT advantage seen in Figure 4 is of comparable size to the RT differences seen among positions within quadrants in Figure 3; in particular, the RVF advantage in Figure 4 is 26.11 ms and the position 2-3 difference for the horizontal target in Figure 3 is 27.54 ms, indicating that the magnitude of the right-left perceptual asymmetry needed support the RVF RT advantage is about the same as the sum of the visual lobe and radial organization factors for positions as close together as 2 and 3 in Experiment B.
Guided Search
177
Experiment C: Changes in Position & Field Effects with Learning The purpose of Experiment C (Expt 1 in Yund & Efron, 1996) was to study changes in the position and field RT effects as subjects learned to detect the target more quickly and accurately. In Experiment C, 18 subjects with no prior experience in visual search experiments performed 21 sessions of the horizontal-target multiple-pattern condition used previously in Experiment A. We used the horizontal target because it displayed larger within-field position differences in RTs than the vertical target, and we wanted to track and compare changes in the position and field effects as learning progressed. Not surprisingly, there were large differences among subjects in how quickly they learned to detect the target without error and in how quickly their RTs approached their final levels of performance. These differences in learning rate among subjects, however, are not of particular interest in the present context and will not be considered further here. The changes in position and field effects for all subjects combined are plotted in Figure 5. Each curve in this figure represents an average of 6 sessions beginning with sessions 1, 4, 7, 10, 13, and 16, from the top to the bottom of the graph and plotted with crosses, inverted triangles, diamonds, uptight triangles, circles and squares, respectively. The first six sessions produced a U-shaped RT gradient in both right and left fields, as would be expected for the horizontal target and the combination of visual lobe and radial organization factors. There was also a large RVF advantage in the first six sessions (48.42 ms). As learning proceeded, RTs decreased at all target locations, but the decrease was not the same at all locations. In the right field, where the Ushaped RT gradient was initially more shallow than in the left field, the RT gradient became quite flat and the mean level stopped changing at approximately session 10; the last three right-field curves, representing sessions 10-15, 13-18 and 16-21, overlap so much that it is difficult to see shape differences and their mean RTs are 422.53, 421.45 and 420.30 ms, respectively. In the left field, however, the changes in shape and mean level continued throughout the 21 sessions. It is also interesting that the RT gradient in the left field had about the same shape and amplitude as it did in the right field when the mean left-field RT was approximately equal to that of the right field. Note that the three lower left-field curves and the two upper right-field curves are quite similar in shape, as well as mean RT.
178 Yund
(o
o) n~ od
E
C
0
lira
t~ (.1
0 _1
4-1
" --' In
(1,1
.,q. J
tll
O1
L_
i-
_J (9 .J __J ,qp,,,,,
1
J
I
I_
I
J
I
I
I
I
I
I o
to
to
to
,q-
,q.
,q.
o~1 xr
(stu) F i g u r e 5: Results of Expt C. Each curve represents mean. of 6 sessions starting with 1 (cross), 4 (inverted triangle), 7 (diamond), 10 (triangle), 13 (circle) and 16 (square). Note that most learning occurred in the first half of the experiment in the RVF, but that learning continued throughout the experiment in the LVF.
Guided Search
179
It is beyond the scope of the present chapter to model these results precisely with GSM. It is clear, however, that the pattern of results is entirely consistent with the general predictions of GSM including the visual lobe and radial organization spatial nonuniformities and also an asymmetry in signal strength between the right and left fields. If learning is a result of increasing the selective ability of the top-down guidance, then all of the spatial nonuniformities-including the field asymmetry-should have less and less influence on RTs as learning proceeds, until the RT-gradient curves are essentially flat and the mean RT has reached its minimum, a value that corresponds to the target always attaining the top position on the prioritized list. Given the hypothesized right-left asymmetry in signal strength, we would expect the right field gradient to become flat and reach the RT minimum before the left field RT gradient, just as it does in Figure 5. Similarly, we would expect the left-field performance to continue to improve after the right-field performance had reached its upper limit because the only limiting factor within the model corresponds to a location always being on the top of the list whenever the target was present there. Furthermore, we would expect the shape of the left field gradient to be similar to the right field gradient when they had similar mean levels because the generation of the gradient shape and the mean level are tightly linked within the model. For a more complete description of the logical structure within which these expectations are derived see the explanation of Figure 12 in Efron & Yund (1996).
Experiment D: Location-Specific Learning The purpose of Experiment D (called Expt 3 in Y und & Efron, 1996) was to determine the extent to which the learning seen in Experiment C was location- or field-specific. This question was of interest for two reasons. First, Karni and Sagi (1991) had shown orientation- and location-specific learning with textural stimuli. Although their stimuli were quite different from ours, it seemed important to determine whether our learning effect was similarly specific. We were not surprised to find that the learning was orientation specific (Yund & Efron, 1996: Experiment 2), but it seemed less likely that our subjects were learning to identify targets at particular locations rather than in a more general way throughout the visual field. The demonstration by Shiu and Pashler (1992) of location-specific learning for an orientation-discrimination
180 Yund task, however, suggested that such learning can be restricted to the locations where the task has been practiced. The second reason for interest in the question of location specificity directly involves concepts of visual-field/cerebral-hemisphere asymmetries: Perhaps the learning is specific and somewhat independent in the left and right visual fields, as might be expected if there were critical independent processing and/or learning occurring in the two cerebral hemispheres. I n order to determine the location- and field-specificity of target learning, we tested three separate conditions, each on 10 subjects who had no previous experience in visual search experiments. All three conditions involved our standard circular configuration of 12 locations and concurrent search for the horizontal and vertical targets, only one of which was present on any trial. Unbeknownst to the subjects, the two different targets also were presented in different subsets of the 12 locations. Half of the subjects in each condition started with horizontal target in one subset of locations and the other half of the subjects started with the vertical target in those locations. After 7 sessions were run with the initial target-location relationship the relationship was reversed (again unbeknownst to the subject), such that in the following sessions the horizontal and vertical targets were being detected at locations where they never had been presented in previous sessions. If the learning were location specific, the target- location switch would be expected to cause a major performance decrement (an RT increase). The only difference among the three conditions was in the division of locations into subsets. The "alternating" condition employed subsets consisting of alternating locations (6L, 4L, 2L, 1R, 3R, 5R versus 5L, 3L, 1L, 2R, 4R, 6R), to determine if the learning was restricted to individual locations. The "right-left " condition employed subsets consisting of right- and left-field locations (1R, 2R, 3R, 4R, 5R, 6R versus 1L, 2L, 3L, 4L, 5L, 6L), to determine if the learning might be restricted to the visual field/cerebral hemisphere. The "up-down" condition employed subsets consisting of upper- and lower-field locations (3L, 2L, 1L, 1R, 2R, 3R versus 4R 5R, 6R, 6L, 5L, 4L). The up-down condition was designed as a control for the right-left condition: If location-specific learning occurs over broad areas that include 2 or more adjacent pattern locations, then no evidence of it would be found in the alternating condition although it could account for half-field specificity that might be found in both the right-left and up-down conditions. 7
Guided Search
181
The results of Experiment D are illustrated in Figure 6. In considering the location specificity demonstrated in the present experiment, it is important to understand the numbering of the sessions as plotted on the abscissa: The sessions using the initial target-location relationship are numbered -6 to 0 and session 1 is the first to use the reversed relationship. If location specific learning has occurred in the initial sessions, then there should be a performance discontinuity between sessions 0 and 1, with session 1 showing slower RTs than session 0. For the results of the alternating condition (top panel of Figure 6), there was no indication that anything had changed between sessions 0 and 1 and therefore no indication of location specificity in the learning. On the other hand, RTs clearly increased between sessions 0 and 1 for the right-left condition (the middle panel), indicating that learning was location specific when the "location" was the right or left visual field. In the context of hemispheric specialization, it is especially important to distinguish between two possible explanations for the difference in results between the alternating and right-left conditions: (1) Perhaps the alternating locations were too close together for independent learning to occur at alternating locations. When we trained locations 1R and 3R for one target, the neural units responsible for learning to detect that target at location 2R also may have been learning because the "receptive fields" of those units included major parts of locations 1R and 3R. In order for such across-location-subset learning to obscure location specificity in the right-left condition, the "receptive fields" of such units would have to be much larger, including virtually the entire area covered by our circle of patterns. (2) Perhaps the locations have to be restricted to spatial areas served by opposite cerebral hemispheres for independent learning to occur. When we trained locations on opposite sides of the vertical meridian for different targets, we may have found a way to access independent, or partially independent, learning capacities thzt are not essentially spatial-location specific, but rather cerebral-hemisphere specific. The obvious experiment to distinguish between these two explanations for the alternating and right-left results is to divide the circle of patterns in half along the horizontal rather than the vertical meridian, as was done in the up-down condition, the RTs for which are plotted in the bottom panel of Figure 6. There are similarities and differences between the up-down and right-left results. As can be observed by comparing the RT scale on the abscissa of the three panels
-
-
I
~/
<...
'
"F,/
._
3
1~
0
= Z
O"
"-
w
-
j,
0
O~ 9 M
-
&
I
9
I
'
I
'
RT (ms) I
'
0
[3
|
"0
C
I
I
I ....... I
i
''
I
RT (ms) '
I
i
r-
B
=r
(Q
'
'
I
'
I
'
I
" i'
RT (ms) 'i
i '
m
:I r
m
m
Q
I
O0
Guided Search
183
Figure 6 (previouspage). Results of Expt D. In sessions -6 through 0, horizontal and vertical targets were presented at different subsets of locations. In session 1, the targets switched location subsets. An increase in RT between sessions 0 and 1 indicates that location-specific learning occurred. These results indicate locationspecific learning occurred for the fight-left and updown conditions but that adjacent locations were too close for independent learning to occur in the alternating condition.
of the figure, the up-down group of subjects showed faster initial and final RTs and less difference between initial and final RTs than the other two groups. The scale of the bottom panel was adjusted so that the reader can visually compare the plotted sizes of the RT change between sessions 0 and 1 across the different groups of subjects without mentally correcting for the faster initial and final RTs in the up- down condition. Clearly, the up-down condition produced the same type of RT discontinuity between sessions 0 and 1 as the right-left condition. The absolute size of the RT increase in the up-down condition (21.8 ms) was less than in the right-left condition (31.7 ms), but if those absolute values are normalized by dividing by the magnitude of post-switch learning (the difference between sessions 1 and 6 in each condition, 23.0 and 31.9 ms for up-down and right-left conditions, respectively) the resulting percentage values are 94.8% for up-down and 99.4% for fight-left. Thus, the results from the up-down condition demonstrate that the learning is spatial-location specific not cerebral-hemisphere specific. A location specificity with a radius of about 2 would account for the results of all three conditions. 8 One interesting aspect of these results is not obvious in Figure 6: Comparison of RTs in Experiment D with those of previous experiments (e.g. Experiment C, Figure 5) indicates that the subjects had no more difficulty searching concurrently for the horizontal and the vertical targets than previous subjects had searching for one or the other. This is a little surprising because, in a previous experiment, subjects who had experience searching for one of these targets showed slowed RTs when they switched to searching for the other. In that experiment (Experiment 2, Yund & Efron, 1996), however, the target pattern in one half of the experiment was a distractor in the other half of the experiment, and it may have been the added burden of rejecting the previously-correct target pattern or accepting the previously-incorrect target pattern that produced the orientation specificity of the learning in that experiment.
184 Yund Another interesting aspect of these results is the implication concerning the nature of the top-down selection process of GSM. Initially, the GSM selection process was characterized as a facilitation of the responses of the units encoding target features. In contrast, Treisman (1988) suggested selection by inhibition of units encoding distractor features. Our location-specific results are more consistent with target-feature facilitation than with distractor-feature inhibition because only the target was different at the different locations. Each location in our experiments had randomly equalized experience with all of the distractors, independent of any target-specific experience at that location. GENERAL DISCUSSION
There have been a few experiments in other laboratories concerning visual field effects in visual search. Polich (1982, 1984) reported RVF advantages in experiments directed primarily at parallel/serial distinctions. In the first study (Polich, 1982), subjects decided whether two strings of 2-4 characters (letters or symbols) presented sequentially in the RVF or LVF were the same or different; the RT was measured from the onset of the second string of characters. In the second study (Polich, 1984), subjects had to decide whether a square array of 4, 9, or 16 X symbols contained a single O. Although there are so many differences between the experiments of Polich and our experiments that precise comparisons are impossible, there is nothing in the results of Polich to indicate that his RVF advantages were due to any mechanism other than GSM with the same right-left nonuniformity discussed here. LaBerge and Brown (1986) have reported RVF RT advantages for subjects identifying a second target stimulus that was presented immediately after an initial foveal target stimulus. The targets were different in the two sequential parts of the task, and the subjects responded only when the assigned targets were present in both parts. GSM was not developed for, nor has it been applied to, such rapid sequential tasks, but it is clear that the combination task of LaBerge and Brown does involve a rapid shift of attention from foveal target identification (stimulus duration 150 ms) to RVF or LVF target identification (stimulus immediately following with a duration of 180 ms). Furthermore, the RVF advantage was not important for the purposes of the study and was not discussed in any detail by the authors. Nevertheless, as in the studies of Polich, there is nothing in the methods
Guided Search
185
or results of LaBerge & Brown to indicate that the RVF advantages cannot be due to a fight-left spatial asymmetry in the bottom-up stages of a mechanism like GSM. Previc and Blume (1993) and Previc (1996) have reported RVF advantages in both feature-search and conjunction-search tasks where the target was differentiated from the distractors by shape and/or size. These authors interpret their results in terms of attentional nonuniformities in extrapersonal space perception mechanisms. Although neither report specifically discusses GSM, both their results and their interpretation seem quite consistent with our own results and our explanation in terms of GSM. Kingstone, Enns, Mangun and Gazzaniga (1995) studied "guided search" in the RVF and LVF of three split-brain patients and a group of 10 normal controls. In the experiment, subjects searched for a blackcircle or gray-square target (in separate parts of the experiment) in the presence of varying numbers of black- square and gray-circle distractors. Either there were equal numbers of the two distractors (D= trials) or the number of gray circles was greater in the ratio of 5:2 (D~ trials). 9 D= trials (with 2, 4, 8, or 16 items) and D;~ trials (with 8 or 16 items) were randomly intermixed, as were unilateral- and bilateral-presentation modes. Normal-control subjects showed faster RTs to RVF than LVF targets and faster RTs in D~ than in D= trials for the 16-item displays. Two of three patients also showed faster RTs in D~ than in D= trials for the 16-item displays, but only for RVF targets. The authors conclude "that guided search is a left-hemisphere process" and that they have localized the top-down component of guided search, as opposed to the bottom-up component of guided search, because the pattern of results "do not fit the expected profile for a hardwired perceptual process". Discussing the results of Kingstone et al. (1995) in the context of this chapter is made difficult by the vast difference between the concept of GSM presented here and the sense in which the term "guided search" is used there. As presented here, GSM is a complex model with multiple bottom-up and top-down components, all of which can contribute to the signal levels in the activation map, which then determines the order of the attentional scan within one trial. Bottom-up components act automatically and in the same way on whatever stimuli are present: One red target "pops out" in a large group of green distractors, but two red items in a similar group of green distractors also stand out for the same reason, although perhaps not as much because there are two red items
186 Yund rather than only one. Bottom-up processing is, to some extent, spatially local and spatial nonuniformities in feature sensitivity also affect bottom-up signal levels. Top-down components are controlled according to information concerning the identity of targets and distractors. Top-down guidance can be applied to more than one feature at a time and increases in efficiency as a subjects gains experience with the stimuli, but cannot be adjusted rapidly enough to respond to the particular distribution of items present on an individual trial. In Kingstone, et al. (1995), "guided search" is a simple processselecting all items having a particular feature (black or square in the two parts of the experiment) which are then examined in random order. "Guided search" is occurring when asymmetrical distractor distributions yield faster RTs than symmetrical distractor distributions. Their discussion also seems to use two further assumptions about "guided search". First, the results must be due to either bottom-up or top-down processing and, second, top-down selection can act on only one feature dimension at one time. 10 Interpreting the results of Kingstone et a1.(1995) within the more complex context of GSM, we must first consider the different effects of bottom-up processing in the D~ and the D= trials: in D~ trials, bottomup feature contrast should put the target and (to an even greater extent) the minority distractors higher on the prioritized list than the majority distractors. This would occur without any change in top-down selection, merely because minority-distractor features (one of which is shared by the target) are less common in the display than they are in D= trials. As noted by Kingstone et al. (1995), we would not expect major differences in bottom-up processing between the RVF and LVF in normal or split-brain subjects, but then how can we account for the RVF/LVF performance difference on D~ and D= trials in the patients? It is critical to examine how both bottom-up and top-down processes affect the activation levels that determine the order in which the items are examined. When both bottom-up and top-down processes are considered, the expected results can be counterintuitive. For example, the most efficient strategy for the D~ trials in this experiment would be to concentrate the top-down selection on the feature the target shared with the majority distractors, and to rely primarily on the bottom-up contrasts to differentiate the target from the majority distractors. Such an optimal strategy would be one way to explain a virtually fiat slope like that seen for the D~ trials in the RVF of patient JW. This could occur because
Guided Search
187
bottom-up activation would select the target and the minority distractors while top-down control would select the target and the majority distractors, allowing only the target to stand out. 11 In comparison, concentrating top-down selection on the feature the target shared with the minority distractors would not help GSM to differentiate the target from minority distractors. The top-down selection would be needed most to differentiate the target from the minority distractors which would have even greater bottom-up activation than the target on D~ trials because minority distractors differ from the majority distractors in both features while the target differs only in one. Furthermore, any RT advantage resulting from the improved bottom-up contrast in the D~ trials relative to the D= trials would depend on the selection strategy, as well as the efficiency, of GSM's top-down selection. Given the best possible top-down selection in D= trials (the target is always the first item scanned), the RTs cannot be better in D;~ trials. Given the worst possible top-down selection in D= trials (the target's position on the list is random), the RTs should not improve in D~ trials because the primary effect will be to put minority distractors higher on the prioritized list and majority distractors lower on the list leaving the position of the target essentially unaffected. The advantage in D~ trials for intermediate capacities of top-down selection, as implied above in the description of the optimal strategy, will depend more on the efficiency of top-down selection for the feature the target shares with the majority distractors because that is the critical feature to discriminate the target form the minority distractors on the D~ trials. Thus, the results of Kingstone, et al. (1995) may indicate a different control strategy for the guided-search top-down selection process in the RVF compared to the LVF of the split-brain patient, rather than the presence of "guided search" in the RVF and its absence in the LVF. There are three experiments that have not found an RVF advantage in visual search and all three have used color as one (or the only) targetdefining feature. Arguin, Joanette and Cavanagh (1990) and Woods and Ogawa (1997) used color and orientation to differentiate the target bars from distractors. Efron and Yund (1996) used color alone (Experiment G) with multiple different-colored distractors. In addition to no significant field asymmetry, Woods and Ogawa (1997) found that RTs depended primarily on the number of target-colored distractors, suggesting that the color feature was the primary one used for "guided search" in their experiment. If the subjects in Arguin, Joanette and
188 Yund Cavanagh (1990) also used primarily color in their "guided search", then all three experiments are consistent with each other and indicate that there is no fight-left asymmetry for visual search guided by color. Given that color does not yield an RVF advantage in visual search, what is the nature of the stimulus features that do? Polich (1982, 1984) used printed letters in lines or square arrays to obtain RVF advantages, but lines or square arrays of letters are quite similar to the texture patterns we have used. The RVF advantages seen in the control subjects in Kingstone, Enns, Mangun and Gazzaniga (1995) and in LaBerge and Brown (1986), Previc and Blume (1993) and Previc (1996) also suggest that shape and/or orientation might be important. Furthermore, the entire texture patterns are not necessary for the RVF advantage in our experiments: Reducing the stimuli to single bright bars of the same size as one bar of our patterns, yielded an RVF advantage for both horizontal- and vertical-bar targets (Efron & Yund, 1996, Expt E). The single-bar results also showed the same RT-gradient shapes as those in Figure 1 for the horizontal-and vertical- pattern targets. The only clear difference between the single-bar and whole- pattern tasks was that the single-bar task was more difficult, probably due to the weaker orientation-specific signal of one bar in comparison to four cycles of a square-wave grating pattern. The relatively weak orientation signal for single bars also may explain why color was so dominant over orientation in visual search guidance in the study of Woods and Ogawa (1997). The idea that orientation-specific signals are critical for RVF advantages is consistent with recent results concerning the role of high spatial frequencies in RVF advantages (Christman, 1989; Christman, Kitterle & Hellige, 1991; Kitterle, Christman & Hellige, 1990; Kitterle, Hellige & Christman, 1992; Kitterle & Selig, 1991). The spatial frequency components carrying the critical orientation information in our experiments were at and above 3 cycles per degree (cpd). However, it is not clear why high spatial frequencies should favor one or the other visual half-field, because the visual world has no such obvious asymmetries and the value of vision is in its ability to extract useful information from the visual world. On the other hand, the magnitude of the right-left spatial nonuniformity, that GSM can amplify into a large visual-search RVF advantage, may be so small that it is otherwise unimportant for visual information processing in the fight and left visual fields. There is one visual activity within which the demands for visual information are not right-left symmetrical: reading. Rayner (1993)
Guided Search
189
describes how the reading process has been studied using the eyecontingent display change paradigm. In this paradigm, only the visual information immediately around the fixation is presented clearly (or accurately). Varying the size of the information displayed on either side of the fixation and measuring the effect on reading, yields an estimate of the size and asymmetry of the information window used by the reader. As summarized by Rayner (1993), "the span of effective vision extends about 14 to 15 letter spaces to the right of fixation[, but] no more than about 4 letters to the left of fixation for readers of English." Combining this fact about reading with the idea that reading makes primary use of high-frequency orientation-specific visual information, and remembering our location specific learning results, suggests a possible source for the RVF advantage: Perhaps the fight-left asymmetry is learned (conditioned, might be a better word) during reading. It must be admitted that this hypothesis is little more than idle speculation at this time, and that there are many objections that easily can be raised, but there are also two more bits of data in its favor. Our very first encounter with this visual search process (Efron, Yund & Nichols, 1987) involved a linear array of patterns with three on each side of the fixation point; we found an RVF advantage and evidence of a detectability gradient with the highest detectability just to the right of fixation and then decreasing to the extreme right, followed by the extreme left and then decreasing toward the fixation point in the LVF. This gradient is essentially the path of reading: to the right from the current fixation with a return to the left margin after reaching the right margin. In later experiments we avoided this linear arrangement of patterns with the fixation in the center in order to avoid this "confusion" with reading behavior and also to avoid the difficulties introduced by resolution differences at different distances from the fovea. Ostrosky-Solis, Efron & Yund (1991) directly tested the effect of reading on the visual search detectability gradient by comparing the performance of socioeconomic-matched groups of literate and illiterate subjects. Of the four groups in the experiment (male or female, readers or nonreaders), only the male nonreaders failed to show the usual RVF advantage although their overall target detection was as good or better than the other groups (as would be expected if they merely lacked a spatial nonuniformity that the other groups had). Incontrast, the female nonreaders did show the usual RVF advantage and thus the effect of
190 Yund literacy remained equivocal in spite of the very unusual lack of an RVF advantage in the male nonreaders. In addition to the lack of strong evidence for the hypothesis that the fight- left asymmetry in our results is conditioned by reading, there are other difficult questions that have to be answered. One such question is how an asymmetry conditioned during reading could extend over as large an area as studied in the present experiments. Of course, large amounts of the visual field contain text during reading, but we might expect any asymmetrical conditioning to be restricted to the immediate parafoveal area along the horizontal meridian, corresponding to the span of effective vision for reading (about 20 letter spaces predominantly to the fight of fixation). Other questions would be whether readers of languages not read from left to right show different asymmetries and whether the typical font in which different languages are printed affects the type of stimuli that show fight-left asymmetries. Short vertical and horizontal line segments seem to be dominant in modern English fonts, but what would occur for the reader of a language whose typical font was primarily curvilinear? Further work will be needed to answer these and other questions and/or to develop better hypotheses for the source of the RVF advantage in these visual search tasks. CONCLUSIONS The Guided Search Model provides a mechanism for amplifying virtually insignificant spatial nonuniformities into major spatial performance asymmetries when multiple stimuli are present. Investigators of visual-field/cerebral-hemisphere asymmetries must be especially careful with tasks that involve stimuli with high spatial frequencies carrying orientation information. Results that suggest major cerebral asymmetries may be caused by no more than minor spatial nonuniformities in the bottom-up stages of attention guidance. Notes
1. Other models of visual search (see Duncan & Humphreys, 1989; Hoffman, 1978, 1979; Treisman, 1988, 1993; Treisman & Gelade, 1980) include components similar to GSM to deal with these same results. Bundesen (1996) provides an extensive review of models of visual attention and visual search. GSM is used in the present context because of the ease of integrating the effects of spatial
Guided Search
191
nonuniformities into the structure of GSM, which clearly defines how bottom-up and top-down effects influence the signal levels in the activation map. 2. Elsewhere (Efron & Yund, 1996), we have called GSM with the added effects of spatial nonuniformities mGSM (for modified GSM). The addition of the spatial nonuniformities, however, is a very small modification of the basic model and, for this reason, I have dropped the mGSM designation for the purposes of this chapter. 3. At the present time the capacity of GSM's top-down guidance is not well understood, except that it must be limited because we know that we cannot always immediately find any target in a visual scene merely because it is uniquely defined in that scene by a known set of perceptual attributes. A limit in the effectiveness of the top-down selection mechanism, plus the competition of top-down guidance with differential bottom-up activation, may be all that is necessary to prevent the target from appearing at the top of the list on every trial of a typical visual-search task. Without spatial nonuniformities affecting bottom-up activation, it was necessary to include noise in the calculation of the activation levels (Cave & Wolfe, 1990). In effect, spatial nonuniformities may be one source of this noise. 4. Note that GSM provides an explanation for the difference in difficulty between single-feature and conjunction search without need of any special feature conjunction process, as hypothesized in Treisman's Feature Integration Theory (FIT). See Kim & Cave (1995) for evidence of similar attentional shifts in single-feature and conjunction searches. The question of the need for a feature-conjunction (or other feature-binding) process in explaining other aspects of cognition, is far beyond the scope of the present discussion of GSM and visual search nonuniformities. 5. It is not necessarily the case that no target features are defined, just because the task "does not involve visual search." The subject may use some general features of the stimuli in the top-down guidance mechanism merely to avoid distractions from the assigned task. Such a use of top-down guidance, however, should not produce spatially asymmetrical performance unless the top-down guidance mechanism itself is spatially asymmetrical. 6. There is no way to be sure that the stimuli near the meridian could not have been processed in the other hemisphere which had no other stimuli to process. In effect, the stimuli may not have been sufficiently isolated to one cerebral hemisphere to reveal an asymmetry that was present. 7. In informal debriefing at the end of the experiment, none of the 7 subjects demonstrated any awareness of the spatial segregation of the targets (even in the fight-left and up-down conditions). Furthermore, after completing the first session after the switch, several subjects made excuses for their subjectively poorer performance in that session (e.g. "out late last night", "not feeling completely well"). Another subject said that the stimuli "seemed shorter" and asked if we had
192 Yund changed their duration. It still seems surprising that none of the subjects noticed the spatial segregation of the two targets, but there was no reason for the subjects to take note of where particular targets appeared nor to integrate that information across trials to discover the pattern. It is true that we cannot rule out the possibility that some type of partial awareness shaped a subject's target-detection strategy, but a more precise partial-awarenessstrategy model would be needed before it could be tested against simple location-specific learning that was too coarse for independent learning to occur at adjacent locations in the alternating condition. 8. A further implication of this location-specific hypotheses was also tested by Yund & Efron (1996). It would predict that the RT step between sessions 0 and 1 should be greater for locations farther away from the dividing meridian than for locations closer to the dividing meridian. This expected difference was seen in both up-down and fight-left conditions, although it did not reach statistical significance in the up-down condition alone. 9. The authors call the D trials "standard-search" trials and the D trials "guidedsearch" trials. I have not used the authors' trial names because they might confuse us into thinking that the subjects were using a different topdown search strategy on the two trial types. Strategy switching would be impossible, however, because trial types were randomly intermixed with no cue to indicate which was coming. Furthermore, strategy switching is not necessary to produce, and might undermine the theoretical significance of, the expected effect of the manipulation of distractortype probabilities on guided search (Kaptein, Theeuwes, van der Heijden, 1995). 10. In their study of visual search guided by the feature color, Kaptein, Theeuwes & van der Heijden (1995) chose a difficult orientation discrimination as the feature to compete with a red-green color discrimination, specifically to avoid the possibility of subjects using the orientation feature for top-down guidance. If subjects had used the orientation feature for guidance, the observed effect of colorguidance would have been reduced or eliminated. In Kingstone, et al. (1995), however, the experimental design required that the subjects use one feature with one target and the other feature with the other target. 11. A subject would not have to understand GSM in order to adopt such a strategy. If a few distractors and the target "popped out" on several trials the subject might try to focus on the feature that differentiated the target from those more intrusive distractors and in this way set the top-down control to select for the feature shared by the target and the majority distractors.
Acknowledgments P r e p a r a t i o n of this c h a p t e r was s u p p o r t e d by the D e p a r t m e n t of Veterans Affairs and by N I N D S grant NS 32893. I am grateful to
Guided Search
193
Robert Efron, Dell Rhodes, and David L. Woods for many very helpful discussions on the critical issues of visual search and visual field asymmetries, and also for their excellent comments of an earlier version of this chapter. References
Arguin, M., Joanette, Y., & Cavanagh, P. (1990). Comparing the cerebral hemispheres on the speed of spatial shifts of attention: Evidence from serial search. Neuropsychologt'a, 28, 733-736 Buckles, K.M., Yund, E.W., & Efron, R. (1991). Visual detectability radients: Effect of high-speed visual experience. Brain & Cognition, 7, 52-63. Bundesen(1996). Extending "guided search": Why "guided search"
~
~!~!i'i~t~l~
v~2ar~s~ ~O0PeW~sK~
tnh, Me G~ :~Y~~
Y g Cave, K.R., & Wolfe, J.M.(1990). Modeling the role ofparallel processing in visual search. Cognitive Psychology, 22, 225-271. Christman, S. (1989). Perceptual characteristics in visual laterality research. Brain and Cognition, 11, 238-257. Christman, S., Kitterle, F.L., & Hellige, J. (1991). Hemispheric asymmetry in the processing of absolute versus relative spatial frequency. Brain and Cognition, 16, 62-73. Curcio, C.A., Sloan, K.R.,Jr., Packer, O., Hendrickson, A.E., & Kalina, R.E.(1987). Distribution of cones in human and monkey retina: Individual variability and radial asymmetry. Science, 236, 579-582. Duncan, J. & Humphreys, G.W.(1989). Visual search and stimulus similarity. Psychological Review, 96, 433-458. Efron, R. & Yund, E.W. (1996). Spatial nonuniformities in visual search. Brain and Cognition, 31, 331-368. Efron, R., Yund, E.W., & Nichols, D.R.(1987). Scanning the visual field without eye movements: A sex difference. Neuropsychologia, 25, 637-644, Efron, R., Yund, E.W., & Nichols, D.R.(1990a) Serial l~rocessing of visual spatial patterns in a search paradigm. Brain and c'ognition, 12, 17-41. Efron, R., Yund, E.W., & Nichols, D.R.(1990b). Detectability as a function of target location: Effects of spatial configuration. Brain and Cognition, 12, 102-116. Efron, R., Yund, E.W., & Nichols, D.R.(1990c). Visual detectability gradients: The effect of distractors in contralateral field. Brain and Cognition, 12, 128-143. Hoffman, J.E. (1978). Search through a sequentially presented visual display. Perception and Psychophysics, 2 3, 1-11. Hoffman, J.E. (1979). A two-stage model of visual search. Perception and Psychop_hysics, 2 5, 319,327. Kaptein, N.A., Theeuwes, J. & van der Heijden, A.H.C. (1995). Search for a conjunctively defined target can be selectively limited to a
194 Yund subset of elements. Journal of Experimental Psychology: Human Perception and Performance, 21, 1053-1069.
color-defined
Karni, A. & Sagi, D. (1991). Where practice makes perfect in texture discrimination: Evidence for primary visual cortex plasticity. Proceedings of the National Academy of Science, 88: 4966-4970. Karni, A. & Sagi, D. (1993). The time course of learning a visual skill. Nature, 365: 250-252. Kim, M.-S. & Cave, K. R. (1995). Spatial attention in visual search for features and feature conjunctions. Psychological Science, 6, 376380. Kitterle, F.L., Christman, S. & Hellige, J.B (1990) Hemispheric differences are found in the identification ~,ut not the detection, of low versus high spatial frequencies. Perception and Psychophysics, 48, 297-306. Kitterle, F.L & Selig, L.M. (1991). Visual field effects in the discrimination of sine- wave gratings. Perception and Psychophysics, 50, 15-18. LaBerge, D. & Brown, V. (1986). Variations in size of the visual field in which targets are presented: An attentional range effect. Perception and Psychophysics, 40, 188-200. Ostrosky-Solis, F., Efron, R., & Yund, E.W. (1991). Visual detectability gradients: Effect of illiteracy. Brain and Cognition, 17, 42-51. Polich, J. M. (1982). Hemispheric differences for visual search: Serial vs parallel processing revisited. Neuropsychologia, 30, 297-307. Pofich, J. M. (1984). Hemispheric patterns in visual search. Brain and Cognition, 3, 128-139. Previc, F. H. (1996). Attentional and oculomotor influences on visual field anisotroj~ies in visual search. Vision Cognition, 3, 277-301. Previc, F. H. & t~lume, J. L. (1993). Visual search asymmetries in threedimensional space. Vision Research, 33, 2697-2704. Rijsdijk,.J:P., Kroon, J.N., & van der Wildt, G.J.(1980). Contrast sensitivity as a function of position on the retina. Vision Research, 20, 235-241. Rovamo, J., Virsu, V., Laurinen, P., & Hyv~irinen, L. (1982). Resolution of gratings oriented along and across meridians in peripheral vision. Investigative Ophthalmology and Visual Science, 23, 666-670. Scobey, R.R., & van Kan, P.L.E. (1991). A horizontal stripe of displacement sensitivity in the human visual field. Vision Research, 31, 99-109. Shiu, L-P & Pashler, H. (1992). Improvement in line orientation discrimination is retinally local but dependent on cognitive set. Perception & Psychophysics, 52: 582-588. Temme, L.A., Malcus, L. & Noell, W.K. (1985). Peripheral visual field is radially organized. American Journal o f Optometry and
Physiological Optics, 62, 545- 554.
Toet, A. & Levi, D.M. (1992). The two-dimensional shape of spatial interaction zones in the parafovea. Vision Research, 32, 1349-1357. Treisman, A. (1988). Features and objects: The Fourteenth Bartlett Memorial Lecture. Quarterly Journal of Experimental Psychology, 40A, 201-237.
Guided Search
195
Treisman, A. (1993). The perception of features and objects. In A. Baddeley & L.Weiskrantz (Eds.), Attention: Selection, awareness, and control, pp. 5-35. Oxford, England: Clarendon Press. Treisman, A & Gelade, G. (1980). A feature integration theory of attention. Cognitive Psychology, 12, 97-136. Treisman, A & Gormican, S. (1988). Feature analysis in early vision: Evidence from search asymmetries. Psychological Review, 95, 15-48. Treisman, A & Souther, J. (1985). Search asymmetry: A diagnostic for preattentive p r o c e s s i n g of separable features. Journal of Experimental Psychology: General, 114, 285-310. Wolfe, J. M. (1994). "Guided search" 2.0: A revised model of visual search. Psychonomic Bulletin & Review, 1, 202-238. Wolfe, J. M. (1996). Extending "guided search": Why "guided search" needs a preattentive "item map". In A. F. Kramer, M. G. H. Coles, & G. D. Logan (Eds.), Converging Operations in the Study of Visual Selective Attention, pp. 247-270. Washington, DC" American Psychological Association. Wolfe, J. M., Cave, K.R., & Franzel, S.L. (1989). "Guided Search"" An alternative to the feature integration model for visual search. Journal
of Experimental Psychology: Human Perception and Performance, 15, 419-433. Woods, D. L. & Ogawa, K. (1997). Factors guiding displacement of attention in visual search. (Submitted). Yund, E.W., & Efron, R., (1996). "guided search": The effects of learning. Brain and Cognition, 31,369-386. Yund, E.W., Efron, R., & Nichols, D.R.(1990a). Detectability gradients as a function of target location. Brain and Cognition, 12, 128-143. Yund, E.W., Efron, R., & Nichols, D.R.(1990b). Detectability as a function of spatial location: Effects of selective attention. Brain and Cognition, 12, 42- 54. Yund,-E.W., Efron, R., & Nichols, D.R.(1990c). Target detection in one visual field in the presence or absence of stimuli in the contralateral field by_ right- and left-handed subjects. Brain and Cognition, 12, 117-127.
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
197
Chapter 7
Hemispheric Coordination of Spatial Attention James T. Enns University of British Columbia & Alan Kingstone University of Alberta It is now widely accepted that the control of spatial attention in human vision is governed by a distributed network of brain regions (Posner & Raichle, 1993). This chapter is a progress report on our research examining the ways in which the two cerebral hemispheres coordinate their activities in two behavioral tasks designed to index spatial attention: visual search and object identification. Our strategy has been two-pronged: (1) We have studied the performance of healthy adult observers viewing displays presented to one or the other visual field, and (2) We have tested observers who had their corpus callosums surgically sectioned, thereby preventing the normal neural communication that occurs between hemispheres. In a first series of experiments, we asked whether the two hemispheres showed signs of being specialized for visual search based on particular kinds of visual attributes. For example, would the left hemisphere be better able to search for targets defined at a local level of
198
Enns and Kingstone
detail, and the right hemisphere be more adept at search for targets defined by global features, as one might expect based on previous studies of hemispheric specialization? The answer was surprising to us. We found very little evidence for local vs. global specialization in the visual search task. This suggested to us that the control of attention might not be as lateralized as had been assumed. We next considered whether each hemisphere might be able to guide search independently. Although some earlier reports had indicated that split-brain observers (surgical callosotomy) were able to search more rapidly when items were displayed in both visual fields than when the same items were displayed in a single field, we found that this was not generally true. Instead, we found that the two cerebral hemispheres work together in performing visual search tasks, both in split-brain observers, and in individuals whose corpus callosal connections are intact. Finally, we explored the generality of the claim that there is a subcortical system that is shared by the two hemispheres in tasks involving spatial attention. We used an object identification task under conditions of divided vs. focused attention. We found a similar twoobject cost for items presented in the same visual field as for items presented in different visual fields in normal observers. However, a splitbrain observer showed strong evidence of hemispheric competition when two items were simultaneously displayed in separate visual fields. These findings are discussed with reference to four questions: (1) Why do normal observers demonstrate hemispheric competition when performing visual search? (2) Why does a split-brain observer show hemispheric competition when performing visual search and item identification? (3) What are implications for understanding spatial attention? and (4) What are the implications for understanding hemispheric specialization?
Hemispheric specialization in visual search? One of the most widely-cited examples of hemispheric specialization is the association of the right hemisphere (left visual field) with global visual processing and the association of the left hemisphere (fight visual field) with local visual processing (e.g., Kitterle, Christman, & Conesa, 1993; Lamb, Robertson, & Knight, 1990; Robertson & Lamb, 1991). The prevailing view is that the right hemisphere is specialized to perform a rapid and low-pass spatial frequency analysis of the visual image,
Hemispheric Coordination
199
whereas the left hemisphere is biased in favor of a slower, high-spatial frequency analysis (Badcock, Whitworth, Badcock, & Lovegrove, 1990; Hughes, Fendrich, & Reuter-Lorenz, 1990). In a first experiment we asked whether local-global specialization would be evident in a visual search task involving target items that could differ from distractor items at either a local or a global level of perceptual structure. Visual search has proven to be a simple and powerful tool for the study of spatial attention. In this task, observers try to indicate rapidly and accurately whether a prespecified target item is present in a display. The critical factor is the number of other (non-target) items that are also present. Because search time tends to increase linearly with this factor, two different measures can be derived: a baseline measure reflecting the sensory, decision, and response components of the task and a slope measure reflecting the incremental cost associated with selecting the target from a larger set of candidate items. Theories of visual search link the baseline measure to an early preattentive stage of processing, in which simple features are registered automatically and in parallel across the visual field (Treisman & Gelade, 1980; Duncan & Humphreys, 1989; Wolfe, Cave, & Franzel, 1989). They also link the slope measure to an attentive stage, in which the spatial relations between visual features are determined by slower, but more adaptive, processes. Theories vary considerably in their proposals for specific attention mechanisms (e.g., serial scanning by a spotlight vs. limited-capacity parallel operations), but they all share the assumption that the search slope reflects the relative involvement of attention. We used the search items and displays shown in Figure 1. These stimuli were prompted by observations that there are a number of factors contributing to whether a local or global target will be at an advantage (Kinchla & Wolfe, 1979; Lamb & Robertson, 1990; Martin, 1979; Navon, 1983) and that search for a target at either level is attentiondemanding only if some amount of element grouping is required (Enns & Kingstone, 1995). Therefore, we designed stimuli in which the degree of grouping at either level could be varied systematically. Elements of the same sign of luminance contrast with respect to the background group quite naturally and effortlessly; elements of opposite contrast group only with considerable effort. The standard search task was modified only slightly to permit an examination of visual field effects. Observers indicated with one of two response keys (index fingers of the left and fight hands) whether the tar-
l-ens!A goo I sns.~aA leqOla ~q pau!jap slaBarel aoj qo.mas lenS.tA u! uo.tlez!le!oads o.uaqds!maq aoj lsal ol pasn Ss qoaeas lenS!A jo saldmex H "I aantl!d
la~ae~L leq o l o _
!]!i!i!i!•!i!i!i!i!•!i!•!i!i!i!•!i!i!]!•!i!•!i!i!i!i!i!i!i!i!i!i!i!i!•!i!i!•!•!•!•!i!J!•!i!•!i•iii•i!i!i!i!i!i!i!i!i!i!i!i!i!i!•!•!•!i!•!•!ii!i!•!i!ii!iiii!•i•i•i•!•i•ii!i!i!]ii!iiii•!ii!iiii!i!ii•i ~i~!~i~!~i~i~i~i~i~i~!~!~i~i~!~i~i~!~i~i~i~!~i~i~i~!~i~i~i~i~i~i~i~i~i~!~i~i~i~i~!~i~i~i~i~i~i~i~i~i~i~!~i~i~i~i~i~i~i~i~;ii!i!i!i!i!i!i!i!!!i!i!!!!!!!~!!i!i!~!!!i!i!i!i!!!!!!!i!!!~!~!i!!!i!i!i!i!i!i!;i!i!~!~i~i~!!i!i!
•i•i•i•i•i•i•i•i;i•i•i•i•i•!i•!i•C i!•;i•i;•!i•!i•i•i•i•i•i•i :!:i:i:i:!:i8 i:i..~lg..i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i.~lg. i:i:i:i:i:i:i:!:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:!:i:!:!:!:!:i:i:!:i:i:!Si:i:i:i:!:i:i:i:!!!:i:i!i:i!i!i!i!i!i!i!i!i!i!i!i!i!i!!!i!i!i!i~i!i!!!i~i!!!i iili!i!i!!!i!i!i!i!!!i!i!i.~..! !i!i!i!i!i!i!i!i!i!!!!!i!i!i!i!i!i!i!i!i!i!i!i!!!!!i!i!
iii!iii{ii{i{i::::: iiii i iiiiiiiiiiiiiiii!i!{ iiiiiiii;;:::: iiiii i iiii{iiiiiiiiiiiiiiiiiiiii!iiiiiiiiiii iiiiiii!i! i iiiiiiiii i iiiiiiiiiiiiiii la:Sae.L leOOq
!i!!~:~.ii!iii i!ii!:.i.:~:.iii:.i!i!i iiiiiiiiiiiiiiiiiiiii!i!iiiiiii!iiiiiiiiiii!iiiiii!iiiiiiiiiiiiiiiiiiiiiiiiiii{iii!iii!iiiiiii! :.:..:.:.~:..:.:.::.:.:.:.:..~:::.:.:.:.:.:.:.:.:.:.:!.:i:i:i:i:i:i:: ~.~.~-..~.~-~.``.-.~...-.~.~-.....-~..~.-~v~-.....~.....-...........~....~.....~.....~.................~..~..-.:.'~...-...........-...~.~...~...~...~.....-....-.....~.....~....~-.....~....
- . " -. . . . . .
~.~..
................................................
.... ::::::::::i:i:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::. :::::::::::::::::::::::::::.::.:.:.:.:.:?:.:._:.:._:._:.:.:.:~_~...~:: : : : : : : : : : : :::::::::::::::::::::::::::::::::::!:::!:!::8!:!8!:. .:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:!:~'~
iiiiiii!iiiiii! i!iiiii!i i!iiiiiii !i!i!i!i!i!ii!i !iii !iiii{!!!!i!i i!ii!iiiiiii!ii{i!!i!ii!! ii! ! !ii!!i!i! !iiiiiiiiii!iiiiiii!iiiiiii!iii i~i~i~!~i~i~i~i~i~i~!~i~!~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~!~i~i~i~i~i~i~i~i~i~i~i~i~i~i~!~i~i~i~i~i~i~i~i~::~i~i~!~i!i~i!i!i!i!~!i!~!~!~!!~!~!~!~!i!i!i!~i!i!i~i~i~i~:~:d!i!i! }ii{{•!ii{i{{{iii{{{•ii•i{i!i•i{i!!{{•{i{•••i•iii!•••{{i•giii{{{i!i•i•{ii{i{{i{{i{{{{{{{{{{{•{ii•••i!iiii•iiiii•i{iii!i•i{i{iiii{i{{{{•iii i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i{i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i~i
qaaeaS PaeH IeqOlD aUOlSl~U!~ pu~ suuq
OOE
Hemispheric Coordination
201
structure. In each example, a total of 10 dot-cluster items are presented, with half of the items in each visual field. The target item contains an oblique arrangement of dots, with the oblique signal coming either from the local level (smallest dot pair) or the global level (dot pair groupings). The oblique arrangement of dots in the target could also involve grouping over different signs of luminance contrast (hard search) or grouping within the same sign of luminance contrast (easy search).
get was entirely absent, no response was made. Ten college students served as observers, each contributing 360 trials. Results are shown in Figure 2. Aside from the expected effects of display size and target discriminability, search was somewhat easier for difficult local targets in the fight visual field (left hemisphere) than in the left visual field (fight hemisphere): RT slopes were 38 ms and 44 ms per item, respectively, and this difference was significant. Conversely, search was slightly easier for global targets in the left visual field than in the right visual field: RT slopes were 22 ms and 32 ms per item, respectively, and this difference was also significant. Thus, although we obtained some evidence of hemispheric specialization for local and global targets in a visual search task, this evidence was not very impressive for a number of reasons. First, it was spotty, in that the effects were not seen in all conditions. Second, the effects seemed to be entirely due to a non-linearity in the search slopes between display sizes l0 and 18. This suggested that a closer examination of the response speed versus accuracy trading relation might be in order. Third, the obtained hemisphere differences were not very large in any practical sense, involving RT slope differences of only 6-l0 ms per item. Finally, to confirm these findings we repeated the experiment using different global-local stimuli and replicated the same pattern of results (Enns & Kingstone, 1995). To see whether these differences might not be more pronounced when the two hemispheres were unable to communicate directly with each other, we conducted the same experiments with two split-brain observers (JW and V J). As shown in Figure 3, these data were on the whole remarkably similar to that of the observers with intact corpus callosums. There were no consistent visual field differences in search for either the local or global targets. In order to be fully convinced that the expected left hemispherelocal advantage and fight hemisphere-global advantage was not strongly evident in split-brain observers performing this search task, we repeated
202
Enns and Kingstone
Local Hard Search 1400
Global Hard Search Local Target
1300
OLeft Visual Field Right Visual FieldJ
P-- 1200 r.4.--" 1100 r
a> 1000 o O c
Global Target
900
Global Target
800
700
Local Target
6O0 500 400
a
'
a
I
I
I
2
10
18
2
10
18
Display Size Figure 2. Results from normal college-age observers searching for target items defined at either the local and global level of visual structure.
the experiment, but now with an emphasis on response accuracy following a brief display of the items. One advantage of a brief display is that there is no opportunity for the observer to make an eye movement while the display is visible. This increased our confidence that each visual field was stimulated uniquely. A second advantage was that the accuracy emphasis reduced our concerns about speed-accuracy considerations influencing the results. Observers were no longer under any speed pressure and so only response accuracy was at stake. In one session of this experiment, display duration was set to 200 ms and JW contributed 240 trials. In two further sessions, display duration varied randomly between 100 and 200 ms and JW contributed 480 trials. VJ was then tested under these same conditions for 240 trials. The results are shown in Figure 4. As expected, search became less accurate
Hemispheric Coordination
Local Hard Search
203
Global Hard Search
2200 -
Local Target
2000
Left Visual Field Right Visual Fiel
I-IT" 1800 r,_) 1600 (1) 01400
Global Target
~
9
C1200
Global Target
(1) 1000
Local Target.
80O 600
T
I
I
2
10
18
2
10
18
Display Size Figure 3. Results from two split-brain observers (JW and VJ) searching for target items defined at either the local and global level of visual structure. with display size (slope was now expressed as a decrease in accuracy) and search was more difficult when we deliberately made the target less discriminable at either the local or global level. However, the only visual field effect concerned steeper search slopes for low-discriminability targets in the left visual field (local targets in the local-hard condition, global targets in the global-hard conditions). This suggests that visual search under these conditions is generally more efficient in split-brain observers when items are presented to the left hemisphere, It is not consistent with hemispheric specialization for local or global features.
204
Enns and Kingstone
Global Hard Search
Local Hard Search
Local Target
90 ro (1) ~- 80 O
9
(1) 03 41---I
t(l) ro
(1)
Global Target
Global Target
70
60
Local Target ~~ Left Visual Field Right Visual Field)
50
[2_ 40 30
I
I
I
I
I
I
2
10
18
2
10
18
Display Size 4. Results from two split-brain observers (JW and VJ) in the local-global search task performed under brief exposure conditions.
Figure
These visual search experiments were designed to provide every opportunity for hemisphere biases to be revealed in visual search. The overall difficulty of the search task was varied over a large range, from local and global targets that popped out of the displays, to local and global targets that required serial and effortful scrutiny. Nonetheless, evidence of hemispheric specialization was difficult to find. In observers with intact connections between hemispheres, there was a small tendency for the left hemisphere to favor local detail and the fight hemisphere to favor global features. However, in observers with split-brains, where
Hemispheric Coordination
205
global-local specialization should have been even easier to see, none could be found. Instead, we observed a consistent bias favoring the left hemisphere in visual search for targets of both kinds. Why was specialization so hard to find in visual search task, when it seemed so much easier to find in tasks involving the identification of shapes at either a local or global level (Kitterle, Christman, & Conesa, 1993; Lamb, Robertson, & Knight, 1990; Robertson & Lamb, 1991)? The possibility we considered next was that the task of visual search, performed by either hemisphere alone, or by each hemisphere concurrently, draws on a common system for spatial attention. Hemifield differences in unilateral vs. bilateral visual displays
Previous research has shown that split-brain patients search through visual displays twice as fast as normal observers when items are divided evenly between visual fields, as though each disconnected hemisphere possessed its own attention system for performing visual search (Luck, Hillyard, Mangun, & Gazzaniga, 1989, 1994). In these studies, the range of display sizes tested was relatively small (2, 4, and 8 items). This meant that each hemisphere was presented with either 2, 4 or 8 items on unilateral displays, while on bilateral displays each hemisphere received only 1, 2, or 4 items. Visual search involving such small display sizes must be considered in light of the evidence that the unit of serial search is not necessarily the individual item (Pashler, 1987; Treisman, 1982). In particular, the unit for serial search seems to vary with the discriminability of the target. For example, if the target differs from distractors on the basis of a highlydiscriminable feature, but the items in the display are clustered into groups, then search increases linearly with the number of clusters, not the total number of items (Treisman, 1982). In another study, a finegrained analysis of visual search in display sizes ranging from 2 to 24 items indicated that the 2:1 slope ratio for negative and positive responses, traditionally taken as evidence for a serial self-terminating search, held only for display sizes larger than 8 (Pashler, 1987). For display sizes between 2 and 8, the slope ratio was close to 1: 1. This suggests that search may be serial over clusters of items, the size of the cluster being determined by the discriminability of the target. These considerations suggest that the rapid search of split-brain observers on small bilateral displays reflects hemispheric independence
206
Enns and Kingstone
in preattentive rather than attention-demanding processes. At the very least, serial search may have occurred only in some of the conditions. For example, assume that a hemisphere must initiate serial search when 4 or more items are presented in an experiment involving 2, 4, or 8 items. Unilateral displays will then initiate serial search on 2/3 of the trials (display sizes 4 and 8). In contrast, serial search will be required on bilateral displays only 1/3 of the time (display size 8). It is therefore possible that split-brain observers in the Luck et al (1989, 1994) studies only appeared to search more rapidly through bilateral displays, since only one bilateral, but two unilateral, displays demanded a serial search. To test this question, we repeated the Luck et al. (1989, 1994) search task but tested display sizes from 2 to 24. Our initial question was whether the 2:1 ratio of search slopes for bilateral versus unilateral displays would hold up even when the search task unquestionably required some degree of serial search. The results proved to be much more interesting that we had imagined. The search items were similar to those of Luck et al (1989, 1994) and are shown in Figures 5a and 5b. Displays consisted of between 2 and 24 items, presented either in one visual field (unilateral) or divided equally between fields (bilateral). Only one target could appear on any trial; display size and field of presentation were randomly intermixed within a block of trials. The observer's task was to press a left-hand key when the target was presented in the left visual field (fight hemisphere) and to press a fight-hand key when it was presented in the right visual field (left hemisphere). When the target was not presented (20% of trials) observers were instructed to withhold any response: a keypress was therefore an error. Search displays remained visible until the subject responded or until 1800 ms had passed. To ensure that eye movements did not play a role, gaze was directed to a central fixation point and eye position was monitored by a zoom-lens video camera. Observers were tested on 560 trials in two conditions varying in level of difficulty: all items oriented vertically (easy) or all items oriented horizontally (hard). The first experiment in this series tested 10 college-age observers with intact corpus callosums. The results for correct RT are shown in Figure 6. As expected, RT increased with display size and search slopes were steeper for the horizontally-oriented than for the verticallyoriented items. The finding of greatest importance for the present question, however, was that visual field had a much stronger influence on search difficulty in bilateral than in unilateral displays.
Hemispheric Coordination
207
Unilateral Display Figure 5a. Examples of unilateral display used to test the hypothesis that hemisphere is independently capable of visual search for a target defined solely by a spatial relation. In each example display, a total of 16 rectangular items are presented, each consisting of an abutting white and black square. The 15 distractor items have white tops, whereas the target item has a black-top. In unilateral search displays, all items are presented in either the left or the fight visual field.
Consider first the unilateral displays: search time increased with display size at about the same rate for items presented exclusively to the left or fight visual field (12 ms and 10 ms per item respectively). Now compare search times in the bilateral displays. For fight-field targets the search slope was consistently around 10 ms per item. However, for leftfield targets, the search slope was shallow for display sizes 2 to 8 (6 ms per item) and then steeper for larger displays (14 ms per item). Search
208
Enns and Kingstone
B ilateral Di splay Figure 5b. Example of bilateral display. In bilateral displays, the items are evenly divided between the two fields.
through the bilateral displays was therefore much easier in the left visual field, especially in the smaller display sizes. In the larger display sizes, this difference disappeared. This suggests at a minimum that visual search is not performed in the same way when items are presented to one versus two hemispheres. More interestingly, it suggests that the two hemispheres are in competition for resources when both are stimulated. If so, the results in the present search task indicate that the right hemisphere won this competition, at least for small display sizes, permitting search to be conducted more efficiently for targets on the left side of the display.
Hemispheric Coordination
209
Aside from this interaction with visual field of presentation, there was very little difference in search for unilateral vs. bilateral displays. The ratio in search slopes between these conditions overall was very close to 1:1 for both the smaller and larger displays. Search times increased approximately 16 ms per item between 2 and 8 items and 8 ms per item between 12 and 24 items. A split-brain observer (JW)was tested in the same way for a total of 1200 trials distributed over 4 sessions. The results are shown in Figure 7. Unlike the normal observers, there was no interaction between visual field and whether one or both hemispheres were stimulated. Overall, targets in the left field were detected earlier than those in the fight field (mean difference = 64 ms), but the search slopes did not differ reliably with visual field or type of display (range = 17-20 ms per item). This indicates that the hemispheric competition seen in normal observers searching through bilateral displays does not occur when the corpus callosum is sectioned. When the data for JW were averaged over visual field, there was a tendency for search to be more rapid on bilateral than unilateral displays, but only in the range from 2 to 8 items. The mean search slopes over these display sizes were 42 ms per item for unilateral and 25 ms per item for bilateral displays. This benefit for search through bilateral displays for a split brain observer is similar to the one reported by Luck et al (1989, 1994). However, when search slopes were compared on the larger display sizes (8-24 items), this difference disappeared, with the unilateral displays now yielding slopes that were somewhat more shallow (13 ms per item) than the bilateral displays (18 ms per item). As we had done in the global-local visual search study, we repeated these experiments with brief displays, both to eliminate any role of unintended eye movements, and to minimize concerns over speedaccuracy trading relations by emphasizing only response accuracy. The exposure duration of the search displays was therefore set to 200 ms. The results for observers with intact corpus callosums are shown in Figure 8. The interaction seen in the RT experiment was replicated here. On unilateral displays, accuracy decreased with display size at about the same rate for left (1.6% per item) and right visual field displays (1.8% per item). However, on bilateral displays the accuracy for left field targets was consistently higher. The fight field accuracy slope decreased by only 0.1% per item until display size 8, and then decreased by 2.3% per item; the left field slope decreased by about 2.1% per item over the
2 l0
Enns and Kingstone
1 200
~
1100
~
1000
Unilateral Displays
"
Bilateral Displays
(.~ 900 800 (1)
:~ 700 600
t
I
I
2 4
I
I
I
I
I
6
6 8 12 16 24
I
I
I
I
8 12 16 24
100 -
15
(1) O
(_>
90 8O
(D
o') t~
70
cO O
60
4.--I
O
n
o L e f t Visual Field Right Visual Field~
500
i
l
2 4
i
l
i
i
i
6 8 12 16 24
i
i
i
2
4
6
i
i
i
i
8 12 16 24
Display Size F i g u r e 6. Results from normal college-age observers in the unilateral versus bilateral visual search task.
r~
ira.
~=.,.
-~
o
N
n
O~
CO
O~
-I~
P~
P~
0
I
0
I
l
I
0 I
-rl
--
~"
"1"I
r= 9
I
0 I
0 I
I
0 I
Percentage Correct I
0 I
f,~
C~
.ib
r'~
.==i
O0
r,J
,==,=5
O0
c~
r~
0 0 I
I
0 0 I
'
I
0 0 I
0 0 I
I
I
0 0 I
0 0 I
I
Mean Correct RT 0 0 I
"
I
0')
--
c~~
l
0 0
t,,.1
0
i,=.~ ~
~
l,.=J 9
,...-
(1)
e
o
~,,~~
f~
9 ~.
r~
=~=o
o
~
oo
9
~
0
-I~
0
I
'
0 I
O~ "
0 I
"q '
0 I
CO "
0 I
c.~
Percentage Correct '
I
-"
0
0
O0
O~
Po
O0
0
0
=
O~
'
I
0
0
~
'
I
0
0
O0
'
I
0
0
~
'
I
0
0
0
'
I
0
0
"-~
Mean Correct RT '
I~O
--(I)
~m__
I
0
0
0
Hemispheric Coordination
Bilateral Displays
Unilateral Displays
1 200 1100
213
l-rr
1000
o ~- 900 0
o
800 - 700 600 500
I
I
I
2 4
I
.,I
I
I
6 8 12 162_4
I
I
I
I
2
4
6
I
I
I
8 12 16 24
90
'~176 f 80
70 6o c
50 ~
~
40
~_0 o 30 s 20 10
[~0__ Left Visual Field "~ Right Visual Field~ I
' 2 4
I
I
I
i
I
6 8 12 16 24
I
I
2 4
i
6
I
I
i
I
8 12 16 24
Display Size Figure 9. Results from a split-brain observer (JW) in the brief exposure version of the unilateral versus bilateral search task.
214
Enns and Kingstone
whole range. In comparison with unilateral displays, accuracy on bilateral displays was better for left field targets and worse for right field targets. This confirms that the two hemispheres were competing for something they both needed when searching in bilateral displays, even when displays were brief and response time was not at issue. As in the response time version of the task, the right hemisphere won the competition. The results for a split brain observer (JW) are shown in Figure 9. There are two differences from the data of normal observers that are immediately apparent. First, the interaction of hemifield and type of display seen for normal observers was not evident here. The left hemisphere showed a tendency for more accurate search in both unilateral and bilateral displays, especially as display size was increased. This replicated the RT task in showing that the competition seen normally between hemispheres during visual search on bilateral displays does not occur when the corpus callosum is sectioned. A second difference from the search data of normal observers was the markedly steeper search slopes in bilateral than in unilateral displays. Over the small display sizes (2 to 6 items) the search slopes were similar (5% per item for unilateral and 7% per item for bilateral). For larger display sizes, however, the search slope was almost flat for unilateral displays (0% per item), yet it decreased 1.5% per item for bilateral displays. This finding points to an important competition between the hemispheres of a split brain observer that is not evident in normal observers. Note too that this finding runs completely counter to the implication derived from Luck et al (1989, 1994), and replicated in our RT task for small display sizes (Figure 8), that the separated hemispheres of split brain observers have independent systems for visual search. On the contrary, these data indicate that the separated hemispheres of split brain observers draw on a common system of attention in a visual search task. These experiments therefore point to two different aspects of hemispheric coordination in the visual search task. First, there is an apparent competition between the hemispheres for visual search mechanisms that is only evident when the connections between hemispheres are intact. Faced with bilateral displays, the right hemisphere searched more efficiently and the left hemisphere less efficiently than they did when presented with the same items in a unilateral display. Disconnecting the hemispheres by sectioning the
Hemispheric Coordination
215
corpus callosum eliminated this form of competition, allowing visual search in each hemisphere of the split brain observer to be largely unaffected by these display differences. Second, there is a competition between the hemispheres for subcortical mechanisms necessary for visual search. This competition can only be seen following callosotomy. It is glaringly apparent in visual search performed by split brain observers when large display sizes are compared under unilateral and bilateral conditions. The split brain observer become very inaccurate when searching through large numbers of items that were distributed across both visual fields. Note that the large number of items are not at issue. Search by JW through the same number of items was much more accurate when the items were presented to only one hemisphere. In fact, in these displays each hemisphere was actually viewing twice the number of items, yet performing more accurately. The brief exposure duration was also not at issue for a similar reason: accuracy was much better for the same number of items, provided they were all presented in the same visual field. Finally, these results could not be explained by any simple form of hemispheric specialization. JW showed the same tendency for a left brain advantage when searching in either unilateral or bilateral displays. We were therefore compelled to conclude that there must be subcortical mechanisms, important to visual search, that are in demand by each hemisphere in the split brain observer. When the corpus callosum is intact, these mechanisms are presumably used by both hemispheres in a coordinated fashion. Any competition is resolved before it has the opportunity to express itself in behavior. It is only when the hemispheres lose the ability to communicate directly, that an overt competition for these mechanisms is observed.
Hemifield competition in object identification How general is the finding that the hemispheres compete for subcortical structures in tasks involving spatial attention? We addressed this question with a divided attention task that was much simpler than visual search (Duncan, 1980, 1984, in press). In this task, only one or two items are presented briefly on each trial and the observer is asked to identify one of them by indicating whether or not it corresponds to a subsequent probe item. The general result is that accuracy in this task is consistently lower (and correct response times are slower) when two
216
Enns and Kingstone
items have been presented than when only one is shown. This effect is often referred to as a two-object cost. In the present study we asked whether the two-object cost was any different for unilateral and bilateral visual displays. Examples of the stimulus displays are shown in Figure 10. Each trial began with a view of four sets of location markers, at the comers of an imaginary square 3 degrees from fixation, showing where display items would be presented. Four shapes (vertical or horizontal bars) were then flashed briefly (60, 100, or 150 ms). One or two of the shapes were black, the remainder were white. Observers were told to attend only to the black shapes and were asked to indicate the shape of one of them immediately following a brief mask (180 ms) at each of the four locations. The probe, a black shape that was either the same or opposite to the target, stayed on view until the observer responded or until 2 s had elapsed. Observers pressed a left-hand key if a match was detected on the left side of the display and they pressed a fight-hand key if a match was detected on the fight side. Displays were evenly divided between one-target, two-target unilateral (targets in the same field), and two-target bilateral (targets in different fields). The key factor in this experiment was whether the two targets were presented unilaterally or bilaterally. If the two connected hemispheres could each perform object identification separately, then the two-object cost should be reduced or eliminated in normal observers viewing bilateral displays. If each of the hemispheres were capable of identifying the items independently, but only when direct connections between hemispheres were severed, then the two-object cost should be reduced in split brain observers viewing bilateral displays. This would be consistent with the proposal of independent attention systems (Luck et al., 1989, 1994). Finally, if the two hemispheres drew on shared subcortical structures, there should be an increased two-object cost in one hemisphere in the split brain observer (in the hemisphere losing the competition), along with a reduced two-object cost in his other hemisphere (in the hemisphere winning the competition). This would be consistent with the proposal of a shared subcortical system as seen in the previous visual search study. Results for two groups of normal college-age observers are shown in Figure 11. Each observer contributed 320 trials. One group was tested with a stronger mask than the other group (random black and white shapes) in order to examine the accuracy pattern at two different levels
Hemispheric Coordination
9
9
9
9
i
217
9
9
i))))))))J)))!i)J!ii!!)i)ili /, -i))))i))))))))i) 9
9
Probe
mm
m r----Im
Mask (180 ms)
9
9
9
,,m
mm
9 Display(60-150 ms)
Preview ("/"00 ms)
/
Figure 10. Examples of displays used to test for hemispheric independence of object identification. The target items (black rectangles) could appear in only one location (one object), in two location in the same visual field (two objects--same side), and in two locations in opposite visual fields (two objects--different side). Remaining locations in the display consisted of white rectangles randomly oriented vertically or horizontally.
of difficulty. This had no effect on the results other than changing the overall level of task difficulty. In each case, a two-object cost was obtained to approximately the s a m e degree in both unilateral and bilateral displays. The visual field of the item to be reported had no additional influence on accuracy. The results for JW are shown in Figure 12, averaged over 640 trials of testing. Averaged over visual field, the pattern of results was identical to that for the normal observers: a similar two-object cost was observed
218
Enns and Kingstone
Strong Mask
Weak Mask 100 o 0
(~
90
Left Visual Right V'~sual
~. 80 e--
~
70
~" 60 50
One Object
Two Object Two Object Same Different
One Object
Two Object Two Object Same Different
One Object
Two Object Two Object Same Different
One Object
Two Object Two Object Same Different
1000 900 EE 8O0
t~ 700
0 0 ~- 6o0 ~: 500
Figure 11. Results from normal college-age observers in the unilateral versus bilateral object identification task.
in unilateral and bilateral displays. However, there was also a large interaction with visual field. The fight hemisphere (left field) showed no significant two-object cost on bilateral displays, whereas the left h e m i s p h e r e (right field) showed a two-obJect cost that was approximately twice as large on bilateral than on unilateral displays. This finding is consistent with the proposal of a shared subcortical system for spatial attention tasks.
Hemispheric Coordination 100
~~
o 90 O
o 80 r "-' 70 ~ 9 4, 60 IX. .
1
Left Visual Right Visual
1000
)
E
219
I-tr 900 O
[ 8oo
L_ 0
--7O0
6OO
50
5oo
One Object
Two Object Two Object Same Different
One Object
Two Object Two Object Same Different
Figure 12. Results from a split-brain observer (JW) in the unilateral versus bilateral object identification task.
Discussion In this section we discuss various aspects of the results from our three studies, examining threads common to all of them, as well as pointing out important differences between the studies. We do this by attempting to answer four questions that emerge quite naturally from the results of these studies. Why did normal observers demonstrate hemispheric competition when performing visual search on bilateral displays? Visual search was most efficient for normal observers on bilateral displays when the target was presented to the right hemisphere. This result stood in sharp contrast to their search in unilateral displays, where targets were detected in a very similar way in both visual fields. It also stood in contrast to the results of a split brain observer, who performed visual search for left and right targets in very much the same way on unilateral and bilateral displays. A scanning bias? One possibility that must be considered before loftier theoretical speculations arc made is that this result was the consequence of a strategic decision by normal observers. Perhaps normal observers adopted a left-to-right scanning strategy, thereby giving targets in the left visual field an advantage over those in the right field. This would explain an advantage for targets on the left in bilateral
220
Enns and Kingstone
displays. However, it encounters difficulties with several other aspects of the data. First, this account predicts a left field advantage that should grow along with display size, since an increasing number of left field items would have to be inspected before the right field items could be inspected. This prediction was not supported by either the response time task (Figure 6) or the brief exposure task (Figure 8). In both cases, the advantage for left side targets on bilateral displays remained fairly constant across display size, and in some cases performance actually converged with display size for targets in the two visual fields. Second, there were strong hints of nonlinearities in the way performance changed with display size for left side targets. Such nonlinearities are also not predicted by a scanning bias explanation. Consider the slopes from display size 2-8 and then again from display size 8-24 in the two tasks (Figures 6 and 8). In each case, slopes are relatively shallow in the small range and then much steeper in the larger range, to the extent that target detection is similar on both sides at the largest display sizes. This points to a qualitative difference, rather than merely a quantitative difference, between search by the right hemisphere in unilateral versus bilateral displays. Third, a scanning strategy that produced a general left-side advantage should have been evident in normal observers performing the global-local search task (Figure 2), in a split brain observer performing visual search tasks (Figures 3 and 7), and in normal and split brain observers performing the object identification task (Figures 11 and 12). This feature was clearly not a general one of spatial attention tasks, and as such, it strengthens our confidence in its importance to visual search for these specific items. Fourth, a left-to-right scanning strategy should have produced results that were similar for unilateral displays and for targets on the left side in bilateral displays. This hypothesis was examined by comparing he average performance on unilateral displays with that on bilateral displays for the corresponding display sizes. This is shown in Figure 13. Note that in this comparison, a unilateral display size of 12 corresponds to a bilateral display size of 24, a unilateral display of 8 corresponds to a bilateral display size of 16 items, etc. These are the comparable conditions for someone scanning the displays from left to right. The comparison given in Figure 13 shows that this hypothesis was not supported. Except for left side targets in the smallest three display
Hemispheric Coordination
1200
100
221
-
.i-...
r
F-- 1100 t-v-
a >go L_ k..
Q
1000 o 0
/
0
080
/
900
~70 r
r
800
~60
700
a..50
%
4>
600
2
4
6
8
12 16 24
Unilateral Display Size
40
I
I
I
I
I
I
I
2 4 6 8 1 2 16 24 Unilateral Display Size
Figure 13. Performance of normal observers in the bilateral visual search task compared with performance on unilateral displays that have been equated for the number of search items presented in each visual field. Search in bilateral displays is more difficult than in unilateral displays, even when the target is in the left visual field. Open circles: Bilateral - Left Visual Field; Solid squares: Bilateral - Right Visual Field; Shaded circles - Unilateral - Left and Right.
sizes, search for both left and fight side targets was consistently less efficient than for the comparable display sizes in unilateral displays. This points to a competition between the hemispheres that goes well beyond a scanning bias. A competition for mechanisms of feature integration? T h e hemispheric competition was observed in only one task for normal observers, the visual search task patterned after Luck et al (1989, 1994). It was not observed in the global-local visual search task, nor in the object identification task. What distinguished this task from the others? Although there are many differences that could be considered, one theoretically important one is that of feature conjunction. Only the Luck-style task involved search for targets defined solely by the spatial relations among identical elements (white and black squares). This is equivalent to a conjunction of brightness level and relative location (see Figure 5). The global-local search task and the object identification task, on the other hand, involved search for targets defined by the simple feature of item orientation. It is true that in the global-local task, item
222
Enns and Kingstone
orientation had to be determined on the basis of dots that were nonabutting and sometimes of different levels of brightness. Nonetheless, it is likely that the resulting signal concerning item orientation is simpler than the signal for the conjunction of brightness and relative location. Perhaps feature integration involves communication between the hemispheres, at least in observers with an intact corpus callosum. This possibility will clearly have to await additional tests before it can be fully accepted. However, we think it warrants some consideration, in particular, concerning the question of why feature integration may be susceptible to hemispheric competition. If nothing else, such speculation will help guide future experiments that disconfirm the hypotheses. One possibility, therefore, is that the intact brain is organized to permit conjunctions of features to emerge into consciousness for only one location or object at a time. Such an organization would promote unified and coherent action to objects in the visual environment. There is also growing support for such a view among researchers studying the neuropsychological conditions of neglect and extinction (Baylis, Driver, & Rafal, 1993; Cohen & Rafal, 1991; Cohen, Ivry, Rafal, & Kohn, 1995). In the split brain observer, these constraints would no longer be at work and so feature conjunction might be able to proceed independently in the two hemispheres. Another possibility is that visual search for targets defined by feature conjunctions involves more than one attention system: a cortical and a subcortical one. Neurophysiological research suggests that feature integration is performed by mechanisms in the temporal lobe of the cortex. Receptive fields of neurons in this brain region are large in size, often crossing the vertical meridian, and are sensitive to complex combinations of features (Laberge, 1995; Moran & Desimone, 1985). The mechanisms important to search, on the other hand, have large subcortical components. Eye movements to various locations in a display, as well as covert movements of the mind's eye, involve mechanisms in the superior colliculus. The engagement of attention on a new item is governed by mechanisms in the pulvinar nucleus of the thalamus (Laberge, 1995; Posner & Raichle, 1993). The competition seen in visual search by normal observers may therefore be an interaction between these two systems. Search for conjunction-defined targets may activate a competition for activity in temporal lobe neurons, as well as a competition for control of neurons in the thalamus and superior colliculus. Severing the direct connections between the hemispheres, as
Hemispheric Coordination
223
in the split brain observer, may therefore eliminate the cortical aspect of this competition. The loser-wins? One form of hemispheric competition that may be playing itself out in visual search is that between bottom-up (data driven) and top-down (strategic or intentional) mechanisms. As discussed in the introduction to these experiments, visual search is guided by a combination of data driven mechanisms that include spatially-parallel operations for rapid grouping and segmentation, and strategic mechanisms which control the locus of the attentional gaze. It is also strongly possible that one hemisphere is more adept in the use of one of these mechanisms than the other. On the basis of past research on visual search by split-brain observers (Kingstone et al, 1995; Luck et al, 1989, 1994), one might suspect that the left hemisphere was most adept, or at least most dominant, in assuming strategic control over the voluntary locus of attention. Finally, the present results from the splitbrain observer in visual search and object identification tasks indicates that the division of attention between the hemispheres involves a competition for a common mechanisms. Armed with these premises, the following scenario can be considered. In an effort to control visual search on bilateral displays, the left hemisphere searches diligently through the items on the fight side of the display. Control of attention by the left hemisphere in this way leaves little, if any, of the voluntary control mechanisms for use by the fight hemisphere. As such, fight hemisphere search is conducted, at least in the small range of display sizes, by the bottom-up mechanisms of rapid grouping and segmentation. Such mechanisms might indeed be able to point to a target when the display size is small, or when the target is highly distinctive (Pashler, 1987; Treisman, 1982). However, they will also fail at some point, when the relative salience of the target is too low relative to the distractors. From that point on, controlled visual search will be required for targets on the left side as well. This account provides a reason for the nonlinearities seen in the search slopes for targets on the left side in bilateral displays. Slopes should be shallow for small display sizes, reflecting the operation of bottom-up mechanisms in the right hemisphere. The finding that these shallow search slopes were also associated with response times and accuracy levels that were better, in absolute terms, than the performance levels associated with voluntary control of attention, is not predicted explicitly by this account. However, it does fit reasonably well with the
224
Enns and Kingstone
rest of the account, given the slower time course of endogenous versus exogenous orienting (Cheal & Lyon, 1991). The hemispheric competition, in this view, is thus for brain structures important for the voluntary control of attention. This competition is "won" by the left hemisphere, which is often dominant in these matters. This has the unintended consequence of permitting the right hemisphere to perform search primarily with bottom-up mechanisms. These are actually more efficient than voluntary control mechanisms, at least for small display sizes, and so the hemisphere which "loses" the hemispheric competition actually "wins" the behavioral race. The finding that the split-brain observer failed to show this form of competition points further to a competition that involves interhemispheric communication. Why did the split-brain observer show hemispheric competition when performing visual search and item identification on bilateral displays? The split-brain observer was at a large disadvantage in searching through bilateral displays, in comparison to search through the same number of items in unilateral displays. This was true both for targets presented to the left hemisphere, which was generally more adept at search, and for targets presented to the fight hemisphere. This observer also showed a large interaction between visual field and type of display in the object identification study. In this case, his right hemisphere showed an advantage in bilateral displays, showing no interference from the item displayed to the left hemisphere, whereas the left hemisphere suffered noticeably in object identification when another item was presented to the right hemisphere on bilateral displays. What are the subcortical mechanisms that are shared between the disconnected hemispheres in these spatial attention tasks? Posner and Raichle (1993) point to two systems that are each relevant to these tasks. The first is a midbrain structure known as the superior colliculus. Neurons in this structure are not only highly active during the initiation and execution of eye movements, but are also active when the mind's eye moves covertly from one scene location to another (Holtzman, 1984, Wurtz, 1996). When the superior colliculus is damaged in animals, eye movement responses to cues become delayed. A disease in humans known as progressive supranuclear palsy causes selective damage to this brain region, dramatically slowing both eye movements and covert shifts of attention (Rafal, Posner, Friedman, Inhoff & Bernstein, 1988). Finally, studies of covert orienting in split brain observers confirm that cues presented to one visual field can produce orienting responses to
Hemispheric Coordination
225
targets in the other visual field (Holtzman et al., 1981; Holtzman, 1984). This indicates that information is being passed between hemispheres in a spatial attention task involving the superior colliculus, but that the route of communication cannot be cortical. A second relevant subcortical structure is the thalamus. It has long been assigned an important role in the selection of information from multiple sources (Crick, 1984). More recently, PET studies have provided direct evidence of its importance in a visual filtering task (Laberge & Buchsbaum, 1990). During the time an observer was selectively attending to a target item surrounded by distractors, neuronal activity increased in the pulvinar nucleus of the thalamus, but not in other regions of the thalamus, or in cortical visual regions. The functions supported by each of these subcortical structures were important in the efficient performance of the tasks in our studies. Visual search and bilateral object identification both require controlled shifts of attention between items in a visual display; they also require that some items be filtered out from further consideration (i.e., distractors in visual search; the white items in object identification). Our findings therefore indicate that these subcortical functions cannot be performed independently for each of the two hemispheres at the same time. In unilateral displays, the entire system can be devoted to a single hemisphere because of the absence of competition. However, in bilateral displays, one hemisphere gains control over the system, causing a direct detriment in performance for the other hemisphere.
What are implications for understanding spatial attention? The larger point made by these patterns of hemispheric competition is that the performance of complex tasks involves the coordination of activity in a number of specialized brain regions (Zeki, 1995). Tasks involving spatial attention are no exception. Multiple specialized brain regions are involved and so speeded and accurate responses require the coordination of a network of distributed systems (Posner & Raichle, 1993). One unique contribution of the present s t u d i e s is the demonstration that some of these issues can be studied in normal observers. Previous to this study, one form of hemispheric coordination had been studied in patients with damage to the parietal lobes. Damage to this region, especially if it is on the right side, often results in various
226
Enns and Kingstone
forms of neglect of the opposite (left) side of visual space (Kinsbourne, 1977; Rafal, 1994). These patients often fail to notice objects on the left side of the vertical meridian, sometimes they neglect only the left side of objects wherever they appear in the visual field, and sometimes they neglect the left side of their own bodies. It is as though the damaged fight parietal lobe no longer competes with the left lobe for control over attentional mechanisms. A more subtle form of visual neglect, termed visual extinction, is observed in some patients, who tend to show contralateral field neglect only when the ipsilateral visual field is also stimulated (Baylis, Driver and Rafal, 1993; Bradshaw & Mattingly, 1995; Valler, Rusconi, Bignamini, Geminiani, & Perani, 1995). In some cases, even this extinction is modulated by the visual similarity and perceptual coherence of the items in the two visual fields (Ward & Goodrich, 1996). Extinction is strongest when the item in the "good" visual field is most similar to that in the damaged field; extinction is weakened if the items in the two fields are grouped into a single object via connectedness or other Gestalt principles. This clinical phenomenon is therefore additional evidence for hemispheric competition when one region of the brain has been damaged. We are aware of only one previous study of hemispheric competition in normal observers (Reuter-Lorenz, Kinsbourne, & Moscovitch, 1990). In that study, observers were given a line bisection task in conjunction with a stimulus that selectively activated one or the other hemisphere. Errors in line bisection were biased by the hemisphere activation stimulus, such that a longer line segment was perceived in the visual field contralateral to the activated hemisphere. In addition, a rightward orienting bias (left hemisphere dominance) was observed when the two hemispheres were placed into direct competition. These results were interpreted within the framework of the activation-orienting hypothesis (Kinsbourne, 1977), which proposes that control over attentional orienting is governed by a dynamic balance between opponentprocesses in the two hemispheres. These processes are ordinarily in rough balance, with a slight tendency to favor the right side of space. However, selective injury to either side of the brain will result in an imbalance, producing the clinical conditions of hemifield neglect and extinction. From this perspective, the present evidence of hemispheric competition in the visual search task points to yet another behavioral domain in
Hemispheric Coordination
227
which hemispheric coordination can be studied in normal observers. We suspect that other domains will be found. It appears that the key ingredient to observing it in our study was the comparison between performance on unilateral and bilateral displays. Informal anecdotes we have begun collecting from other researchers suggest that this may not be an unusual occurrence. For example, judgments of the direction of coherent motion in a random-dot display are severely compromised by the presence of an irrelevant motion pattern in the opposite visual field (Jane Raymond, personal communication). Competition between the two hemispheres of normal observers will perhaps be seen in any situation in which the two hemispheres are unequal in their abilities, or at least unequal in their demand for preeminence. The challenge for researchers will be to design behavioral tasks that allow the competition to be visible in performance.
Implications for understanding hemispheric specialization Most of the emphasis in past research on hemispheric differences has been on specialization. The guiding questions have been "What tasks are each hemisphere best equipped to perform?" and "What kinds of specialized equipment does each hemisphere bring to a task?" Implicit in this approach is the assumption that specializations are hard-wired, or least very well-established in their home hemisphere as a result of experience and maturation. We are calling for a different emphasis. Instead of viewing evidence of specialization as a direct reflection of the resident equipment of that hemisphere, one should view specialization as the outcome of a competition for control over the network of distributed resources required to perform the task. That is, specialization may not reflect a difference in hemispheric circuitry or algorithms so much as hemispheric dominance in the competition. One reason we propose this view of specialization is because of the diverse pattern of hemispheric specialization seen in split brain observers. Consider JW, without a doubt the most thoroughly studied of these observers. His left hemisphere consistently outperforms his right hemisphere in the Luck-style search task (Luck et al, 1989, 1994; present study), yet his right hemisphere is dominant over the left in the object identification task. Both of these tasks place a heavy demand on the division of attention between spatially separated items; yet the story
228
Enns and Kingstone
on "specialization" is different. We think that there is no reason to expect a uniform pattern of hemisphere dominance if the dominance reflects the outcome of a competition for distributed resources rather than inherent differences in processing. Instead, one might expect dominance to vary a great deal with the task, since each task uses these distributed resources in a slightly different way. Perhaps the most vivid example of this view of specialization comes from JW's visual search performance when a conjunction-defined target is placed among an imbalanced ratio of distractors. An intelligent search strategy would be to search among the smaller of the two subsets of distractors that share a feature with the target. Although normal observers employ this "guided search" strategy for targets in either visual field, JW and one other split brain observer did so for targets on the right side (Kingstone et al, 1995). It therefore appeared that the left hemisphere was specialized for guided search. However, subsequent testing has revealed that JW can perform guided search with his right hemisphere (left field displays), provided that these displays occur in a block of trials in which the left hemisphere is not presented with an imbalanced ratio of distractors (Kingstone & Enns, unpublished). We conclude that the apparent specialization of the left hemisphere for guided search should more correctly be seen as a dominance of the left hemisphere when placed in competition with the right hemisphere for task-relevant mechanisms. How should questions of specialization be distinguished from issues of hemispheric coordination in future research? One of the necessary first steps is to design experiments that systematically vary the degree of competition between hemispheres. The traditional approach of specialization studies involves comparisons of unilateral left and right field displays, or comparisons of bilateral displays in which the target item is in the left or the fight field. This does not manipulate the level of competition. It is only when unilateral and bilateral displays are compared for the same visual field that competition can be studied. Our studies, involving a simple comparison of this kind, represent only the beginning of what is possible with this methodology. In future studies we plan to systematically vary the nature of the stimulus in the competing visual field, in a similar way to how this has been done in studies of extinction following parietal lobe damage (Baylis, et al, 1993; Ward & Goodrich, 1996).
Hemispheric Coordination
229
Acknowledgments The research described in this chapter was supported by grants from the Natural Science and Engineering Research Council of Canada to both authors and by a grant from the Alberta Heritage Foundation to A. Kingstone.
References Badcock, J. C., Whitworth, F. A., Badcock, D. R., & Lovegrove, W. J. (1990). Low frequency filtering and the processing of local-global stimuli, Perception, 19, 617-629. Baylis, G. C., Driver, J., & Rafal, R. D. (1993). Visual extinction and stimulus repetition. Journal of Cognitive Neuroscience, 5, 453-466. Bradshaw, J. E. & Mattingly, J. B. (1995) Neuropsychology: Behavioral and brain science. San Diego: Academic Press. Cheal, M., & Lyon, D. R. (1991). Central and peripheral precuing of forced-choice discrimination. Quarterly Journal of Experimental Psychology: Human Experimental Psychology, 43A, 859-880. Cohen, A., & Rafal, R. D. (1991), Attention and feature integration: Illusory conjunctions ion a patient with a parietal lobe lesion. Psychological Science, 2, 106-109. Cohen, A., Ivry, R. B., Rafal, & Kohn, C. (1995). Activating response codes by stimuli in the neglected visual field. Neuropsychology, 9, 165-173. Crick, F. (1984). The function of the thalamic reticular complex: The searchlight hypothesis. Proc. Natl. Acad. Sci. USA, 81, 4586-4590. Duncan, J. (1980). The locus of interference in the perception of simultaneous stimuli. Psychological Review, 87, 272-300. Duncan, J. (1984). Selective attention and the organization of visual information. Journal of Experimental Psychology: General, 113, 501-517. Duncan, J. (in press). Similarity between concurrent visual discriminations: Dimensions and objects. Perception &
Psychophysics.
Duncan, .L, & Humphreys, G. W. (1989). Visual search and stimulus similarity. Psychological Review, 96, 433-458. Enns, J. T. & Kingstone, A. (1995). Access to global and local properties in visual search for compound stimuli. Psychological Science, 6, 283-291. Holtzman, J. D. (1984). Interactions between cortical and subcortical visual areas: Evidence from human commissurotomy patients. Vision Research, 24, 801-813. Holtzman, J. D., Sidtis, J. J., Vol~, B. T., Wilson, D. H., & Gazzan!ga, M. S. (1981). Dissociation of spatial information Ior stimulus localization and the control of attention. Brain, 104,861-872. Hughes, H. C., Fendrich, R., & Reuter-Lorenz, P. A. (1990). Global versus local processing in the absence of low spatial frequencies. Journal of Cognitive Neuroscience, 2, 272-282.
230
Enns and Kingstone
Jane Raymond (personal communication). Department of Psychology. University of North Wales, Bangor, Wales. Kinchla, R. A., & Wolfe, J. M. (1979). The order of visual processing: "top-down", bottom-up" or "middle out". Perception & Psychophysics, 25, 225-231. Kingstone & Enns, unpublished. The lateralization of guided search: "Hardware" versus "software". Kingstone, A., Enns, J. T., Mangun, G. R., & Gazzaniga, M. S. (1995). Guided visual search is a left-hemisphere process in split-brain patients. Psychological Science, 6, 118-121. Kinsbourne, M. (1977). Hemi-inattention and hemispheric rivalry. In E. A. Weinstein & R. P. Freidland (Eds.), ttemi-attention and hemispheric specialization: Vol. 18. Advances in Neurology (pp. 4149). New York: Raven Press. Kitterle, F. L., Christman, S., & Conesa, J. (1993). Hemispheric differences in the interference among components of compound gratings. Perception & Psychophysics, 54, 785-793. Laberge, D., & Buchsbaum, M. S. (1990). Positron emission tomographic measurements of pulvinar activity during an attention task. Journal of Neuroscience, 10, 613-619. Laberge, D. (1995). Computational and anatomical models of selective attention in object identification. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 649-663). Cambridge, MA: MIT Press. Lamb, M. R., & Robertson, "I5. C. (1990). The effect of visual angle on global and local reaction times depends on the set of visual angles presented. Perception & Psychophysics, 47, 489-4%. Lamb, M. R., Robertson, L. C., & Kni.ght, R. T. (1990). Component mechanisms underlying the processing of hierarchically organized patterns: Inferences from patients with unilateral cortical lesions.
Journal of Experimental Psychology: Learning, Memory, & Cognition, 16, 471-483.
Luck, S., Hillyard, S. A., Mangun, G. R. & Gazzaniga, M. S. (1989). Independent hemispheric attentional systems mediate visual search in split-brain patients. Nature, 342, 543-545. Luck, S. J., Hillyard, S. A., Mangun, G. R., & Gazzaniga, M. S. (1994). Independent attentional scanning in the separated hemispheres of split-brain patients. Journal of Cognitive Neuroscience, 6, 84-91. Martin, M. (1979). Local and global processing: The role of sparsity. Memory & Cognition, 7, 476-484. Moran, J., Desimone, R. (August, 1985). Selective attention gates visual processing in the extrastriate cortex. Science, 229, 782-784. Navon, D. (1983). How many trees does it take to make a forest? Perception, 12, 239-254. Pashler, H. (1987). Detecting conjunctions of color and form: Reassessing the serial search hypothesis. Perception & Psychophysics, 41, 191-201. Posner, M. I., & Raichle, M. E. (1993). Images of mind. NY: Scientific American Library. Rafal, R. (1994). Neglect. Current Opinion in Neurobiology, 4, 231-236.
Hemispheric Coordination
231
Rafal, R. D., Posner, M. I., Friedman, J. H., Inhoff, A. W., & Bernstein, E. (1988). Orienting of visual attention in progressive supranuclear palsy. Brain, 111, 267-280. Raymond, J. (personal communication). Department of Psychology. University of North Wales, Bangor, Wales. Reuter-Lorenz, P. A., Kinsbourne, M., & Moscovitch, M. (1990). Hemispheric control of spatial attention. Brain & Cognition, 12, 240266. Robertson, L. C., & Lamb, M. R. (1991). Neuropsychological contributions to theories of part/whole organization. Cognitive Psychology, 23, 299-330. Treisman, A. (1982). Perceptual grouping and attention in visual search for features and for objects. Journal of Experimental Psychology: Human Perception & Performance, 8, 194-214. Treisman, A., & Gelade, G. (1980). A feature integration theory of attention. Cognitive Psychology, 12, 97-136. Valler, G. Rusconi, M. L., Bignamini, L., Geminiani, G. & Perani, D. (1995). Anatomical correlates of visual and tactile extinction in humans: A clinical CT scan study. Journal of Neurology, Neurosurgery, & Psychiatry, 57, 464-70. Ward, R, & Goodrich, S. (1996). Differences between objects and nonobjects in visual extinction: A competition for attention. Psychological Science, 7, 177-180. Wolfe, J. M., Cave, K. R., & Franzel, S. L. (1988). Guided search: An alternative to the feature integration model for visual search. Journal
of Experimental Psychology: Human Perception & Performance, 15,
419-433. Wurtz, R. H. (19%). Vision for the control of movement. Investigative Ophthalmology and Visual Science, 37, 2131-2145. Zeki, S. (1995). A vision of the brain. Boston, MA: Blackwell.
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 1997 Elsevier Science B.V.
233
Chapter 8
Asymmetries in the Flanker Compatibility Effect Frederick Kitterle Northern Illinois University Mark R. Ludorf & Jeremy Moreland Stephen F. Austin State University One issue of concern in studies of selective visual attention is degree to which attention can be narrowly focused on a given spatial location such that stimuli falling within this region are fully processed, whereas those falling outside the focal region are excluded from processing. Several studies indicate that the ability to do this is limited. That is, stimuli that fall outside the zone of attention, which are irrelevant to the task, in fact may interfere with the efficient performance of the task, cannot be ignored and thus, influence the processing of stimuli within the focal region. For example, Eriksen and Eriksen (1974) found that when a target letter that was associated with a given response was flanked by letters that were associated with the same response (response compatible condition), time to identify the target was somewhat faster (although not consistently so) than if flanked by letters that has no assigned response (response neutral condition). On the other hand when the target was surrounded by letters that were associated with a different response (response incompatible condition), reaction time to identify the target was considerably longer than the response neutral condition. The difference in reaction time between the incompatible and the compatible
234
Kitterle, Ludorf, and Moreland
conditions reflects the degree to which the flankers influence reaction time to the target. This is referred to as the flanker compatibility effect (FCE; see Eriksen, 1995 for a review). Thus, even though the task explicitly requires identifying the central target letter in a string of three letters and ignoring the flanking letters, the FCE indicates that the flanking letters cannot be ignored; rather they are also identified and influence RT to the central target letter. Thus, there appears to be a limit on the degree to which attention can be narrowed. Several studies have addressed the issue of where in the information processing stream that limit is set, that is, whether the FCE reflects an early or late selection process. Early selection processes (e.g., Treisman, 1964) hypothesize that the processing of stimuli outside of the focus of attention is confined to rudimentary physical properties whereas, for late selection theories (e.g., Deutsch & Deutsch, 1963), stimuli outside the focus of attention are fully identified. Eriksen and Eriksen (1974) found that the magnitude of the FCE decreased with spatial separation. Nevertheless, the effect is quite robust; it is still present at large spatial separations (e.g., Miller, 1991). Eriksen and Eriksen (1974) have interpreted this as evidence for an imperfect late selection process, which reflects response competition and is driven by the spatial allocation of attention. Other research supports the view of a late selection response competition interpretation. For example Coles, Gratton, Bashore, Eriksen, and Donchin (1985) demonstrated this physiologically. Miller (1987) has shown that neutral letters that are correlated with a particular response act as congruent stimuli for this response. However, early selection processes also play a role in the flanker effect. Studies have shown that the physical characteristics of the flankers also contribute to the FCE. Response neutral flankers that are similar to response incompatible flankers cause a slower response than response neutral flankers that are similar to response compatible flankers (Eriksen & Eriksen, 1974). Flankers that are response compatible and identical to the target produce faster reaction times that flankers that are only response compatible (Eriksen & Eriksen, 1979; Eriksen & Schultz, 1979). Yeh and Eriksen (1984) found that physical similarity between target and flanker (e.g., both upper case) has a greater effect than name similarity (e.g., target upper case and flankers lower case). LaBerge, Brown, Carter, Bash, and Hartley (1991) have shown that the effect of flankers could be reduced by manipulating the focus of attention. They presented in the same spatial location a digit
Flanker Effect
235
prior to presenting a target flanker array and required that subjects identify both the preceding digit and the target letter. LaBerge et al. (1991) manipulated the focus of attention by shortening the duration of the first target. They assumed that, with shorter durations, the focus of attention would be narrowed and carried over to the target-flanker presentation. They showed that shorter durations of the digit target reduced the influence of the flankers and with slight increases in spatial separation between the letters, the effects of the flankers was virtually eliminated. These results suggest that there is an early selection spatial component to the FCE in addition to any response selection components because the focus of attention was set prior to the onset of the targetflanker display. Baylis and Driver (1992) demonstrated that the FCE decreases when the target and flankers are different colors. These results complement other work which indicates that perceptual grouping principles influence the magnitude of the FCE (Harms & Bundesen, 1983; Kramer & Jacobson, 1991). The fact that the impact of flankers varies with distance from the target is consistent with spotlight or zoom lens models of selective attention (Eriksen & Eriksen, 1974; Posner, 1980). That is, flankers falling within the attentional beam are processed, whereas those falling outside are not. For most experiments, the flanker to the left and to the fight of the target are both consistent, both inconsistent, or both neutral. An important question which this study addresses is whether the magnitude of interference from each flanker is equal or whether interference with target processing is asymmetrical and depends upon the spatial position of the flanker. This question is motivated by recent work showing that target identification in multielement arrays is influenced by spatial position. For example, Efron, Yund, and Nichols (1990) proposed a serial scanning mechanism which is biased to begin scanning at the top and right of a display. An implication of this hypothesis for the FCE is that in horizontal displays - the fight flanker will have a greater effect than the left. In vertically oriented displays, the top flanker will have a greater effect than the bottom. Also, research on hemispheric differences in information processing suggests the possibility of differences in the effectiveness of the left and right flanker because in the three element display, the left and fight flankers fall in different visual fields. Although differences in the relative effectiveness of flankers are not predicted by either the zoom lens or spotlight models, there is evidence,
236
Kitterle, Ludorf, and Moreland
which suggests a greater influence on target processing from the left flanker. This is based upon the studies indicating a left side advantage in the processing of letter like strings (Posner, Snyder, & Davidson, 1980; Pashler, 1984). On the other hand other research has found right side advantages in the shifting of attention (Posner et al., 1980; Laberge & Brown, 1986; 1989). Dowling and Pinker (1985) found that the distribution of attention was asymmetric, the right visual field showing an advantage in facilitating responses and in the detection of a luminance increment (Hughes & Zimba, 1985). These results suggest that flanker on the fight should have a greater effect on the processing of a central target than those on the left. However, Hommel (1995) has argued that in strings that closely approximate words, asymmetries resulting from the automatic initiation of reading-like habits may also account for the fact that the left flanker has a considerably greater effect on the FCE than the fight flanker. Preliminary research on the FCE indicates complex relationship in which the effectiveness of the left or the fight flanker depends upon stimulus characteristics (Hommel, 1995). For example, with letter strings, flankers to the left of the target letter produce greater response compatibility and incompatibility effects than those on the fight (Beach, 1995; Harms & Bundesen, 1983; Hommel, 1995). However, other research has found fight-side flanker effects with mirror-image letters and with geometrical forms (Hommel, 1995). In summary, regardless of which flanker may be dominant in the FCE, it is important to note that neither the spotlight nor the zoom lens model assumes asymmetries in the relative effectiveness of the left or right flanker (that is, of course assuming that the center of the distribution of attention is focused at the center of the target letter). Thus, the purpose of this study is to determine further those factors leading to asymmetries in the FCE.
Expt. 1: Left-right asymmetries in the FCE: M and W letter arrays As indicated earlier, research indicates the existence of left-right asymmetries. However, the flanker that is most effective appears to depend upon stimulus characteristics (Hommel, 1995). For example, with letter strings flankers to the left of the target letter produce greater response compatibility and incompatibility effects than those on the right (Beach, 1995; Harms & Bundesen, 1983; Hommel, 1995). In
Flanker Effect
237
contrast to these data, Kitterle and Ludorf (1993) reported preliminary data indicating flanker asymmetries in which the right flanker produced a greater compatibility effect than the left. In that study, the letters M and W were used as targets and the flankers. In their study there were eight stimulus conditions that resulted from orthogonally varying the left and right flankers and factorally combining them. It should be noted that the letters M and W have component line segments that are highly similar in contrast to the letters used by others (e.g., Beach, 1995; Harms & Bundesen, 1983; Hommel, 1995). It might be argued that identification of these letters is more critically dependent upon discriminating local features. Given research on hemispheric differences in the processing of local and global stimuli, a fight visual field/left hemisphere difference might be expected. In light of these apparent discrepancies, this experiment is designed to determine the direction of flanker asymmetries using the same letters as in Kitterle and Ludorf (1993). In their study, Kitterle and Ludorf (1993) made the general assumption that the effects of two incompatible flankers on the FCE was greater than a display with only one incompatible flanker. Consequently, when one flanker was response compatible (C) and the other response incompatible (I), then reactions time (RT) for left incompatible/right compatible, RT(I-C), or left compatible/right incompatible, RT (C-I), should fall between the RTs when both flankers were compatible, RT(C-C), or when both were incompatible, RT(I-I). Kitterle and Ludorf (1993) also proposed the following specific hypotheses: Hypothesis 1: Under the assumption that both flankers exert equal interference and one incompatible flanker produces less interference than two [left compatible-right incompatible (C-I) or left incompatiblefight compatible (I-C)], then RT (C-I) = RT (I-C) and RT(C-C) < RT(C-I or I-C) < RT (I-I). Hypothesis 2: Flanker interference is asymmetrical, one incompatible flanker produces less interference than two, then RT (I-C) RT (C-I)and RT(C-C) < RT(C-I) < RT(I-C) < RT (I-I) or RT(C-C) < RT(I-C) < RT(C-I) < RT (I-I).
238
Kitterle, Ludorf, and Moreland
Hypothesis 3: Interference is asymmetrical, one flanker totally accounts for interference, then RT (I-C) ~ RT (C-I) and RT(C-C) < RT(C-I) < RT(I-C) = RT (I-I) or RT(C-C) < RT(I-C) < RT(C-I) = RT (I-I). The present experiment tests these hypotheses as well as considering experimental conditions in which the stimulus display contains one response incompatible flanker and one response neutral (N) flanker (that is, displays in which there is no response assigned to one of the flanker letters). Thus, we also test conditions for flanker asymmetries of the form RT(I-N) vs. RT(N-I) as well as RT(C-N) vs. RT (N-C). As noted earlier, the left and the right flankers project to the right and left hemisphere, respectively. Given the fact that the FCE reflects, in part, response competition, it is of interest to examine how hand of response and flanker position interact to determine the magnitude of the FCE. For example, it might be assumed that in the letter array WMM (or WMN), the magnitude of the FCE might be larger with the left hand responding to the target letter M (ML) and the right hand to the target letter W (We,) than vice versa. The basis for this assumption is that if the left flanker has a greater effect on the FCE because of an automatic left to right scanning process (Hommel, 1995), then the left flanker should prime the hemisphere that controls response, namely the right hemisphere. In this case, more inhibition may be needed to suppress the primed response "W" if the correct response "M" is to be made. On the other hand when the "W" projects to the hemisphere that controls the response "M", there is less priming of the "W" response and consequently less inhibition with a resulting smaller FCE effect. Subjects. Forty undergraduates participated in this experiment and received extra credit for participation. Subjects had normal or correctedto-normal vision and were naive about the purpose of this study. Apparatus and stimuli. PC-compatible computer workstations were used to present stimuli and collect responses and latencies. Subjects were seated at a work station 54 cm from the monitor (14" SVGA color monitor), which was positioned at eye level. A standard keyboard in front of and below the monitor was used to record responses and latencies. Subjects responded to the target stimuli using the "z" and "/"
Flanker Effect
239
keys located on the bottom left and right sides of the keyboard, respectively. Mapping of targets to response keys was counterbalanced. The stimuli consisted of three-letter arrays that were horizontally oriented with the central letter falling in the fovea. The letters were upper case with a width of 15 min. and height of 30 min. The distance between each letter was 5.4 min. The central target letter (M or W) was flanked on the left and right of the array by one of four letters randomly chosen letters from the set (M, W, E, or G). A second neutral letter was chosen so that the total number of neutral letters equaled the number of target letters. Thus, for the left and right flankers there were four conditions: response compatible and response incompatible plus two response neutral conditions. The left and right flankers were orthogonally varied and crossed with each other to produce 8 arrays (e. g., MWM, WMM, EWW, GWM, etc.). Procedure. Subjects performed a block of 32 practice trials followed by a block of 128 experimental trials. On each trial, a fixation cross was presented for 1000 msec, followed by a 500 msec warning tone. After a variable ISI of 700 to 1200 msec. the fixation cross was extinguished and a three letter array was presented for 500 reset. Subjects were instructed to identify as quickly and accurately as possible whether the central target letter was an 'M' or a 'W' by using the "z" key or "/" key as response input, respectively. Subjects were given a total of 2500 msec. to respond. For half of the subjects the target letter M was responded to with their left hand and the target letter W with the right hand and for the other half this was reversed (i.e., MLWR vs. MR WE, respectively). Results. Preliminary analyses of the data indicate that there were no significant differences between the neutral conditions. Consequently, they were collapsed. The data, which are shown in Table 1, presents correct mean reaction times as a function of stimulus condition. These data were analyzed by means of a 2 Hand of Response (between group factor: MLWR, MRWL) X 3 Left Flanker (compatible, incompatible, neutral) X 2 Target Name (M or W) X 3 Right Flanker (compatible, incompatible, neutral) split-plot ANOVA. There were significant main effects of Left Flanker [F(2,76)= 69.40, p< .0000001] and Right Flanker [F(2,76)= 30.25, p<.0000001]. For the Left Flanker the mean RTs for the various conditions was as follows: response compatible (591 reset), response incompatible (643) and response neutral (608 msec). The response compatible RT was
240
Kitterle, Ludorf, and Moreland
Table 1. Results of Experiment 1. Hand of R e s p o n s e : M
(left)
Target Letter
W
(right)
- M
Right Flanker Left Flanker Compatible Incompatible Neutral
Compatible
Incompatible
563 641 588
607 650 627
575 646 592
Hand of R e s p o n s e : M . . . . . .
.
.
Neutral
.
.
.
.
(left)
W
(right)
.
Target Letter
- W
Right Flanker Left Flanker Compatible Incompatible Neutral
Compatible
Incompatible
577 614 592
Neutral
607 612 606
595 616 605
Hand of R e s p o n s e : M
(right)
Target Letter Right Left Flanker Compatible Incompatible Neutral
Compatible
- M
Flanker
Incompatible
575 649 607
Neutral
648 656 638
586 661 594
Hand of R e s p o n s e : M
(right)
Target Letter Right . . . . . . . .
Left Flanker Compatible Incompatible Neutral
.
.
.
.
.
.
.
Compatible 589 652 611
.
.
.
.
.
.
.
.
W (left)
W (left)
- W
Flanker .
.
.
.
.
.
.
.
Incompatible 608 664 643
.
.
.
.
Neutral 587 646 617
Flanker Effect
241
significantly faster than the neutral condition indicating facilitation of RT and the response incompatible condition was significantly slower than the neutral condition, which indicates response interference. Clearly, the response incompatible condition was significantly slower than the response compatible condition. For the Right Flanker condition the RTs were as follows: response compatible (603 msec), response incompatible (631 msec), and response neutral (608 msec). In this condition, there was a significant difference between the response compatible and response incompatible conditions, indicating a flanker compatibility effect. However, the response neutral condition did not differ significantly from the response compatible condition. Thus, in this case there was no apparent facilitation of RT for the compatible condition. For the other two main effects, neither the Hand of Response or the Target Name were significant [F
242
Kitterle, Ludorf, and Moreland 660 -
640 620
RT 600
580 560
I
Compatible
"
I
'
Incompatible Left
I
Neutral
Flanker
Figure 1: Reaction time as a function of left flanker condition for response compatible fight flanker (circles), response incompatible fight flanker (squares), and response neutral fight flanker (triangles).
both a response compatible and response incompatible flanker. As seen in Fig. 1, a flanker that is response compatible produces a faster response if it is on the left (C-N) than the fight (N-C; 585 msec vs. 600 msec, respectively). Furthermore, a response incompatible flanker on the left (I-N: 642 msec) produces a slower RT than on the fight (N-I: 629 msec.). Thus, these more extensive results contradict those of Kitterle and Ludorf (1993) and indicate that it is the left flanker, not the fight, which dominates in the FCE. Moreover, the results of this study indicate that the left flanker dominance does not depend upon whether the right flanker is response compatible or response neutral. The second order interaction, Hand of Response X Target Name X Left Flanker [F(2,76)=6.86, p<.002], was also significant. These data are presented in Table 2. Consider first the results when the flanker is inconsistent. In the condition in which the hand of response M(left) W(right) for the target condition WM (e.g., left flanker W and center target M ) o r the hand of response M(right) W(left) for the target condi-
Flanker Effect
243
Table 2: Reaction time for Expt. 1 as a function of hand of response, target name, and left flanker condition.
RESPOND M(left) W(right) M(right) W(left)
INCON SISTENT Left Flanker-Center Target WM MW 649 655
CON SISTENT Left Flanker-Center Target MM WW
614 654
582 603
593 595
tion MW (e.g., the left flanker M and the center target W), the inconsistent flanker is processed by the hemisphere controlling the hand responding to the target. In contrast, the condition in which the hand of response M(left) W(right) for the target condition MW (e.g., the left flanker M and the center target W) or the hand of response M(right) W(left) for the target condition WM (e.g., left flanker W and center target M), the inconsistent flanker is processed by the hemisphere that does not control the hand responding to the target. The former yields an RT of 652 msec. and the latter an RT of 635 msec. These results indicate that the magnitude of the FCE depends upon whether or not the incompatible flanker is processed by the hemisphere that controls both the response to the target stimulus and the flanker. If this is the case, then the FCE is increased, whereas, if it is not, then the FCE is smaller. In summary, interference is asymmetric, with left flankers producing greater interference effects than fight flankers. In order to determine the generality of these results and the degree to which they are consistent with those of other researchers (e.g., Beach, 1995; Hommel, 1995), we have utilized different letters to test for asymmetries in the FCE. Experiment 1B- FCE with H, V letter arrays The target letters used by Hommel (1995) were "S" and "H". Another study demonstrating a greater left flanker effect (Beach, 1995) used the letter set "C", "H", "S" and "K" with two of the target letters assigned to one response and the other two to a different response. The letters in these two studies differ considerably in the orientation of their constituent components, whereas in Experiment 1A the component features are quite similar in orientation. For this reason we explored the
244
Kitterle, Ludorf, and Moreland
asymmetries of the FCE with the letters "V" and "H", which share fewer features than the letters "M" and "W". Subjects, apparatus, and procedure. Fifty four subjects with normal or corrected-to-normal vision participated in this experiment for course credit. They were naive about the purpose of this study and did not participate in the earlier studies. As in the previous experiment, PC-compatible computer workstations were used to present stimuli and collect responses and latencies. Subjects were seated at a work station 54 cm from the monitor (14" SVGA color monitor), which was positioned at eye level. A standard keyboard in front of and below the monitor to was used to record responses and latencies. Subjects responded to the target stimuli using the "z" and "/" keys located on the bottom left and right sides of the keyboard, respectively. Mapping of targets to response keys was counterbalanced. Subjects performed a block of 32 practice trials followed by a block of 256 experimental trials. On each trial, a fixation cross was presented for 1000 msec, followed by a 500 msec warning tone. After a variable ISI of 700 to 1200 msec. the fixation cross was extinguished and a three letter array was presented for 500 msec. Subjects were instructed to identify as quickly and accurately as possible whether the central target letter was a 'V' or an 'H' by using the "z" key or "/" keys as response input. Subjects were given a total of 2500 msec to respond. Target letter and response key were counter-balanced across subjects. Results. The results of this experiment were analyzed by means of a 2 Left Flanker (compatible, incompatible) X 2 Right Flanker (compatible, incompatible) repeated measures ANOVA on correct RTs. There were significant main effects of Left Flanker [F(1,53)= 116.02, p<.00001] and Right Flanker [F(1,53)=44.37, p<.00001]. For both flankers there was a significant difference between the RTs for response incompatible vs. response compatible flankers. In addition, there was a significant Left Flanker X Right Flanker interaction [F(1,53)=10.93, p<.002]. These results are plotted in Figure 2. Fine grain analyses of these results indicated significant differences between the CC (574 msec) vs. the CI conditions (604 msec) [t(1,53)= 6.06, p<.001], between the CC and IC conditions (619 msec) [t(1,53)= 10.67, p<.0001], between the CC and II (634 msec) [t(1,53)=11.25, p<.00001] and between the CI and IC conditions [t(1,53)=3.38, p<.001]. Thus, there is a significant left flanker asymmetry. Thus, the
Flanker Effect
245
640 -
620
600 RT 580
560
540
|
c-c
Left-Right
|
|
C-I Flanker
I-C
|
I-I
Compatibility
F i g u r e 2: Results of Expt. 1B for letters H and V. RT as function of left-right
flanker compatibility (e.g., left flanker compatible and fight flanker incompatible = C-I).
significant left flanker asymmetry found in Expt. 1A cannot be due to the particular letters that were selected. In the previous experiments, both target and flanker letters had the same case. The next experiment explores the effects of letter case on the flanker asymmetry effect. E x p e r i m e n t 2 - E f f e c t s o f letter c a s e
Physical similarity between the target and flankers affects the magnitude of the FCE. Eriksen and Eriksen (1974) have shown that response neutral flankers that were physically similar to incongruent flankers caused a slower response than neutral flankers, which were physically similar to congruent flankers. Moreover, flankers which were congruent and identical to the target yield faster reaction times than flankers which were congruent but not identical (Eriksen & Eriksen, 1979; Eriksen & Schultz, 1979). These studies have used targets and flankers that were similar in case, that is, upper case letters. Variations in lettercase have between the target and flanker permit an evaluation of
246
Kitterle, Ludorf, and Moreland
the effect of both the name and the physical similarity of the flankers to the target on the magnitude of the FCE. For example, Yeh and Eriksen (1984) found that targets surrounded by flankers which had the same name and case were responded to faster than targets presented in isolation and also faster than targets presented with flankers having the same name but different case. In contrast, targets surrounded by response incompatible letters were, in general, slower when flankers were the same case as the target. However, it should also be noted that, in the specific case in which an upper case target was surrounded by lower case letters of the same name but the lower case flankers were physically similar to incompatible targets, RT was the same as that found with upper case letters from the opposite response set. Thus, both the physical similarity and the name similarity influence the magnitude of the FCE. However, physical similarity has a greater effect than name similarity. The purpose of this experiment is to determine the effect of name and physical similarity on the magnitude of the FCE. Hellige and Webster (1981) explored the effects of physical and name identity on the time to determine whether two letters presented in either the fight or the left visual field had the same name. RT was faster when the two stimuli were physically identical than when they had the same name (i.e., one was upper and the other lower case). For left visual field presentations, the lettercase effect was as large on different name trials as on same name trials. For right visual field presentations, however, the letter case effect was restricted to same name trials. Thus, there is a qualitative visual field difference that depends upon letter-case. Sergent (1984) has shown that time to identify a three letter word was the same for upper and lower case when presented in the left visual field. However, identification for lower case words was longer than upper case words for fight visual field presentations. In light of these results, it is of interest to determine whether there are differences in the degree to which physical similarity and name similarity influence the magnitude of the flanker asymmetries in the F C E . The effects of letter case on the magnitude of the FCE is interesting in light of research by Harms and Bundesen (1983). They showed a reduced effectiveness of flankers when they differed from the target in color. It may well be that letter case, like color, isolates flankers from targets in processing with a resultant decrease in flanker effectiveness. Given the findings of our earlier experiments on flanker asymmetry, it is
Flanker Effect
247
of empirical interest to determine whether letter case modulates the asymmetrical FCE. The influence of letter case and name is also important from a theoretical perspective. Hommel (1995) has suggested that the greater influence of the left flanker on the FCE is due to an automatic left-tofight, attentional scanning mechanism. The specific role of the structural characteristics of the stimulus are thought to be minimal. However, his interpretation of the structural characteristics of the stimulus pertained only to the spatial separation of the letters and not to their case. If, in fact, structure does not play a role and an automatic scanning process is initiated when a stimulus is presented, it might be expected that letter strings in which the target and flankers letters are similar in name and case should produce RTs that are equivalent to stimuli with the same name and different case. Moreover, for letter strings in which the left flanker differs, the asymmetry effect should be present as well. Subjects. Twenty six right-handed males participated in this experiment for course credit. They all had normal or corrected to normal vision and were naive about the purpose of this study. None of the subjects in this experiment participated in the earlier studies. Apparatus and Stimuli. The stimulus array consisted of a target letter surrounded on either side by flanking letters. The central target letter was either an upper- or lowercase B or D. The flanking stimuli were upper and lower case letters B and D. At a viewing distance of 54 cm, the letters subtended 30 min X 15 min and the separation between letters was 5.4 min. These letters were selected because of the structural differences between the upper and lower case letters and the similarity of the letters of the same case. The stimuli were also selected to equate phonemic similarity to reduce any confounding of phonemic structure with the name code. There were a total of 64 stimuli created by factorally combining four target types by four left flanker by four right flanker types. Procedure. At the beginning of each trial a fixation cross was presented for 500 msec. Following the offset of the fixation cross there was a 200 msec warning tone and after a random foreperiod (200-800 msec) the stimulus array was presented for 200 msec. Subjects were required to respond press the "/" to indicate one of the target letters (e.g.., B) was presented and the "z" to indicate the presence of the other target letter (e.g.., D). Subjects responded in the 2500 msec interval before the next trial. The mapping of target letter to response key was
248
Kitterle, Ludorf, and Moreland
counterbalanced. There were ten practice trials followed by 320 experimental trials (5 replications of the 64 conditions) divided into 5 blocks of trials. Results. The correct RTs of this experiment were averaged over Target Name and Target Case and then analyzed by means of a four factor [2 Left Flanker Name (consistent vs. inconsistent) X 2 Left Flanker Case (same vs. different) X 2 Right Flanker Name (consistent vs. inconsistent) X 2 Right Flanker Case (same vs. different)] repeated measures ANOVA. Note that the variable left (right) flanker name refers to whether the flanker letter required a response that was consistent with the response to the target or inconsistent with it, whereas the variable left (right) flanker case refers to whether the flanker letter was the same case or differed in case from the target letter. These data are summarized in Table 3. Both the main effects of Right Flanker Name and Left Flanker Name were significant [F(1,25)=6.93, p<.01 ; F(1,25)=59.25, p<.0001, respectively]. In general, RTs were slower for inconsistent than for consistent flankers. However, the main effect of Flanker Case was not significant for either right or left flankers. There were significant first order interactions of Right Flanker Case X Right Flanker Name [F(1,25)=5.22, p<.03], Left Flanker Case X Left Flanker Name [F(1,25)= 11.89, p<.002], and Right Flanker Case X Left Flanker Name [F(1,25)=5.09, p<.03]. In addition, there was a second order interaction of Right Flanker Name X Left Flanker Case X Left Flanker Name [F(1,25)=4.80, p<.04]. These effects are not discussed in detail because their interpretation is constrained by the significant interaction of Right Flanker Case X Right Flanker Name X Left Flanker Case X Left Flanker Name [F(1,25)=5.81, p<.03]. In order to more clearly appreciate the way in which target case and name influences the magnitude of the flanker asymmetries, these data are plotted in Figure 3. This figure shows RT as a function of the leftright flanker case in relation to the target case (e.g., left same-right different: SD). Filled squares plot the results when the left flanker is response incompatible and the right flanker response compatible (I-C). Filled circles plot the results for the left flanker compatible and the right flanker incompatible (C-I). As can be readily seen, the left flanker has a greater impact on target processing than the right (647 vs. 604 msec) when both have the same case. Changing the case of response incompatible flankers reduces the FCE for both IC and CI conditions (618 vs.
Flanker Effect
249
Left Flank Name Left Flank Case Right Hank Name Right Flank Case
RT
Consistent
T a b l e 3: Summary of Experiment 2.
Consistent
Inconsistent
Inconsistent
Consistent
Inconsistent
Consistent
Consistent Inconsistent
568 598
Inconsistent
Consistent Inconsistent
604 592
Consistent
Consistent Inconsistent
596 608
Inconsistent
Consistent Inconsistent
602
Consistent
Consistent Inconsistent
647 618
Inconsistent
Consistent Inconsistent
631 627
Consistent
Consistent Inconsistent
610 617
Inconsistent
Consistent Inconsistent
634 619
614
592 msec., respectively). In fact, the FCE is virtually non-existent for the CI condition, as RT in this condition is indistinguishable from RT to the CC condition regardless of the case of either the left or right flanker. In contrast to these results, note that when the lettercase of the response compatible flanker differs from the target letter case, there is no difference for the CI and IC conditions (614 vs. 618 msec, respectively). Thus, lettercase appears to have the greatest effect on the response incompatible flanker. In addition, when flankers are both response compatible or incompatible, changing lettercase of the left or right flanker has little effect on RT (a 2 msec difference for the former condition and a 7 msec difference for the latter). Unlike other research showing right visual field effects for changes in lettercase in a letter matching task, our results indicate that changing the lettercase of the incompatible flanker has a greater effect on the left than on the fight flanker.
250
Kitterle, Ludorf, and Moreland 650 640
0
c-I
I I
I-C
630 -
RT
620" 610" 600
590
I
I
I
I
S-S
S-D
D-S
D-D
Relationship of Target Case to Flanker Case Figure 3: Reaction time as a function of the left-fight case of the flanker (leftfight same case: SS, left same-right different: SD, left different-right same: DS, and both different: DD) for left compatible-fight incompatible (C-I: circles) and left incompatible-fight compatible (I-C: squares).
Experiment 3 - Target-Flanker Spacing
The results of the previous experiments have shown an asymmetrical influence of flankers on the magnitude of the FCE. The present experiment explores the effectiveness of left and right flankers on the FCE by independently varying the distance of the left or right flanker from the target letter. Based upon research that indicates that the flanker compatibility effect decreases as the distance between the target and the flankers increases, it is expected that spatial separation should have a greater influence on the FCE for left flankers than for right flankers (Eriksen & Eriksen, 1974; Eriksen & Hoffman, 1973; Harms & Bundesen, 1983; Hommel, 1995; Miller, 1991; St. James, 1990). Subjects. Forty subjects participated for course credit. They had normal or corrected to normal vision, were naive about the purpose of this study, and had not participated in the previous experiments. Apparatus and Stimuli. Stimuli consisted of uppercase letters "V" and "H", presented via a PC workstation on a 14" SVGA color monitor.
Flanker Effect
251
The letters were black on a white background. At a viewing distance of 54 cm, each letter subtended a visual angle of 30 min X 25 min. In the letter string both flankers could be response compatible, response incompatible, left compatible-right incompatible, or left incompatiblefight compatible. There were two spacing conditions: a near condition in which the distance of the outer edge of the target to the inner edge of the flanker was 5 min or a far condition in which this distance was 20 min. Levels of compatibility were completely crossed with the spacing variable yielding eight experimental conditions. Procedure. On each of the 32 practice and 64 experimental trails, subjects viewed a central fixation cross for 500 msec. After a 500 msec interstimulus interval the three letter array was presented for 500 msec. Subjects were instructed to identify the target letter by pressing one key on the computer keyboard for the letter "V" and the other for the letter "H" and to ignore the flanking letters. Mapping of the response keys "z" and "/" were counterbalanced across subjects. There was a 1500 msec intertrial interval before the next presentation. Results. The results of this experiment were analyzed by means of a 2 Response Compatibility (compatible vs. incompatible) X 4 Spacing (Contiguous, Left one space, Right one space, and both left and right one space) repeated measures ANOVA based upon correct RT. The data are presented in Figure 4. There was a highly significant main effect of Compatibility [F(1,39)= 161.09, p<.0001]. Response compatible flankers yielded faster RT than response incompatible flankers (546 vs. 597 msec, respectively). There was a significant main effect of spacing [F(3,117)=5.29, p<.002]. Contiguous flankers had a greater effect than separating both flankers from the target (583 vs 562 msec., respectively). Contiguous right flanker/separated left flanker yielded a weaker flanker effect than contiguous left flanker/separated fight flanker (566 vs. 575 msec., respectively). There was a significant Compatibility X Spacing interaction [F(3,117)=2.86, p<.04]. Fine grain analyses indicate that the simple main effect of spacing differs significantly for incompatible flankers [F(3,117)=7.87, p<.0001] but not for response compatible flankers [F< 1]. The near left-far right condition had a greater effect than the near right-far left (605 vs. 588 msec., respectively). Note also that the near left-far right condition differed significantly from the far condition when both flankers were incompatible [605 vs 582 msec; t(39)=3.20,
252
Kitterle, Ludorf, and Moreland
v --
ResponseCompatible ResponseIncompatible
I
620 600 580 RT
560 ~,
540
520
0
0
I
I
I
I
Contiguous
Left
Right
Both
Target
Flanker L e t t e r
Spacing
Figure 4: RT as a function of letter spacing for response compatible (circles) and response incompatible flankers (squares). Left (Right) indicates a separation of one letter space to the left (Right) of the target, Both indicates a space to the left and the fight, and Cont indicates that the flankers and targets were contiguous.
p<.003] but did not differ from the near condition when both flankers were compatible [588 vs. 582 msec]. The results of the present experiment support the existence of a left flanker d o m i n a n c e in the flanker c o m p a t i b i l i t y effect. As the comparisons between the left flanker near/right flanker far vs. left flanker far/fight flanker near conditions indicate, when the left flanker is separated from the target, the incompatibility effect decreases. However, when the right flanker is separated from the target, the interference effect is virtually identical to the condition in which both flankers are incompatible and near. This clearly indicates that the left flanker plays a dominant role in the FCE. General discussion The present results confirm the existence of asymmetries in the flanker compatibility effect. Specifically, we have shown that the left
Flanker Effect
253
flanker has a greater effect on the FCE than the fight flanker. Typically, in studies of the flanker compatibility effect, the target strings are sufficiently large so that the left and fight flankers fall in the left and fight visual fields, respectively. Because the left visual field projects to the fight hemisphere and the right visual field to the left hemisphere, it might be argued that asymmetries reported in this and other studies reflect hemispheric differences in the processing of visual input. A simple hemispheric explanation would posit that the greater effectiveness of the left hemisphere in processing letters should lead to greater effectiveness of the fight flanker. Alternatively, a left-side flanker effect may reflect a right hemisphere processing advantage for geometrical forms (see Hellige, 1995). However, the results of this study have shown that it is that left flanker, not the fight, that has the greater effect on the FCE. Thus, an interpretation of the differential effectiveness of the left and right flankers on target processing in terms of hemispheric differences does not appear to be supported by the present results. A hemispheric account for flanker asymmetries was also rejected by Hommel (1995) because, contrary to predictions based upon hemispheric processing differences, he found a left-side asymmetry of the flanker compatibility effect with letters. Furthermore, this effect not only disappeared with strings composed of mirrored letters and tiny pictures, but also showed a trend in the data suggesting a shift to a rightside asymmetry. However, based upon research showing that the hemisphere which dominates in a particular task depends upon characteristics of the stimulus (e.g.., spatial frequency content or local vs. global stimulus characteristics)and task demands (see Sergent & Hellige, 1986; Kosslyn, 1994; Robertson, 1995; Robertson & Lamb, 1991), a more complex interpretation of flanker asymmetries in terms of hemispheric processing may still be a viable explanation. Several studies have shown that global stimulus features are processed more efficiently in the fight hemisphere, whereas, local features are processed more effectively in the left hemisphere. Consequently, it might be argued that stimuli in the present experiment are identified on the basis of their global properties, which would lead to an attentional or processing bias to the left flanker. To fully test this theory, stimuli whose local properties were more important than its global ones would have to be presented to see if processing shifted to a bias for the right flanker/left hemisphere.
254
Kitterle, Ludorf, and Moreland
Perhaps the non-letter stimuli used by Hommel (1995) that generated a fight flanker dominance were discriminated on the basis of their local features and this was the reason for the right flanker dominance rather than that fact that they were not letters. Using circles and squares, which were forms easily discriminated on the basis of their global features, we obtained preliminary support for this hypothesis by showing a left flanker dominance in the FCE (Ludorf, Kitterle, Carpenter, & Chun, 1994). Finally, the finding that right flanker dominance can be obtained with mirror-image letters may also support this hypothesis (Hommel, 1995). Mirror-image letters, being unfamiliar, may require local rather than global processing. If this is the case, then a fight, as opposed to a left, flanker dominance would be expected. In summary, there is some suggestion that flanker asymmetries may be accounted for in terms of hemispheric differences in processing stimulus features rather than in the processing of letters versus non-letter forms. More work is needed to determine whether this hypothesis completely accounts for FCE asymmetries or whether additional mechanisms are necessary. Asymmetries in flanker effectiveness may also be accounted for in terms of attentional scanning mechanisms. Rayner and Morris (1992) showed that subjects who read in English focus their attention to the fight of the location of their gaze. Sereno and Rayner (1992) found that during reading, information from the visual field to the left of fixation is not processed, but information from the visual field to the right of fixation is processed. Based upon these results, if there is an attentional scanning mechanism learned by reading, then flankers in the fight visual field should be processed to a greater extent and have a greater influence than those in the left visual field. However, as our results indicate, the left flanker dominates in the FCE with letters. This does not necessarily rule out a scanning mechanism. For example, with briefly exposed letter strings, accuracy decreases from the leftmost to the rightmost element (Crosland, 1931). Moreover, these effects are found even when subjects report the letters in a right to left order (Mewhort, 1974). It is also found when subjects search through a letter string for a target (Krueger, 1976). On the basis of these results, Hommel (1995) suggested that scanning-like effects may more readily account for the left-flanker asymmetry with letters. Specifically, word-like structures and letter strings may more readily induce readinglike habits than other kinds of material. As a consequence, attention
Flanker Effect
255
would be attracted to elements in the left-most part of the string. This would interfere with attending to the target and permit a greater impact of the scanned elements to the left of the target. However, with tiny pictures and mirrored letters, this automatic process would not necessarily be activated and as a result the left-flanker would be abolished. Again, as noted earlier, contrary to Hommel (1995), we found a left flanker effect for forms. It may well be that with highly familiar forms, as with letters, scanning processes are automatically engaged that generally proceed from left to right, whereas with novel forms these scanning processes are not automatically engaged and alternative scanning strategies such as a random walk are invoked. Hommel (1995) has addressed the question of why scanning should take place in a task that requires the identification of a central target while ignoring the flanking letters. He has suggested that letter strings, which are similar to words, evoke a reading-like habit and as a result scanning begins from the left-most letter and proceeds to the right until the target letter is identified. This would interfere with the processing of the target and give greater weight to the scanned elements. Alternatively, incompatible flankers may induce response conflict that can only be solved by localizing the target, that is associating letter identities to spatial positions. To localize the target relative to other letters in the array, the other elements must be given some attention. If reading skills were used for localization, they might start with the leftmost element and proceed until the target position is reached. This would mean that the rightmost flankers would have little effect on target processing. Hommel (1995) used five letter strings in which the target occupied the central of position three in the string. A response compatible or response incompatible letter was placed at one of the other positions to the left or fight of the target. The remaining positions in the string, which were not filled by response compatible or response incompatible letters, were filled with response neutral letters. The results of Experiment 3 of this study indicate that the effects of flanker spacing occur even when the spaces between the target and flanker letters are not filled by response neutral letters. Moreover, the present experiment provides strong evidence that the left flanker dominates in the FCE. The interference effect from incompatible flankers is similar when both flankers are near as when the left is near and the right is far. The asymmetries in flanker spacing indicate that these effects cannot be attributed to a reduction in acuity. If the decrease
256
Kitterle, Ludoff, and Moreland
in the interference from increased spatial separation of the target from the incompatible flanker was due to a loss of acuity, then reductions in flanker effectiveness should be similar for left near/fight far and left far/right near. Moreover, the fact that the asymmetries were only found for the response incompatible and not compatible condition also indicates that reductions in acuity cannot account for these findings. It might be argued that these effects are due in part to the fact that as the space between the letters in a string increases, the activation of left-toright scanning mechanisms or strategies are not engaged because the string is less word-like. However, it should be pointed out that in other work we found a left flanker dominance using geometrical forms. As noted earlier, directional scanning should produce a greater effect of scanned elements on target processing. Given a left-to-right scanning process, the leftmost elements should have a greater effect on target processing than the rightmost. Accordingly, incompatible flankers to the left of the target should produce response conflict that can only be resolved by localizing the target, that is, associating letter identities with positions. With increasing spatial separation between target and flankers, it might be argued that inhibition between channels processing the responses to the target and those to the flanker may decrease, this may facilitate target localization. Conversely, with increasing response conflict, target location may take longer. The results of Experiment 1 support this by showing that the FCE is considerably greater when the incompatible left flanker is processed by the hemisphere that also controls the response to the target. On the other hand, the effect of the left compatible flanker, which does not give rise to response conflict, does not show any relationship between the hemisphere processing the flanking letter and controlling the response to the target. Yeh and Eriksen (1984) have shown that physical similarity plays an important role in modulating the FCE. They showed that the interference from response incompatible flankers was reduced but not abolished when the target and flanker differed in letter case. Our results (Experiment 2 ) c o n f i r m and extend those findings by showing that changing the case of the response incompatible left flanker produces a greater reduction in the FCE than changing the case of the right flanker. However, unlike with the left flanker, changing the case of the right flanker abolishes the flanker compatibility effect. There is a degree of similarity between changing the case of the flanker letter and using color to segregate the target from the flanker (Harms & Bundesen,
Flanker Effect
257
1983). Both manipulations reduce the physical similarity between the target and the flanker. It is interesting to note another similarity between the effects of case changes and color segregation between the target and flankers, namely that the more strongly a flanker interfered with the processing of a target, the greater was the benefit on processing that resulted from segregation. Thus, when there are fewer features shared between the target and the response incompatible flanker, the more easily it is to localize the target letter and to focus attention on it. Thus, it appears that spatial selectivity is enhanced under these conditions. Posner (1978) has suggested that the presentation of a familiar visual form independently activates both a physical code and a name code. The fact that FCE is less when target and flanker differ in case suggests that name codes are not activated to the same degree as physical codes. One possible way in which this could occur has been suggested by Yeh and Eriksen (1984). They hypothesize that features in the developing percept activate name codes, which when partially activated, prime or facilitate the development of the physical code. Thus, name codes are not fully activated but exert their influence solely upon the speed of processing of the physical code. This explanation is supported by the fact that in Experiment 2 there are no laterality effects resulting from changes in the case of the flanker. Clearly if the flanker interference effect reflected the contribution of name codes, there should have been greater effects of changing flanker case for right than for left visual field presentations because of the greater facility of the left hemisphere in processing name codes. The failure to find laterality effects on samedifferent letter matches led Boles (1981) to reject the hypothesis that name codes were involved in these matches. Instead, they found evidence for generation of visual or physical codes in same-different matches. However, the results of this study and earlier work (e.g., Yeh & Eriksen, 1984) indicate that both physical feature and name code of a stimulus are involved in letter discrimination. The asymmetrical effects found in the present study are not predicted by current theories of selective attention. Our results suggest that FCE asymmetries reflect strategies in target selection that result from task irrelevant flankers. Hommel (1995) has attempted to incorporate the idea of a scanning mechanism with recent theories of selective attention. For example, based upon the feature integration theory of Treisman and Gelade (1980), features are coded in parallel. If no feature gives rise to competing responses, then the correct response is
258
Kitterle, Ludorf, and Moreland
selected. However, when the flanker is response incompatible, features need to be localized and identity codes associated with these locations. Strongly developed reading habits may automatically counteract the tendency to direct attention to the center of the string and shift it to the left most elements in the display. Alternatively, the target may need to be localized relative to other elements. The automatic scanning habits may bias localization to the left most elements. Alternatively, stimulus elements may be assigned weights reflecting their similarity to an internal target description (Duncan & Humphreys, 1989). Assume that the strength of those weights can be modified by material- and subject-specific selection strategies. Since most normal letters are selected in a left-to-fight order, the weights for the leftmost letters in a string would be relatively higher that those at other positions in the string. If attention reflects a process of competition between letters that have been assigned these weights, then leftmost letters would compete more with the target selection than fight-side flankers. References
Baylis, G. C. & Driver, J. (1992). Visual parsing and response competition: The effect of grouping factors. Perception & Psychophysics, 51, 145-162. Beach, J. (1995). The effect of foreperiod and memory load on interference and asymmetry in attenhonal filtering. M. S. Thesis in Psychology, University of Texas at Arlington, December, 1995. Boles, D. (1981). Variability in letter-matcliing asymmetry. Perception & Psychophysics, 29, 285-288. Brown, V. & Carter, M. (unpublished). Lateral inhibition and the role of selective attention in response competition. Coles, M., Gratton, G., Bashore, T.R., Eriksen, C.W., and Donchin, E. (1985). A psychophysiological investigation of the continuous flow model of human information processing. Journal of Experimental Psychology: Human Perception and Performance, 11, 529-553. Davidson, R. J. & Hugdahl, K. (1995) Brain Asymmetry. Cambridge, MA: MIT Press: Deutsch, J. A. & Deutsch, D. (1963). Attention: Some theoretical considerations. Psychological Review, 70, 80-90. Dowling., C. J. & Pinker, S. (1985). The spatial structure of visual attention. In M. Posner and Marin (eds.), Attention and Performance XI, pp. 171-187. Lawrence Erlbaum Associates. Driver, J. & Baylis, G. (1991). Target-distractor separation and feature integration in visual attention to letters. Acta Psychologica, 76, 101ll9. Duncan, J. & Humphreys, G.W. (1989). Visual search and stimulus similarity. Psychological Review, 96, 433-458.
Flanker Effect
259
Efron, R. (1990). The Decline And Fall Of Hemispheric Specialization. Hillsdale, N. J.: Lawrence Erlbaum Associates, Inc. Eriksen, B.A. & Eriksen, C. W. (1974) Effects of noise letters upon the identification of a target letter in a nonsearch task. Perception & Psychophysics, 16, 143-149. Eriksen, C.W. (1995) The flanker task and response competition: A useful tool for investigatinga variety of cognitive problems. In Bundesen & Shibuya (Eds.), Visual Selective Attention. LEA, Hillsdale. Eriksen, C. W. & Eriksen, B. A. (1979). Target redundancy in visual search: Do repetitions of the target within a display impair processing? Perception & Psychophysics, 25, 249-263. Eriksen, C. W. & Hoffman, J. E. (1973) The extent of processing noise elements during selective encoding from visual displays. Perception & Psychophysics, 14, 155-160. Eriksen, C. W. & Schultz, D. W. (1979). Information processing in visual search: A continuous flow conception and experimental results. Perception & Psychophysics, 25, 24%263. Harms, L. & Bundesen, C. (1983) Color segregation and selective attention in a nonsearch task. Perception & Psychophysics, 33, 11-19. Hellige, J. B. (1993). Hemispheric asymmetry: What's right and what's left. Cambridge, MA: Harvard University Press. Helllge, J. B. & Webster, R. (1981). Case effects in letter-name matching: A qualitative visual-field difference. Perception & Psychophysics, 17, 179-182. Hommel, I3. (1995) Attentional scanning in the selection of central targets from mulfisymbol strings. In C. Bundesen and H. Shibuya (Eds.) Visual Selective Attention. Lawrence Erlbaum Associates, Hillsdale, NJ. 119-144. Hughes, H. C. & Zimba, L. D. (1985). Spatial maps of directed visual attention: Journal of Experimental Psychology: Human Perception and Performance, 4, 409-430. Kitterle, F. L. (1991). Cerebral Laterality: Theory and Research. Hillsdale, N. J: Lawrence Erlbaum Associates. Kitterle, F. L. (1995). Hemispheric communication: Mechanisms and Models. Hillsdale, N. J: Lawrence Erlbaum Associates. Kitterle, F. L. & Ludorf, M. R. (1993). Right (top) flanker dominates in the flanker compatibility effect. Poster presented at the Annual Meeting of American Psychological Society, Chicago. Kitterle, F.L., Christman, S., & Hellige, J.B. (1991). Hemispheric differences are found in the identification but not the detection of low vs. high spatial frequencies. Perception & Psychophysics, 48, 297-306. Kosslyn, S. M. (1994). Image andBrain. Cambridge, MA: MIT Press. Kramer, A. F. & Jacobson, A. (1991). Perceptual organization and focused attention: The role of objects and proximity in visual processing. Perception & Psychophysics, 50, 267-284. Krueger, L. (1976). Evidence for directional scanning with the order-ofreport factor excluded. Canadian Journal of Psychology, 30, 9-14. LaBerge, D. & Brown, V. (1989) Theory o f attention operations in shape identification. Psychological Review, 96, 101-124.
260
Kitterle, Ludorf, and Moreland
LaBerge, D., Brown, V., Carter, M., & Bash, D. (1991) Reducing the effects of adjacent distractor by narrowing attention. Journal of Experimental Psychology: Human Perception and Performance, 17, 63-76. Ludorf, M. R., Kitterle, F. L., Carpenter, W., & Chun, D. (1994). The
flanker compatibility effect: Visual field asymmetries change for letters versus forms. Paper presented at the Annual Meeting of the
American Psychological Socl-ety, Washington, D.C. Miller, J. (1986)The flanker compatibility effect as a function of visual angle, attentional focus, visual transient, and perceptual load. Perception & Psychophysics, 49, 270-288. Miller, J. (1991). The flanker compatibility effect as a function of visual angle, attentional focus, visualtransients, and perceptual load: A search for boundary conditions. Perception & Psychophysics, 49, 270-288. Mewhort, J. (1974). Accuracy and order of report in tachistoscopic identification. Canadian Journal of Psychology, 28, 383-398. Pashler, H. (1984). Evidence against late selection: Stimulus quality effects in previewed displays. J. Experimental Psychol, 10, 429-448. Posner, M. 1., Snyder, C.R.R., & Davidson, B. J. (1980). Attention and the detection of signals. Journal of Experimental Psychology: General, 109, 160-174. Rayner, K. & Morris, R.K. (1992). Eye movement control in reading: Evidence against semantic preprocessing. Journal of Experimental Psychology: Human Perception and Performance, 18, 163-172. Robertson, L.C. (1995). Hemispheric specialization and cooperation in processing complex visual p a t t e r n s . In F.L.Kitterle (Ed.), Hemispheric communication: Mechanisms and Models (pp. 301318). Hillsdale, N. J: Lawrence Erlbaum Associates. Robertson, L.C & Lamb, M.R. (1991). Neurophysiological contributions to theories of part/whole organization. Cognitive Psychology, 23, 299-330. Sereno, S.C., & Rayner, K. (1992). Fast priming during eye fixations in reading. Journal of Experimental Psychology: Human Perception and Performance, [8, 173-184. Sergent, J. (1984). Role of contrast, lettercase, and viewing conditions in a lateralized word-naming task. Perception & Psychophysics, 35, 489-498. Sergent, J. & Hellige, J. B. (1986). Role of input factors in visual-field asymmetries. Brain and Cognition, 5, 174-199. St. James, J. D. (1990). Observations on the microstructure of response conflict. Perception & Psychophysics, 48, 517-524. Triesman, A. (1964). Monitoring and storage of irrelevant messages in selective attention. Journal of Verbal Learning and Verbal Behavior, 3, 449-456. Treisman, A. & Gelade, G. (1980). A feature-integration theory of attention. Cognitive Psychology, 12, 97-136. Yeh, Y. & Eriksen, C.W. (1984). Name codes and features in the discrimination of letter forms. Perception & Psychophysics, 36, 225-233. Yund, E. W., Efron, R. & Nichols, D. R. (1990). Detectability as a function of spatial location" Effects of selective attention. Brain and Cognition, 12, 42-54.
SECTION IV: EFFECTS OF VISUAL FIELD LOCUS
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
263
Chapter 9
The Relation Between Left-Right and Upper-Lower Visual Field Asymmetries (or: What goes up goes right, while what's left lays low)
Stephen D. Christman Christopher L. Niebauer University of Toledo The development of the tachistoscopic visual half-field technique by Roger Sperry and colleagues provided a straight-forward and noninvasive means of investigating hemispheric asymmetries in normal populations. The reasoning behind this method is grounded in the fact that the temporal and nasal hemiretinae of each eye project to opposite hemispheres. Indeed, there seems to be very little overlap between the two hemiretinae of each eye; estimates of the degree of overlap between the nasal and temporal hemiretinae in humans hover around 0.25 to 0.50 degrees of visual angle (Fendrich & Gazzaniga, 1989; Harvey, 1978; Lines & Milner, 1983). Although there is even debate over whether any overlap exists (cf., Fendrich & Gazzaniga, 1989), there is clear consensus that stimuli presented more than 0.50 degrees from the vertical midline will initially project to a single cerebral hemisphere. Thus, differences in performance for stimuli presented to the left versus right visual fields (LVF vs. RVF) are interpreted as reflecting various differences in the processing competencies of the right versus left cerebral hemispheres (RH vs. LH).
264
Christman and Niebauer
This assumption is so central to how visual laterality experimentation is conducted (i.e., visual field differences are automatically interpreted as reflecting hemispheric differences) that it has become an "invisible" constraint on the interpretation of data, in much the same manner as sensory characteristics of input (e.g., luminance, retinal eccentricity, exposure duration, etc.) were treated as "invisible effects" in laterality experiments until Sergent's (1982; 1983) spatial frequency hypothesis drew researchers' attention to such factors (Hardyck, 1986). However, recent lines of research have opened up opportunities to re-examine this assumption. For example, the use of non-lateralized, free vision tasks, such as the chimeric faces task (e.g., Levy, Heller, Banich, & Burton, 1983), as well as research into the role of reference frames in visual asymmetries (e.g., Robertson & Lamb, 1988), has shown that hemispheric asymmetries are not necessarily tied directly to retinal coordinates; this in turn poses the complementary possibility that left versus right visual field differences may reflect more than just anatomically defined differences in cerebral projection. More explicitly, Robert Efron devoted much of the first two chapters of his recent book, The Decline and Fall of Hemispheric Specialization (1990), to discussion of the dangers in automatically interpreting left-right visual field asymmetries in terms of hemispheric function alone; the core of his argument is that hemispheric specialization is only one of many determinants of visual field differences, yet is almost invariably viewed as the sole cause of leftright differences. Efron's distrust of the cerebral basis of many, if not all, visual field asymmetries was based on data from his lab indicating that visual field differences arise from attentional scanning biases (e.g., stimuli from the right visual field are scanned and processed first, giving them an advantage arising from the rapidly decaying neural representation associated with tachistoscopic p r e s e n t a t i o n ) a n d not from any underlying functional specialization of the hemispheres for processing the specific type of stimulus at hand. A related cause for suspicion of a non-hemispheric basis for leftright visual field differences comes from a 1990 article in Behavioral and Brain Sciences by Fred Previc, which presented an integrative account of functional specialization in the lower versus upper visual fields. Based on ecological and neurophysiological considerations, Previc argued that the upper visual field is specialized for local, high spatial frequency, and visual search processing, while the lower field is
Visual Field Asymmetries
265
specialized for global and low spatial frequency processing. These specializations presumably reflect the different ecological roles of near versus far vision (near space: perception of optically degraded images, visuomotor coordination; far space: visual search), which in turn are predominantly represented in the lower versus upper visual fields, respectively. As Bryden and Underwood (1990) point out in their commentary on Previc's article, the functional specializations that Previc ascribes to the upper versus lower fields map quite nicely onto right versus left visual field specializations, respectively. Indeed, Bryden and Underwood note that Previc's article often sounds like a fairly conventional laterality article, except that the visual world has been twisted by 90 degrees. Two possibilities exist concerning the relation between upper-lower and left-right visual field asymmetries. First, the correspondence between upper and right field processing, on the one hand, and lower and left field processing on the other, may simply be a superficial coincidence, in which case the hemispheric interpretation of left-right asymmetries may stand unaltered. Alternatively, the correspondence may reflect a deeper, functional connection between upper-lower and left-right visual field specializations. Since upper-lower visual field asymmetries cannot reflect hemispheric differences (at least not in any simple and direct way, although the upper and lower visual fields do project to distinct cortical areas, with the upper versus lower visual fields projecting preferentially to the ventral [occipito-temporal] and dorsal [occipito-parietal] visual pathways; see Cragg, 1969, and Zeki, 1969), the question is raised whether left-right differences are not just analogous but also homologous to the upper-lower differences, in which case many left-right differences may n o t reflect hemispheric differences, but rather may reflect other factors. This chapter will be divided into two parts. The first will present a review of upper-lower visual field differences across a variety of tasks, with special emphasis on those task and input factors which have been specifically implicated in studies of hemispheric differences. The second part will offer tentative explanations for the correspondences between upper-lower and left-right visual field differences, allowing an evaluation of the extent to which researchers are safe in interpreting leftfight differences predominantly in terms of hemispheric function versus in terms of non-hemispheric factors.
266
Christman and Niebauer
Correspondences Between Lower/Upper and Left/Right Visual Field Processing An exhaustive survey of functional and anatomical differences between the upper and lower visual fields is beyond the scope of this chapter; interested readers should consult the thorough reviews by Previc (1990) and Skandries (1987). This chapter will restrict its focus to those task and input factors that have been investigated with respect to stimulus placement along both the vertical and horizontal meridians of the visual field.
I. Simple Reaction Time Payne (1967) reported a lower visual field advantage (VFA) for simple reaction time (RT); unfortunately, his method of presenting the data focused on comparisons between the nasal and temporal hemiretinae, precluding a direct comparison of the left versus right visual fields. Similar lower VFAs for simple RT have been reported by Rizzolatti et al. (1987) and by Gawryszewski et al. (1987), among others. Laterality research has not revealed any widely accepted left-right visual field differences in simple RT; however, given the clear ecological differences in information that an organism may detect in the upper versus lower visual fields (cf., Previc, 1990), and the lack thereof for the left versus right visual fields, this may not be surprising. That is, any organism showing a systematic bias towards poorer detection performance in one of the lateral visual fields would be at risk of predators' learning to take advantage of this bias; indeed, fossil trilobites reveal a systematic pattern of scars on the right side of the head (Babcock & Robison, 1989), possibly arising from such a predatory bias. Furthermore, presentation of stimuli to the left and fight visual fields introduces potential interactions between the hemisphere that perceives the stimulus versus the hemisphere that produces the response, thereby obscuring whether left-right differences arise from perception versus responseprocesses. For example, Filbey and Gazzaniga (1969) reported a right VFA for dot detection using a verbal response; the visual field asymmetry, however, disappeared with the use of a motor response. Nonetheless, there appears to be tentative evidence for a left VFA in simple RT. Davidoff (1977) reported left VFAs in dot detection for both RT and threshold exposure duration (but for males only). Similarly,
Visual Field Asymmetries
267
Umilta et al. (1979) also reported a left VFA in a dot detection task. However, any conclusion regarding left-right differences must be taken with caution, as the differences are often quite small, limited to certain subject groups (e.g., Davidoff, 1977), or even absent (e.g., McKeever, Gill, & VanDeventer, 1975).
II. Resolution/Acuity There is tentative evidence that the visual field asymmetries in simple RT discussed in the previous section are a function of certain scales of spatial resolution. For example, data from a number of studies indicates that the lower VFA in simple RT is confined to the low-to-moderate spatial frequency range (e.g., Lundh, Lennerstrand, & Derefeldt, 1983; Murray, MacCana, & Kulikowski, 1983; Rijsdik, Kroon, & van der Wildt, 1980); upper-lower visual field differences at high spatial frequencies seem to be attenuated or absent. Acuity measures (reflecting the finest resolvable details which are carried by high frequencies) yield a somewhat different picture, with some studies reporting lower VFAs in acuity (e.g., Finke & Kosslyn, 1980; Millodot & Lamont, 1974) and other reporting upper VFAs (e.g., Julesz, Breitmeyer, & Kropfl, 1976). However, acuity measures are at best only an approximation of sensitivity to high spatial frequencies (cf., DeValois & DeValois, 1988). Furthermore, there is a possible confounding between performance based on acuity versus hyperacuity (which reflects stimulus localization ability, which in turn may be better in the lower visual field; see section below on Categorical/Coordinate Processing). The evidence is less consistent and robust concerning left-right differences in spatial resolution. In terms of contrast sensitivity, there is little evidence for left-right differences. Numerous studies have reported no significant lateral field differences in contrast sensitivity (e.g., Blake & Mills, 1979; Fiorentini & Berardi, 1984; Kitterle, Christman, & Hellige, 1990; Kitterle & Kaye, 1985; Rao, Rourke, & Whitman, 1981), although it is worth mentioning that the studies by Blake and Mills (1979) and Kitterle et al. (1990) found nominal left VFAs for low, relative to high, frequencies. Fiorentini and Berardi (1984) also report data indicating nominal left versus right VFAs for low versus high frequency detection, respectively, although this effect was confined to left eye viewing; fight eye viewing led to nominally greater sensitivity in the right visual field at all frequencies (equivalent comparisons for the
268
Christman and Niebauer
studies by Kitterle & Kaye and by Rao et al. are not possible because of hemiretinal confounds and/or a lack of explicitly presented data). RT measures of left-fight visual field differences tell a similar story: neither Kitterle et al. (1990)nor Previc (1982) reported any visual field differences as a function of spatial frequency; however, Kitterle et al. (1990) reported significant left VFAs in RT across all frequencies tested for both threshold and suprathreshold detection which may be related to the visual field differences in simple RT discussed in the previous section. Also, it is worth noting here that, although there is little evidence for left-fight asymmetries as a function of spatial frequency in the type of detection tasks reviewed here, there is clear evidence for left versus right VFAs in the identification of low versus high frequencies, respectively (Kitterle, Christman, & Hellige, 1990). Finally, while relatively less attention has been devoted to upper/lower visual field differences in spatial frequency processing, Pointer and Hess (1989) reported upper-lower (as well as left-right) visual field symmetry in contrast sensitivity for both high and low spatial frequencies. Hemifield differences in temporal resolution have also been reported (also see section below on Motion processing). The general consensus regarding upper-lower differences is that a lower VFA exists for temporal resolution as measured by the critical flicker fusion technique (e.g., Hylkema, 1942; Singer, 1985; Tyler, 1987). The picture with regard to left-fight differences is somewhat different. Rao et al. (1981) reported left versus right VFAs in thresholds for low (e.g., _< 2 Hz) versus high (e.g., >__4 Hz) temporal frequencies, respectively. Blake and Mills (1979) reported no left-right differences for 3.5 Hz flicker thresholds. Christman (1988) also reported no left-fight differences in a two-pulse temporal gap detection task. Mecacci and colleagues have reported data indicating left versus right VFAs in evoked potential magnitude to low versus high temporal frequencies, respectively (e.g., Mecacci, 1993; Rebai, Mecacci, Bagot, & Bonnett, 1989; Spinelli & Mecacci, 1990). However, their data also indicates interactions between visual field, temporal frequency, and spatial frequency. For example, Mecacci (1993) reports greatest right visual field-left hemisphere EPs to low spatial frequencies presented at high temporal frequencies; the largest left visual field-right hemisphere EPs were obtained for high spatial frequencies presented at low temporal frequencies. Alternatively, Rebai, Bagot, and Viggiano (1993) reported that the fight visual field-left hemisphere advantage for high temporal
Visual Field Asymmetries
269
frequencies was constant across a range of spatial frequencies, whereas left visual field-right hemisphere responses were a function of both spatial frequency and mode of stimulation (i.e., larger amplitudes to low temporal frequencies at lower versus higher spatial frequencies for reversal versus onset stimulation, respectively). Thus, with regards to spatial resolution, lower VFAs for lower frequencies appears to be fairly well established; the evidence is quite mixed, however, regarding upper-lower differences in high frequency processing. With regards to the left versus right fields, differences in sensitivity as a function of spatial frequency, if they exist, are very small, although there may be hints of left VFAs at low spatial frequencies. In terms of temporal resolution, there appear to be lower and right VFAs for high temporal frequencies, and left VFAs for low temporal frequencies, which are qualified, however, by effects involving spatial frequency and method used to assess temporal processing. Three last effects potentially related to resolution are worth noting. First, Pennal (1977) reported better color matching ability in the lower left visual field relative to the other three quadrants. Previc (1990), however, argued that these results may reflect temporal processing differences, as the extremely short exposure duration of 30 msec employed by Pennal may have prevented the sluggish opponent-color channels from being activated, thereby transforming the task into a luminance matching task. Second, Edgar and Smith (1990) employed a task in which observers matched the spatial frequency of pairs of gratings, with one presented to the left and one to the right visual field, or with one presented to the upper and one to the lower visual field. They reported that the spatial frequency of stimuli presented to the left and lower visual fields was overestimated relative to stimuli in the right and upper visual fields. Finally, Berardi and Fiorentini (1991) reported lower and left VFAs in a spatial phase discrimination task (phase refers to the absolute position in space of a spatial frequency component). Given that phase judgments involve the precise localization of visual stimuli, these lower and left VFAs may be related to lower and left VFAs in processing coordinate spatial relations (see below).
III. Local-Global Processing The presence of left versus right VFAs for the processing of global versus local levels of form, respectively, is a widely accepted finding
270
Christman and Niebauer
among laterality researchers (e.g., Sergent, 1982; Van Kleeck, 1989). Based on his analysis of the types of visual information required for typical processes associated with near space/lower visual field (e.g., diplopic images, degraded input) versus far space/upper visual field (e.g., need for high resolution in conjunction with visual search and object recognition), Previc (1990) predicted lower versus upper VFAs for global versus local processing, respectively. Christman (1993) confirmed this prediction, reporting upper (and right) versus lower (and left) VFAs for local versus global processing, respectively. Interestingly, across both experiments reported by Christman (1993), the magnitude and reliability of the upper/lower differences were larger than those for left-right differences. While there was no evidence for additivity of these effects (e.g., global processing was not significantly better in the lower left quadrant), this lack of effect may have simply reflected a lack of power, as there were only 12 observations per subject per trial type per location. There were hints of additivity of such effects, however, as processing of local information in the lower left quadrant was nominally slowest and yielded the second highest error rate out of 8 locations; similarly, global processing was most accurate in the lower left quadrant (this distinctive nature of lower left field processing, relative to the other quadrants, will resurface in the section below on categorical/coordinate processing). The data on visual field differences in local-global processing thus display a similar pattern to the data on acuity/resolution: upper and right VFAs for local processing versus lower and left VFAs for global processing. This correspondence probably reflects in part the relationship between spatial frequency content and local versus global levels of form: local versus global structure is carried predominantly by higher versus lower frequencies, respectively (e.g., Kitterle, Christman, & Conesa, 1993; Shulman, Sullivan, Gish, & Sakoda, 1986). Thus, it is still not clear as to the extent to which visual field differences in local-global processing reflect differences in the processing of spatial frequency versus hierarchical structures p e r se.
IV. Categorical/Coordinate Processing Niebauer and Christman (1997) suggested that Previc's hypothesized specializations for the upper and lower visual fields bore a systematic relation to Kosslyn's (1987) theory of categorical versus coordinate
Visual Field Asymmetries
271
ss~l 540
RT
530
( m s e c ) 520 510 500
Uppert.eft
Uppe~ght LowerLeft LowerRight Visual Field _
Figure 1. Data from Niebauer and Christman (1997). Reaction time on Categorical task as function of visual field quadrant.
650 640 RT 630 (msec) 620 610 600 Upped_eft
Upperl~ght
LowerLeft LowerRight
V i s u a l Field Figure 2. Data from Niebauer and Christman (1997). Reaction time on Coordinate task as function of visual field quadrant.
representations of spatial relations. In particular, they argued that the requirements of visuomotor coordination (associated with the lower visual field in Previc's theory) are consistent with the nature of coordinate representations, in which the precise position of objects and parts is represented; alternatively, the requirements of object identi-
272
Christman and Niebauer
fication (associated with the upper visual field in Previc's theory) have been specifically linked by Kosslyn to categorical representations, which represent structural invariances. Accordingly, Niebauer and Christman predicted and obtained upper and right versus lower and left VFAs for categorical versus coordinate judgments, respectively (see Figures 1 and 2). Interestingly, these field differences were driven entirely by performance in the lower left visual field, which was fastest for coordinate judgments and slowest for categorical; there were no differences between the upper right, upper left, and lower right visual fields within either task condition. Thus, the left-right differences were confined to the lower visual field, and the upper-lower differences were confined to the left visual field, suggesting a functional linkage between left-fight and upper-lower field differences. The presence of left versus right VFAs for coordinate versus categorical processing is well established (e.g., Hellige & Michimata, 1989; Kosslyn et al., 1989; see also chapter by Hellige, this volume). Thus, we once again find a systematic relation between upper/lower and left/right differences, with upper and right VFAs for categorical processing, and lower and left VFAs for coordinate processing.
V. Stereopsis Stereopsis can be measured via stereoacuity tasks (which rely primarily on high spatial frequency, local correspondences)or via detection of random-dot stereograms (which rely primarily on low frequency, global correspondences). Paralleling the results for resolution/acuity, no upper-lower visual field differences are found for stereoacuity (Richards & Regan, 1973). Use of stereograms has revealed upper versus lower VFAs for uncrossed, divergent (far) versus crossed, convergent (near) disparities, respectively (e.g., Breitmeyer, Julesz, & Kropfl, 1975; Julesz, Breitmeyer, & Kropfl, 1976). However, Manning et al. (1987), using a procedure that they claimed was a more sensitive test of stereosensitivity than that employed by Julesz and colleagues, found no upper-lower differences; at the very least, their results indicate that visual field anisotropies in stereopsis are influenced in complex ways by such factors as stimulus size, retinal eccentricity, and the disparity of targets relative to the background field. While there have been reports of left VFAs for stereopsis for both crossed and uncrossed disparities (e.g., Carmon & Bechtoldt, 1969,
Visual Field Asymmetries 273 Durnford & Kimura, 1971; Grabowska, 1983), Breitmeyer et al. (1975) suggest that this effect may have arisen from the use of geometric patterns, thereby raising the question of whether such results arose from left-fight visual field differences in pattern recognition versus stereopsis (Breitmeyer et al.'s task involved simple detection). Other studies not open to such criticisms have also revealed left VFAs in stereopsis (Dimond, Bures, Farrington, & Brouwers, 1975) or shown superior stereopsis in patients with left, relative to right, hemisphere damage (Danta, Hilton & O'Boyle, 1978; Hamsher, 1978). However, these studies employed various localization tasks that may have also have preferentially engaged RH processing. Alternatively, therefore, given (i) evidence for left VFAs in low spatial frequency processing (e.g., Kitterle et al., 1990) along with (ii) the use of coarse-to-fine binocular correspondence strategies (in which output for low spatial frequency channels guides the matching of output for higher frequency channels; see Marr, 1982; Prazdny, 1987), these left VFAs might reflect left field superiority in low spatial frequency processing, not stereopsis p e r se. Previc, Breitmeyer, and Weinstein (1995)examined discriminability of random dot stereograms in the four primary quadrants of the visual field. They reported upper and fight VFAs in RT and accuracy which were qualified by various interactions. First, the upper VFA for both RT and accuracy held only for far, not near, targets; second, the upper VFA for RT was much greater in the LVF than in the RVF, especially for far targets. Thus, the lower left quadrant of the visual field was particularly impaired at resolving far targets, consistent with Previc's (1990) hypothesis of a special role of the lower visual field in near vision. To summarize the results for stereopsis: there appear to be robust upper-lower visual field differences in stereoscopic vision, with upper versus lower VFAs for far (uncrossed disparities) versus near targets (crossed disparities), which map nicely onto the different functional roles of near versus far vision as discussed by Previc (1990). There do not appear to be robust left-fight visual field differences in stereopsis, however; the trend for left VFAs may reflect the importance of low spatial frequencies in binocular matching. Indeed, given the lack of ecological differences between visual input to the left versus fight visual fields (in contrast to the clear differences between upper and lower), it would not appear to be evolutionarily adaptive for an organism to exhibit strong left-fight biases in stereoscopic vision.
274
Christman and Niebauer
VI. Motion
Studies of "short-range" motion (e.g., employing small targets moving at slow velocities) have typically not yielded evidence for upperlower visual field differences (e.g., Anderson, Mullen, & Hess, 1989; Regan & Beverly, 1983; Smith & Hammond, 1989). While Smith and Hammond (1989) reported robust hemifield differences within individuals, with many subjects showing clear upper-lower and left-right differences in perceived velocity, the direction varied from subject to subject, so that no overall group hemifield differences emerged. Studies of more global forms of motion perception, such as those involving motion in depth, do yield upper-lower differences. For example, Regan, Erkelens, & Collewijn (1986) reported that perception of motion in depth was superior in the lower visual field. Edwards and Badcock (1993) also reported greater sensitivity to motion in depth for the lower visual field. In a study examining the perception of looming stimuli, Regan and Vincent (1995) reported that the upper visual field, relative to the lower, left and right fields, was inferior at judging the time to contact. A related finding of larger magnitude optokinetic nystagmus elicited by lower, relative to upper, field motion is also consistent with a general lower VFA for motion perception (Murasugi & Howard, 1989). Interestingly, Raymond (1994) reported that this lower VFA was restricted to upward (i.e., centripetal) motion; there were no upper-lower differences for downward motion. This pattern of lower VFAs in various aspects of motion perception has been linked to the role of the lower visual field in the guidance of locomotion and visuomotor coordination (cf., Previc, 1990), and probably accounts for related findings of lower VFAs. For example, Foley and McChesney (1976) found that ambulatory subjects who had adapted to prismatic displacements of their visual input exhibited a bias to selectively utilize information from their lower visual field. Similarly, Guez et al. (1993) reported that the pseudo-fovea in patients with central scotoma is systematically displaced to the lower (and/or left) visual field, which they hypothesize is beneficial for locomotion. Murasugi and Howard (1989) reported that optokinetic nystagmus, a series of conjugate eye movements that is important in stabilizing the image of a moving world on the retina, is elicited more strongly from lower field stimulation. Finally, it is worth mentioning that this lower VFA for aspects of motion perception may be common across many species; for
Visual Field Asymmetries
275
example, praying mantis' exhibit greater behavioral responsiveness in predatory behavior to stimuli moving in their lower, relative to upper, visual field (Prete, 1993). There has been less attention paid to left-right visual field differences in motion perception, presumably reflecting technical difficulties in presenting dynamic stimuli within the 150-200 msec exposure duration employed by most laterality studies. The few studies that have examined this issue yield rather mixed results. As mentioned above, Smith and Hammond (1989) found no group level differences in velocity perception between the left and right visual fields, although individual subjects did display consistent lateral biases. Calvert (1988) reported faster and longer motion aftereffects for stimuli presented to the fight visual field; this result was interpreted as reflecting a greater role of "sustained" responding in the LH. Kostelyanets et al. (1992) reported a left VFA in sensitivity to centripetal motion; furthermore, the left field was more sensitive to centripetal motion than centrifugal, whereas the right field was equally sensitive to both types of motion. This mirrors the results of Raymond (1994), who reported that the lower field was more sensitive to centripetal motion than centrifugal, while the upper field was equally sensitive to both types of motion. Finally, Christman (1997a) presented observers with a dot that grew or shrank slowly or rapidly; right VFAs were obtained for "grow/shrink" judgments (which were hypothesized to involve categorical spatial relations), while left VFAs were obtained for "fast/slow" judgments (which were hypothesized to involve coordinate spatial relations). Taken together, these studies do not suggest any overall left-fight differences in motion perception, however, there may well be (i) consistent left-right differences within individuals, and (ii) left-right differences in processing specific aspects of motion (e.g., rate versus direction of motion, centripetal versus centrifugal motion). To further complicate matters, there is evidence that left-fight visual field differences in motion perception, especially directional biases, are influenced by reading habits (e.g., Morikawa & McBeath, 1992); thus centripetal versus centrifugal biases are in direct correspondence or opposition with reading habits in the left and right visual fields, but are orthogonal to reading biases in the upper and lower visual fields, complicating the comparison between the vertical and horizontal meridians. In summary, the data concerning differences between the visual fields in motion perception are somewhat equivocal, reflecting the
276
Christman and Niebauer
heterogeneous nature of the different dimensions of motion and the various methods used to assess its perception. Nonetheless, there is tentative evidence for lower and left VFAs in many aspects of motion perception, especially for centripetal and global motion. While further research would be helpful, the short durations associated with tachistoscopic stimulus presentation place inherent limits on the use of dynamic stimuli, as the typical duration maximum of 200 msec limits the visual system's ability to extract motion information. VII. Visual Search
Studies of visual search have yielded strong evidence for both upper/lower and left/right asymmetries. Robert Efron, Bill Yund, and colleagues have presented a systematic series of experiments demonstrating upper and/or right VFAs in visual search tasks (e.g., Chaiken, Corbin, & Volkmann, 1962; Efron, Yund, & Nichols, 1987, 1990a, 1990b; Yund, Efron, & Nichols, 1990; see also chapter by Yund, this volume). Their work has replicated and extended previous research (e.g., Chedru, Leblanc, & Lhermitte, 1973) demonstrating that visual search tends to begin in the upper visual field and proceeds from left-tofight. An unresolved issue, however, concerns where in the visual array search is initiated. Chedru et al. (1973) interpreted their results as indicating that search begins in the upper left quadrant and proceeds in a clockwise direction; however, the data of Efron et al., as well as that reported by Christman and Naegele (1995) suggests that the left-to-right search is initiated from the vertical meridian containing the fixation point, such that stimuli in the upper and right visual fields are at an advantage for primacy in scanning. Arguing that Efron and Yund's results may have arisen from pattern recognition requirements (i.e., their target typically consisted of a moderately high spatial frequency square wave grating, which would be expected to yield right, and possibly upper, VFAs on that basis alone), Christman and Naegele (1995) employed a visual search paradigm closely modeled after that developed by Anne Treisman (e.g., Treisman & Gelade, 1980), involving simple target features (e.g., color and shape) for which no a priori hemispheric asymmetries were expected, and also obtained upper and fight VFAs in visual search (see Figures 3 and 4).
Visual Field A s y m m e t r i e s
600 -
rl
Upper Visual Field
--
Lower Visual Field
277
590 580 RT (msec)
570 560
-
550 540
I
I
I
Left
Medial
Right
Location
within
Visual
Field
F i g u r e 3. Data from Christman and Naegele (1995). Visual search RT as a function of location within the upper and lower visual fields.
r~,., ,IL
1.
600 -
Left Visual Field Right Visual Field
580 RT
560
(msec)540 520 500
I
I
I
Upper
Medial
Lower
Location
within
Visual
Field
F i g u r e 4. Data from Christman and Naegele (1995). Visual search RT as a function of location within the left and fight visual fields.
278
Christman and Niebauer
These field differences appear to arise from scanning biases as such, and not from visual field differences in feature identification nor feature integration, as there is also no evidence for visual field differences in illusory conjunctions (e.g., Christman & Naegele, 1995; Eglin, 1987; Miossec, Kolinsky, & Morais, 1993). It is important to keep in mind, however, that potential hemispheric asymmetries in the processing of visual features that define the targets in search tasks can produce complex interactions with left/right visual field factors (e.g., Polich, DeFrancesco, Garon, & Cohen, 1990; Polich, Lentz, & Crossman, 1986). Previc and Blume (1993) examined visual search in three spatial dimensions (left/right, upper/lower, near/far) and reported RT advantages for upper over lower, fight over left, and far over near conditions. There was no interaction between performance along the vertical versus lateral dimensions; however, performance in the lower left quadrant was nominally slowest and least accurate. Furthermore, echoing findings of Chaiken et al. (1962) and Christman and Naegele (1995), the upperover-lower advantage was nominally larger than the corresponding right-over-left advantage. Importantly, Previc and Blume argue against processing differences p e r se between the various quadrants of the visual field; rather, they interpret their results as reflecting differences in the speed with which focal attention and eye movements can be shifted from one potential target to the next during serial search. This interpretation was confirmed by Previc (1996), who reported that upper and right VFAs in visual search are reduced in less attention demanding situations (e.g., simple, relative to conjunctive, search; fewer distractors); oculomotor biases favoring the upper visual field, along with inefficient saccadic search strategies in the lower left quadrant, were also present. In addition, the upper/lower and left/fight differences summated linearly so that visual search performance was best in the upper right quadrant. An issue related to visual search concerns the size of the visual lobe, defined as the eccentricity contour that yields equivalent detection or recognition performance in visual search tasks. Sanders and Brtick (1991) had subjects search for a letter "c" among a background of "x"s, and examined how the size of the visual lobe increased with increased exposure duration. They reported that the visual lobe extended farther into the fight visual field than into the left, which in turn had a greater extent than the upper and lower boundaries, which did not differ. However, an interaction between duration and direction reflected that fact that increases in duration preferentially favored expansion of the
Visual Field Asymmetries 279 visual lobe into the upper and right visual fields. It should be pointed out, however, that other studies employing unlimited display durations have reported symmetry of the visual lobe about the vertical and horizontal meridians (e.g., Courtney & Chan, 1986). A final set of potentially relevant findings involve visual field differences in saccadic latency. Eye movements are faster to both upper (e.g., Heywood & Churcher, 1980; Honda & Findlay, 1992) and right visual field targets (e.g., Hutton & Palet, 1986). Similarly, Kinsbourne (1972) reported that subjects tend to elevate their eyes when engaged in the scanning of information in memory. Such a pattern is consistent with the aforementioned upper and right VFAs in visual search, as targets that are explicitly scanned first have an obvious advantage. It is important to note, however, that visual field differences in visual search cannot be simply ascribed to visual field differences in saccadic latency, as many of the experiments reporting visual field differences employed exposure durations too brief to allow eye movements (e.g., Christman & Naegele, 1995; Efron et al., 1990a, 1990b). From this perspective, the saccadic latency data probably reflects a deeper attentional bias towards the upper and fight fields, rather than the search biases simply reflecting the saccadic latency data. VIH. Visual Attention
The visual field differences in visual search reviewed in the preceding section are related to more general differences in attentional processing. The most dramatic example involves hemineglect, an inability to attend to stimuli in certain locations following (typically right) parietal lobe damage. Neglect is usually most pronounced in the lower and/or left visual fields (e.g., Colombo, De Renzi, & Faglioni, 1976; Rapcsak, Cimino, & Heilman, 1988; Rubens, 1985). Although the lateral dimension of neglect (i.e., left neglect) has received the most attention, lower visual field neglect is a common conjoint symptom; indeed, in 13 out of 18 patients described by Rubens (1985), neglect was limited to the lower field. Furthermore, the horizontal and vertical dimensions appear to interact in that left neglect is often most severe in the lower visual field (e.g., Ladavas, Carletti, & Gori, 1994). In many ways, the findings from neglect patients provide a counterpoint to the findings on visual search in normals, as the upper and right VFAs in the latter can be seen as being greatly exaggerated in the former.
280
Christman and Niebauer
Other measures of attentional processing provide a less clear picture. For example, one of the best developed models of attention (e.g., Posner, 1992) breaks attention down into three component processes: disengagement of attention from its current focus, movement of attention across space, and engagement of attention to a new location or object. Posner et al. (1984) argued that the symptoms exhibited by neglect patients involve difficulty with disengaging attention; however, they reported no differences in this ability as a function of side of parietal damage in neglect patients. Unfortunately, there do not appear to be any studies of these functions comparing upper versus lower field performance. There is evidence, however, that attention can be shifted more quickly from central fixation to upper, relative to lower, visual field targets (Gunter, Wijers, Jackson, & Mulder, 1994), although Tsal (1989) reported no left-right or upper-lower asymmetries in a letter naming task employing various types of precuing of target location. In this sense, it is possible that the visual search asymmetries described in the previous section reflect visual field differences in the movement of attentional foci across both meridians of the visual field; this conclusion is highly tentative, however, and further work is needed. Related research (Rayner, Well, & Pollatsek, 1980) on the size of the effective visual field during reading indicates a left-right asymmetry, with available information extending farther to the fight of fixation (i.e., about 15 characters) than to the left (about 3-4 characters). However, this asymmetry clearly reflects biases associated with reading directionality, as Hebrew readers exhibit asymmetric effective visual fields that are biased towards the left of fixation. Interestingly, effective visual field during vertical language reading appears to be symmetrical about fixation (Osaka & Oda, 1991), suggesting that the asymmetries obtained with horizontal reading reflect either (i) hemispheric specific factors and/or (ii) attentional scanning mechanisms that operate differently for horizontal vs. vertical directions. Finally, Berlucchi, Tassinari, Marzi, & Di Stefano (1989) examined the phenomenon of inhibition of return, in which presentation of an extrafoveal visual cue has an inhibitory effect on RT to a subsequent extrafoveal visual target. Experiment 1 showed that this effect is much larger when cue and target both appear on the same side of the vertical meridian (i.e., both in the left or both in the right visual field). The authors point out that the standard interpretation of this effect would invoke hemisphere-specific attention mechanisms in which the
Visual Field Asymmetries 281 hemisphere orienting to the initial cue experiences an inhibitory refractory period, leading to slower RT to subsequent targets presented to the same visual field. However, the authors obtained the same visual field specificity in Experiment 2, which compared upper versus lower visual field performance: a cue presented to one side of the horizontal meridian led to inhibition of processing of targets presented to the same side. Thus, the authors point out that the patterns of within-hemifield inhibition cannot be attributed to hemispheric mechanisms, but rather appear to reflect patterns of attentional allocation within, but not across, the major horizontal and vertical meridians of the visual field.
IX. Pattern Recognition Before discussing the issue of visual field differences in pattern recognition, it is important to keep in mind that such differences may often reflect visual field differences in lower-order sensory and perceptual processing (e.g., in terms of resolution, categorical versus coordinate processes, visual search, etc.), rather than in pattern recognition p e r se. The complex interactions among such factors can make it difficult to unambiguously ascribe VFAs to pattern recognition as such. Furthermore, given the very large scope of studies examining left-right visual field differences in pattern recognition, combined with the rather limited scope of studies examining comparable upper-lower visual field differences, the material reviewed in this section is by no means exhaustive and is intended rather to simply highlight some intriguing possible connections. With those caveats aside, a number of interesting visual field differences in pattern recognition can be considered. First, one of the first experiments to employ the tachistoscopic paradigm dominant in most current visual laterality work reported both lower and right VFAs in word recognition (Mishkin & Forgays, 1952); note that this linking goes against the trend evident in many of the preceding sections of this chapter for fight VFAs to be associated with u p p e r VFAs. However, it is quite possible that the VFAs obtained in this study did not reflect word processing as such, but rather arose from attentional scanning biases like those discussed in the section on visual search. For example, Heron (1957) presented four letters arranged in a square to the left versus right visual fields; besides finding an overall right VFA, an advantage for the two upper letters within each square over the two lower letters
282
Christman and Niebauer
presumably reflected scanning biases within the stimulus array (this upper over lower advantage held equally for left and right field arrays). In this case, scanning presumably began at the upper left corner of the square, and following reading habits, continued through the upper right, then lower left, and finally lower fight corners. Conversely, in the Mishkin and Forgays study, scanning, again following reading habits, may have commenced at the fixation point and proceeded rightward and downward to the unitary word stimuli, leading to right and lower VFAs. Of course, this is quite speculative, but the point is highlighted that field differences in pattern recognition, especially for verbal stimuli, may be partly determined by scanning biases, which in turn may be partly driven by reading habits. For example, Ostrosky-Solis, Efron, & Y und (1991) reported that the right visual field scanning advantage in visual search is attenuated in illiterate subjects (although overall visual search performance did not differ between literate and illiterate subjects). Worrall and Coles (1976) examined letter recognition at various points in the visual field corresponding to the positions on a clockface. Their data indicated an overall trend toward a left VFA (which may have reflected the extremely short exposure duration of l0 msec and low luminance level of 0.:5 Ix, both of which limit the visual system's ability to extract high spatial frequency information and would therefore put the right visual field/left hemisphere at a disadvantage). Interestingly, however, a significant right VFA was found along the horizontal meridian only. As the authors point out, this restriction of the fight VFA to the horizontal meridian only argues strongly against an interpretation of the results in terms of hemispheric asymmetry. Given that stimuli are typically displayed along the horizontal meridian in most laterality experiments, one is left to ponder what would happen to all the significant fight VFAs that have been reported for verbal material if the stimuli had been displaced slightly along the vertical dimension away from the horizontal midline. Three other studies of pattern recognition are worth mentioning. Schwartz and Kirsner (1982) employed a letter-matching task in which subjects judged whether two letters were either physically identical (e.g., AA) or had the same name (e.g., Aa). They reported upper and right VFAs for name identity matches, and a lower visual field advantage, but no left-right differences, for physical identity matches. Levy and Kueck (1986) had subjects search for rhyming words out of an array
Visual Field Asymmetries 283 containing 213-327 typed words; thus, their task involves both visual search and pattern recognition. They reported both upper and right VFAs; furthermore, lateral and vertical location interacted such that there were no left-right differences in the upper visual field and a significant fight VFA in the lower field. Finally, Beer, Gallaway, and Previc (1996) reported upper and right visual field advantages in the recognition of airplane silhouettes; conversely, the lower left quadrant tended to exhibit the worst performance (although this effect was of marginal significance in Experiment 1 [p<.08], it was significant in Experiment 2 [p<.01 ]). CONCLUSIONS From an empirical perspective, the experiments reviewed above indicate that functional asymmetries between the upper and lower visual fields are at least as prevalent and robust as left-right visual field asymmetries; furthermore, there is a prevailing trend for processes that yield upper VFAs to yield fight VFAs, and processes yielding lower VFAs tending to yield left VFAs. Upper and right VFAs are obtained for the processing of local structure, categorical spatial relations, and visual search, while lower and left VFAs are obtained for the processing of global structure, coordinate spatial relations, and global motion; the lower and left visual fields have also been selectively implicated in such clinical phenomena as attentional neglect, optokinetic nystagmus, and position of the pseudo-fovea in patients with central scotoma. With regard to visual field differences in pattern recognition, the extremely limited and selective review of the literature reveals a mixed story, with some studies reporting lower and right VFAs in word recognition (although such results may have more to do with attentional search than with word recognition per se), and others reporting upper and right VFAs in letter processing and word search. The only domains reviewed that failed to provide consistent evidence of this upper-right versus lower-left visual field linkage were simple RT, acuity/resolution, and stereopsis. Even here, there is tentative evidence suggestive of lower and left VFAs in simple RT and in the processing of low spatial frequencies. The picture with regard to temporal resolution presents the primary exception to the upper/right versus lower/left linkages, with evidence for better temporal resolution in the lower and fight visual fields; however, given evidence that visual field differences
284
Christman and Niebauer
in temporal frequency processing are modulated by such factors as the spatial frequency content of input and by presentation methods (e.g., on/off versus phase reversal), it is premature to offer any strong conclusions regarding visual field differences in temporal processing at this time. Finally, there do not appear to be visual field differences along either meridian for stereopsis, although (i) upper versus lower VFAs exist for uncrossed versus crossed disparities (reflecting ecological differences between distances, relative to fixation, of objects in the upper and lower fields), and (ii) there may be a left VFA in stereopsis for both crossed and uncrossed disparities (although this likely reflects the importance of low spatial frequency content in guiding binocular registration, rather than a left VFA in stereopsis per se). We are now ready to address the question of the basis for the systematic links between processing in the upper/lower and left/right visual fields. The consistency of these relations argues strongly against their being purely coincidental in nature; rather, the evidence suggests some common functional mechanism(s) underlying processing differences along both the vertical and horizontal meridians. Before considering specific candidates for such common mechanisms, it is important to highlight two issues. First, both the upper versus lower and the left versus fight visual fields differ in striking, albeit very different, ways" the upper and lower visual fields are strongly associated with far versus near vision, respectively, giving rise to clear ecological differences in the types of information that will typically be encountered in the upper versus lower fields; furthermore, the upper and lower fields also project to anatomically distinct neural regions. The left and right visual fields, on the other hand, are not associated with any a priori ecological differences; that is, our environment does not impose pervasive differences in the types of information that we encounter on the left versus right sides of space (with the important exception of reading). The left and fight visual fields, however, project to different cerebral hemispheres, which is, of course, the predominant mechanism invoked to explain left-right differences. Thus, to the extent that the common mechanisms are primarily ecological in nature, the potential ways in which the left and fight fields differ in the types of functions and input they receive need to be addressed (a possible candidate in this regard are left-to-right scanning biases induced by reading; see below). Conversely, to the extent that the common mechanisms are primarily neural
Visual Field Asymmetries 285 in nature, the relation between the dorsal and ventral visual cortical pathways and the left and fight hemispheres needs to be addressed. Second, the fact that left-right differences are often confined to certain positions along the vertical meridian (e.g., Levy & Kueck, 1986; Niebauer & Christman, 1997; Worrall & Coles, 1976) casts doubt on explanations of left-right differences in terms of hemispheric asymmetry alone. Equivalently, these same results can be recast as reflecting the fact that upper-lower asymmetries are often confined to certain positions along the horizontal meridian, suggesting that such differences can not be explained in terms of ecological differences alone. Thus, the relation between upper/lower and left/right visual field asymmetries is likely to involve complex interactions among multiple mechanisms. We will now discuss three potential mechanisms underlying these visual field asymmetries. The first involves attentional biases; in particular, a key component of attention involves the detection of and orienting to objects in the visual field (as in visual search). The upperlower differences in such attentional biases have a clear ecological origin, as the types of objects an organism will typically engage in visual search for are located relatively far away from the organism and hence in the upper visual field. While the natural environment does not produce any comparable left/right anisotropy, directional scanning habits induced by reading may involve comparable phenomena. That is, given that parafoveal presaccadic previews of material about-to-be-read plays an important role in normal reading (e.g., Rayner, McConkie, & Zola, 1980), the attentional system may become "tuned" to more efficiently extract information to the fight of fixation (at least in readers of left-to-right languages). This line of speculation is consistent with evidence that various types of experience can alter the nature of leftright asymmetries in visual search (Buckles, Yund, & Efron, 1991; Ostrosky-Solis, Efron, & Yund, 1991; Yund & Efron, 1996). This suggests that subjects from cultures with fight-to-left reading habits may display reversed left-right, but comparable upper-lower, asymmetries in attentional biases. Unfortunately, there is little relevant data to this question. Hints of such effects, however, are present in the findings of Mishkin and Forgays (1952), who reported right versus left VFAs for reading English versus Hebrew words, respectively, while both sets of words yielded upper VFAs; Orbach (1953) reported similar findings. Similarly, there is evidence that Western subjects prefer stimuli that invoke left-to-right attentional scanning (e.g., Christman & Pinger,
286
Christman and Niebauer
1997), whereas Japanese subjects prefer stimuli evoking right-to-left scanning (e.g., Richie, 1988). Although environmentally and experientially driven attentional biases can account for some of the upper/lower and left/fight visual field differences, it is likely that the upper/lower differences are independent of the left/right differences; that is, the upper/lower differences arise ultimately from the constraints imposed by gravity and an upright posture, whereas the left/right differences are likely to reflect more arbitrary directional biases induced by reading and other human activities. Also, although an attentional bias account handles the visual search and visual attention results reviewed above rather nicely, it cannot readily account for the many basic (and presumably pre-attentional) differences in terms of factors such as spatial and temporal resolution. The second, and by no means mutually exclusive, potential mechanism involves the role of the magnocellular versus parvocellular visual pathways. Previc (1990) reviewed a large body of neurophysiological literature and made the case that the upper versus lower visual fields are disproportionately represented in the parvo- versus magno-cellular pathways respectively. The magnocellular pathway is specialized for lower spatial frequencies, higher temporal frequencies, and transient responses, while the parvocellular pathway is specialized for higher spatial frequencies, lower temporal frequencies, and sustained responding. This dichotomy nicely captures the upper/lower differences in simple RT, spatial and temporal resolution/acuity, local/global perception, and motion perception. Unfortunately, there has been virtually no research examining possible left/right differences in processing visual information carried by the magno versus parvo pathways. We are aware of only three relevant papers. First, Cowin and Hellige (1995) examined categorical versus coordinate processing using green-on-red versus red-on-green stimuli; the red background of the green-on-red stimuli was hypothesized to attenuate the contribution of magnocellular visual pathway. Their results indicated that the red background did indeed produce greater impairment of coordinate, relative to categorical, processing. However, this Task X Background interaction did not involve visual field, leaving open the question of possible left/fight differences in processing output from the magnocellular versus parvocellular pathways. Second, Christman (1997b) presented local-global stimuli as either high contrast (c=0.95) black-on-white stimuli or as low contrast
Visual Field Asymmetries 287 (c=0.03) green-on-red stimuli. Although the red-green stimuli were not physically isoluminant (nor were they equated for subjective isoluminance), their relatively low contrast values presumably attenuated the sensitivity of the magnocellular pathway, relative to the black-white stimuli. The results were an intriguing complement to those reported by Cowin and Hellige (1995). First, background color had no effect on local-global processing as such: robust global precedence and globalon-local interference were obtained in both conditions. This result is in contrast to the presence of Task X Background effects in the Cowin and Hellige (1995)study. Second, background color did have differential effects on the left versus fight visual fields: the predicted left versus right VFAs for global versus local processing, respectively, were obtained for the black-on-white, but not green-on-red, stimuli (again, in contrast to the lack of visual field effects of background color in the Cowin & Hellige study). Although quite tentative, these results suggest that leftright visual field asymmetries may be dependent on hemispheric differences in processing the output of magnocellular pathways. Lastly, Stein (1994) has suggested that normal development of the magnocellular pathways may also promote development of normal hemispheric lateralization for sensory processing; conversely, he also suggests that there may be a common functional basis for the disordered perceptual abilities of dyslexics (e.g., lowered flicker and motion sensitivity), which are linked to magnocellular dysfunction, and the abnormal patterns of hemispheric asymmetry that dyslexics also display. In conclusion, while the important role of the magno and parvo pathways in upper/lower visual field differences appears beyond dispute, the relation of such pathways to left/right differences is not clear (although this lack of clarity is more reflective of a lack of relative evidence, as opposed to a lack of positive evidence or the presence of negative evidence). More generally, however, a magno/parvo account would be consistent with many of the left-right asymmetries reviewed above. Future laterality work would do well to consider a possible left hemisphere-parvocellular, right hemisphere-magnocellular connection. Nonetheless, if magno/parvo differences do underlie left/fight visual field differences (at least in part), it is unlikely that such field differences arise in the same way as upper/lower differences do. This is because there is clear evidence that the upper and lower visual fields project to cortical regions that differ in the relative proportion of magno versus parvo cells, while there is no evidence whatsoever for basic hemispheric
288
Christman and Niebauer
differences in the distribution of magnocellular versus parvocellular pathways. However, potential left/fight differences in processing output from magno- and parvocellular pathways may arise in a similar manner as left/fight differences in processing spatial frequency information. For example, the fact that hemispheric differences in spatial frequency processing are found in identification, but not detection, tasks led Kitterle and Christman (1991) to suggest that such differences do not arise from hemispheric differences in the representation of low versus high spatial frequency channels (which is the mechanism underlying upper/lower differences in magno/parvo processing). Rather, they suggested that the left and fight hemispheres differ in how they weight and evaluate the output from those symmetrically distributed pathways. Thus, the left/right differences may not reflect lateral differences in the distribution of magnocellular and parvocellular pathways, but may instead reflect differences in the manner in which left versus right hemisphere areas "downstream" from striate and extrastriate cortex handle the raw output from the two different pathways. A final possible basis for the links between upper/lowor and left/fight asymmetries concerns a possible effect of head tilt. The idea is very simple: in any study presenting stimuli to the left and right visual fields along the horizontal meridian, the presence of appreciable head tilt among the subjects will lead to a confounding of the left/right and upper/lower visual fields. For example, if a subject tilts their head rightwards, this means that all fight visual field stimuli will also be in the upper visual field, while all left visual field stimuli will also be in the lower visual field; thus, any left/fight differences in such a situation may in fact reflect upper/lower differences. Indeed, a population level bias toward rightward head tilt could account for the upper/right and lower/left visual field links; the evidence is mixed on this question, however, as some studies report greater frequency of rightward head tilt (Greenberg, 1960) and others report greater frequency of leftward head tilt (Previc, 1994). Furthermore, Previc (1994) reports systematic relationships between head tilt and eye dominance, raising the possibility of hemiretinal differences interacting with both upper/lower and left/right differences. In any case, variations in head tilt (whether systematic or random) could certainly result in complex interactions between upper/lower and left/fight visual field differences. One last point needs to be addressed concerning the possible interactions or interdependencies between upper/lower and left/fight visual
Visual Field Asymmetries 289 field differences. The discussion above suggests that, while these visual field differences do appear to have general mechanisms in common, the specific ways in which these mechanisms operate may be different (e.g., upper/lower attentional biases reflecting gravitational and postural constraints and left/right attentional differences reflecting directional scanning biases; upper/lower differences in distribution of magnocellular versus parvocellular pathways and left/right differences in the weighting and evaluation of output from symmetrically distributed pathways). Thus, this account would not directly predict interactions between upper/lower and left/fight visual fields. However, at least some of the studies reviewed above do indicate such effects. In particular, we would draw attention to the results of Niebauer and Christman (1997), who examined categorical versus coordinate processing in the upper fight, upper left, lower right, and lower left quadrants of the visual field and found that visual quadrant variations within each task were confined to the lower left quadrant, which exhibited superior performance on the coordinate task and inferior performance on the categorical task. An admittedly speculative account of these results involves the development of reaching behavior. In particular, early forms of reaching behaviors in neonates display a left-hand bias (e.g., DiFranco, Muir, & Dodwell, 1978; McDonnell, 1979). This finding, coupled with Previc's (1990) arguments about the special role of the lower visual field in reaching behavior, suggests that the visual feedback associated with infants' initial attempts at learning and refining reaching behavior (which requires precise knowledge about the exact location of both the hand and the object being reached for) will be primarily confined to the lower left quadrant of the visual field. In turn, through some sort of behavioral tuning and/or snowball mechanism, this initial lower left visual field advantage may persist into adults and account for the types of findings reported by Niebauer and Christman (1997). In summary, this chapter has had two primary purposes. First and foremost, it constituted an attempt to sensitize laterality researchers to the fact that many of the left/fight visual field differences that they study have comparable analogs (and possible homologs) in terms of upper/lower field differences, which in turn casts doubt on exclusively hemispheric interpretations of left/right visual field differences. Second, we have offered some tentative explanations for the systematic links found between upper and right visual field processing and between lower and left visual field processing in terms of attentional biases
290
Christman and Niebauer
induced by both the environment and reading, and in terms of the role of the magnocellular versus parvocellular visual pathways. Although these accounts are quite speculative at this time, we feel that, regardless of the ultimate explanation, the links between upper/lower and left/fight visual field processing are real and unequivocal.
References
Anderson, S.J., Mullen, K.T., & Hess, R.F. (1989). Asymmetry in the retina for motion and colour detection. Investigative Ophthalmology & Visual Science, 30, 53. Babcock, L.E., & Robison, R.A. (1989). Preferences of the Paleozoic predators. Nature, 33 7, 695-696. Beer, J.M.A., Gallaway, R.A., & Previc, F.H. (1996). Do individual's visual recognition thresholds predict performance on concurrent ~titutde c o ~ r o l f l i g h t tasks? International Journal of Aviation sycnotogy, o, A/J-A~I. Berardi, N., & Fiorentini, A. (1991). Visual field asymmetries in pattern discrimination: A sign of asymmetry in cortical visuaI field representation? Vision Research, 31, 1831-1836. Berlucchi, G., Tassinari, G., Marzi, C., & di Stefano, M. (1989). Spatial distribution, of the inhibitory. . effect of peripheral non-informative. cues on s~mple reaction time to non-fixated v~sual targets. Neuropsychologia, 2 7, 201-221. Blake, R., & Mills, J. (1979). Pattern and flicker detection examined in terms of the nasal-temporal division of the retina. Perception, 8, 549555. Breitmeyer, B., Julesz, B., & Kropfl, W. (1975). Dynamic random-dot stereograms reveal up-down anisotropy and left-right isotropy between cortical hemifields. Science, 187, 269-270. Bryden, M.P., & Underwood, G. (1990). Twisting the world by 90 ~ Behavioral and Brain Sciences, 13, 547-548. Buckles, K., Yund, W., & Efron, R. (1991). Visual detectability gradients: Effect of high-speed visual experience. Brain and Cognition, 17, 52-63. Calvert, J. (1988). Adaptation to movement in fight and left visual fields. Cortex, 24, 589-593. Cartoon, A., & Bechtoldt, H. (1969). Dominance of the right cerebral hemisphere for stereopsis. Neuropsychologia, 7, 29-39. Chaiken, J., Corbin, H., & Volkmann, J. (1962). Mapping a field of short-time visual search. Science, 138, 1327-1328. Chedru, F., Leblanc, M., & Lhermitte, F. (1973). Visual searching in normal and brain-damaged subjects l(c~176 to the study of unilateral inattention). Cortex, 9, 94-1 . Christman, S. (1988). Task Factors and Hemispheri c Asymmetries in Visible Persistence. Presented at the 16th Annual Meeting of the International Neuropsychological Society, New Orleans, January.
Visual Field Asymmetries 291 Christman, S. (1993). Local-global processing in the upper versus lower visual fields. Bulletin of the Psychonomic Society, 3 l, 275-278. Christman, S. (1997a). Hemispheric Asymmetry in Categorical versus Coordinate ProcesSing of Dynamic Visual Input. Presented at the 25th Annual Meeting of the International Neuropsychological Society, Orlando, FL, February. Christman, S. (1997b). Hemispheric Asymmetry in Local-Global Processing: Role of the Magnocellular System. Presented at the 38th Annual Meetingof the Psychonomic Society, Philadelphia. Christman, S., & Naegele, P. (1995). Visual Search as a Function of Retinal Locus of Target. Presented at the 36th Annual Meeting of the Psychonomic Society, Los Angeles, CA, November. Christman, S.D., & Pinger, K. (1997). Lateral biases in aesthetic preferences: Pictorial dimensions and neural mechanisms. Laterality, 2, 155-175. Colombo, A. De Renzi, E., & Faglioni, P. (1976). The occurrence of visual neglect in patients with unilateral cerebral disease. Cortex, 12, 221-231. Courtney, A.J., & Chan, H.S. (1986). Visual lobe dimensions and search performance for targets on a competing homogenous background.
Perception & Psychophysics, 40, 39-44.
Cowin, E., & Hellige, I . (1995). Spatial processing and hemispheric asymmetry: Contributions of the magnocellular system. Presented at the 36th Annual Meeting o f the Psychonomic Society, Los Angeles, CA, November. Cragg, B. (1969). The topography of the afferent projections in the clrcumstriate visual cortex of the monkey studied by the nauta method. Vision Research, 9, 733-747. Danta, G., Hilton, R., & O'Boyle, D. (1978). Hemisphere function and binocular depth perception. Brain, 101,569-589. Davidoff, J. (1977). Hemispheric differences in dot detection. Cortex, 13, 434-A,A.A.. DeValois, R., & DeValois, K. (1988). Spatial Vision. New York: Oxford University Press. DiFranco, D., Muir, D., & Dodwell, P. (1978). Reaching in very young infants. Perception, 7, 385-392. Dimond, S., Bures, J., Famngton, L., & Brouwers, E. (1975). The use of contact lenses for the lateralization of input in man. Acta Psychologica, 39, 341-349. Dumford, M., & Kimura, D. (1971). Right hemisphere specialization for depth perception reflected in visual field differences. Nature, 231, 394-395. Edgar, G., & Smith, A. (1990). Hemifield differences in perceived spatial frequency. Perception, 19, 759-766. Edwards, M., & Badcock, D. (1993). Asymmetries in the sensitivity to motion in depth: A centripetal bias. Perception, 22, 1013-1023. Effort, R. (1990). The Decline and Fall of Hemispheric Specialization. Hillsdale, NJ: Lawrence Erlbaum. Efron, R., Yund, E., & Nichols, D. (1987). Scanning the visual field without eye movements - A sex difference. Neuropsychologia, 25, 637-644.
292
Christman and Niebauer
Efron, R., Yund, E., & Nichols, D. (1990a). Serial processing of visual spatial patterns in a search paradigm. Brain & Cognition, 12, 17-41. Efron, R., Yund, E., & Nichols, D. (1990b). Detectability as function of target location: Effects of spatial configuration. Brain & Cognition, 12, 102-116. Eglin, M. (1987). The effect of different attentional loads on feature integration in the hemispheres. Perception & Psychophysics, 42, 81-86. Fendfich, R., & Gazzaniga, M.S. (1989). Evidence of f0veal splitting in a commisurotomy patient. Neuropsychologia, 27, 273-282. Filbey, R., & Gazzaniga, M. (1969). Splitting the normal brain with reaction time. Psychonomic Science, 17, 335-336. Finke, R., & Kosslyn, S. (1980). Mental imagery acuity in the peripheral visual field. Journal of Experimental Psychology: Human Perception and Performance, 6, 126-139. Fiorentini, A., & Berardi, N. (1984). Right hemisphere superiority in the discrimination of spatial phase. Perception, 13, 695-708. Foley, J., & McChesney, J. (1976). The selective utilization of information in the optic array. Psychological Research, 38, 251-265. Gawryszewski, L., Riggio, L., Rizzolatti, G., & Umilt~, C. (1987). Movements of attention in the three spatial dimensions and the meaning of "neutral" cues. Neuropsychologia, 25, 19-29. Grabowska, A. (1983). Lateral differences in the detection of stereoscopic depth. Neuropsychologia, 21, 249-257. Greenberg, G. (1960). Eye-dominance and head-tilt. American Journal of Psychology, 73, 149-151. Guez, J-E., Gargasson, J-F., Rigaudiere, F., & O Regan, J.K. (1993). Is there a systematic location for the pseudo-fovea in patients with central scotoma? Vision Research, 3 3, 1271- 1279. Gunter, T., Wijers, A., Jackson, J., & Mulder, G. (1994). Visual spatial attention to stimuli presented on the vertical and horizontal meridian: An ERP study. Psychophysiology, 31, 140-153. Hamsher, K. (1978). Stereopsls and unilateral brain disease. Investigative Opthalmology and Visual Science, 17, 336-343. Hardyck, C. (1986). Cerebral asymmetries and experimental parameters: Real differences and imaginary variations? Brain and Cognition, 5, 223-239. Harvey, Jr., L.O. (1978). Single representation of the visual midline in humans. Neuropsychologia, 16, 601-610. Hellige, J., & Mictiimata, C. (1989). Categorization versus distance: Hemispheric differences for processing spatial information. Memory and Cognition, 17, 770-776. Heron, W. (1957). Perception as a function of retinal locus and attention. American Journal of Psychology, 70, 38-48. Heywood, S., & Churcher, J. (1980). Structure of the visual array and saccadic latency: Impl!cations for oculomotor control. Quarterly Journal of Experimental Psychology, 3 2, 335-341. Honda, H., & Findlay, J. (1992). The increase in saccade latencies to the lower visual field. Perception & Psychophysics, 52, 167-174. Hutton, J., & Palet, J. (1986). Lateral saccadic latencies and handedness. Neuropsychologia, 24, 449-451.
Visual Field Asymmetries 293 Hylkema, B. (1942). Examination of the visual field by determining the fusion frequency. Acta Opthalmologica, 20, 181-193. Julesz, B., Breitmeyer, B., & Kropfl, W. (1976). Binocular-disparitydependent upper-lower hemifieId anisotropy and left-right isotropy as revealed by dynamic random-dot stereograms. Perception, 5, 129. Kinsbourne, M. (1972). Eye and head turning indicates cerebral lateralization. Science, 176, 539-541. Kitterle, F., & Christman, S. (1991). Hemispheric Symmetries and mmetries in the Processing of Sine-wave Gratings. In F. Kitterle
~F~.), Advances in Cerebral Eaterality Research: Data and Theory.
Hillsdale, N.J." Lawrence Erlbaum Associates. Kitterle, F., Christman, S., & Conesa, J. (1993). Hemispheric differences in the interference among components of compound gratings. Perception & Psychophysics, 54, 785-793. Kitterle, F., Christman, S., & Hellige, J. (1990). Hemispheric differences are found in the identification, but not detection, of low vs. high spatial frequencies. Perception & Psychophysics, 48, 297-306. Kitterle, F., & Kaye, R. (1985). Hemispheric symmetry in contrast and orientation sensitivity. Perception & Psychophysics, 37, 391-396. Kosslyn, S.M. (1987). Seeing and imaging in the cerebral hemispheres: A computational approach. Psychological Review, 94, 148-175. Kosslyn, S.M., Koenig, O., Barrett, A., Cave, C.B., Tang, J., & Gabrieli, J.D.E. (1989). Evidence for two types of spatial representations: Hemispheric specialization for categorical and coordinate relations.
Journal of Experimental Psychology: Human Perception and Performance, 15, 723-735.
Kostelyanets, N., Kamenkovich, V., & Sharaev, G. (1992). Two mechanisms of recognition of the direction of motion in humans. Sensory Systems, 6, 59-63. Ladavas, E., Carletti, M., & Gori, G. (1994). Automatic and involuntary_ orienting of attention in patients with visual neglect: Horizontal and vertical dimensions. Neuropsychologia, 3 2, 1195-1208. Levy, J., Heller, W., Banich, M., & Burton , L. (1983). Asymmetry of perception in free viewing of chimeric stimuli. Brain and Cognition, 2, 404-414. Levy, J., & Kueck, L. (1986). A right hemispatial advantage on a verbal free-vision task. Brain and Cognition, 27, 24-37. Lines, C.R., & Milner, A.D. (1983). Nasotemporal overlap in the human retina investigated by means of simple reaction time to lateralized light flashes. Experimental Brain Research, 50, 166-172. Lundh, B., Lennerstrand, G., & Derefeldt, G. (1983). Central and peripheral normal contrast sensitivity for static and dynamic sinusoidal gratings. Acta Opthalmologica, 61, 171-182. Marr, D. (1982). Vision. San Francisco: W.H. Freeman. McDonnell, P. (1979). Patterns of eye-hand coordination in the first year of life. Canadian Journal of Psychology, 33, 253-267. McKeever, W., Gill, K., & VanDeventer, A. (1975). Letter vs. dot stimuli as tools for "splitting the normal brain with reaction time". Quarterly Journal of Experimental Psychology, 2 7, 363-373.
294
Christman and Niebauer
Mecacci, L. (1993). On spatial frequencies and cerebral hemispheres: Some remarks from the electropliysiological and neurophysiological points of view. Brain and Cognition, 22, 199-212. Millodot, M., & Lamont, A. (1974). Peripheral visual acuity in the vertical plane. Vision Research, 14, 1497-1498. Miossec, Y., Kolinsky, R., Morais, J. (1993). Illusory conjunctions and the cerebral hemispheres. Perception & Psychophysics, 54, 604-616. Mishkin, M., & Forgays, D. (1952). Word recognition as a function of retinal locus. Journal of Experimental Psychology, 43, 43-48. Morikawa, K., & McBeath, M. (1992). Lateral motion bias associated with reading direction. Vision Research, 32, 1137-1142. Murasugi, C., & Howard, I. (1989). Human horizontal optokinetic nystagmus elicited by the upper versus lower visual fields. Visual Neuroscience, 2, 73-79. Murray, I, MacCana, F., & Kulikowski, J. (1983). Contributions of two movement detecting systems to central and peripheral vision. Vision Research, 2 3, 151- I59. Niebauer, C., & Christman, S. (1997). Upper and lower visual field differences in categorical and coordinate judgments. Psychonomic Bulletin and Review, in press. Orbach, J. (1953). Retinal locus as a factor in recognition of visually perceived words. American Journal of Psychology, 65, 555-562. Osaka, N., & Oda, K. (1991). Effective visual field size necessary for vertical reading during Japanese text processing. Bulletin of the Psychonomic Society, 29, 345-347. Ostrosky-Solis, F., Efron, R., & Y und, W. (1991). Visual detectability gradients: Effects of literacy. Brain and Cognition, 17, 42-51. Payne, W. (1967). Visual reaction times on a circle about the fovea.
Science, 155, 481-482.
Pennal, B. (1977). Human cerebral asymmetry in color discrimination. Neuropsychologia, 15, 563-568. Pointer, LS., & Hess, R.F. (1989). The contrast sensitivity gradient across the human visual field: With emphasis on the low spatial frequency range. Vision Research, 29, 1133-1151. Polich, J., DeFrancesco, D., Garon, J., & Cohen, W. (1990). Hemispheric differences in visual search of simple line arrays. Psychological Research, 52, 54-61. Polich, J., Lentz, W., & Crossman, D. (1986). Hemispheric differences for feature detection. Neuropsychologia, 24, 591-595. Posner, M. (1992). Attention as a cognitive and neural system. Current Directions in Psychological Science, 1, 11-14. Posner, M., Walker, J., Friedrich,.F.,.& Rafal, R. (1984). Effects of arietal injury on covert orienting of attention. Journal of uroscience, 4, 1863-1874. Prazdny, K. (1987). On the coarse-to-fine strategy in stereomatching. Bulletin of the Psychonomic Society, 25, 92-94. Prete, F.H. (1993). Stimulus configuration and location in the visual field affect appetitive responses by the praying mantis, Sphodromantis hneola (Burr.). Visual Neuroscience, lO, 997-1003.
Visual Field Asymmetries 295 Previc, F.H. (1990). Functional specialization in the lower and upper visual fields in humans: Its ecological origins and neurophysiological implications. Behavioral and Brain Sciences, 13, 519-575. Previc, F.H. (1994). The relationship between eye dominance and head tilt in humans. Neuropsychologia, 3 2, 1297-1303. Previc, F.H. (1996). Attentional and oculomotor influences on visual field anisotropies in visual search performance. Visual Cognition, 3, 277-301. Previc, F.H., & Blume, J. (1993). Visual search asymmetries in threedimensional space. Vision Research, 33, 2697-2704. Previc, F.H., Breltmeyer, B.G., & Weinstein, L. (1995). Discriminability of random-dot stereograms in three-dimensional space. International Journal of Neuroscience, 80, 247-253. Rao, S., Rourke, D., & Whitman, D. (1981). Spatio-temporal discrimination of frequency in the right and left visual fields: A preliminary report. Perceptual and Motor Skills, 53, 311-316. Rapcsak, S., Cimino, C., & Heilman, K. (1988). Altitudinal neglect. Neurology, 3 8, 277-281. Raymond,-LE. (1994). Directional anisotropy of motion sensitivity across the visual field. Vision Research, 34, 1029-1037. Rayner, K., McConkie, G., & Zola, D. (1980). Integrating information across eye movements. Cognitive Psychology, 12, 206-226. Rayner, K., Well, A.D., & Pollatsek, A. (1980). Asymmetry of the effective visual field during reading. Perception & Psychophysics, 27, 537-544. Rebai, M., Bagot., J-D., & Viggiano, M-P. (1993). Hemispheric asymmetry in transient visual evoked potentials induced by the spatial factor of the stimulation. Brain and Cognition, 23, 263-278. Rebai, M., Mecacci, L., Bonnet, C., & Bagot, J.-D. (1989). Influence of spatial frequency and handedness on hemispheric asymmetry in visually steady-state evoked potentials. Neuropsychologia, 2 7, 315-324. Regan, D., & Beverly, K.I. (1983). Visual fields for frontal plane motion and for changing size. Vision Research, 23, 673-676. Regan, D., Erke[ens, C.J., & Collewijn, H. (1986). Visual field defects for vergence eye movements and for stereomotion perception. Investigative Ophthalmology~ & Visual Science, 27, 806-819. Regan, D., & Vincent, A. (1995). Visual processing of looming and time to contact throughout the visual field. Vision Research, 35, 18451857. Richards, W., & Regan, D. (1973). A stereo field map with implications for disparity processing. Investigative Opthalmology, 12, 904-909. Richie, D. (1988). Viewing Japanese film: Some considerations. In W. Kissanayake (Ed.), Cinema and Cultural Identity. Lanham, MD: University Press of America. Rijsdik, J., Kroon, J., & van der Wildt, G. (1980). Contrast sensitivity as a function of position on the retina. Vision Research, 20, 235-241. Rizzolati, G., Riggio, L., Dascola, I., & Umilt~, C. (1987). Reorienting attention across the vertical and horizontal meridians: Evidence in favor of premotor theory of attention. Neuropsychologia, 25, 31-40. Robertson, L., & Lamb, M. (1988). The role of perceptual reference frames in visual field asymmetries. Neuropsychologia, 26, 145-152.
296
Christman and Niebauer
Rubens, A. (1985). Caloric stimulation and unilateral visual neglect. Neurology, 3 5, 1019-1024. Sanders, A.F., & BrUck, R. (1991). The effect of presentation time on the size of the visual lobe. Bulletin of the Psychonomic Society, 29, 206-208. Schwartz, S., & Kirsner, K. (1982). Laterality effects in visual informationprocessing: Hemispheric specialisation or the orienting of attention.9- Quarterly Journal of Experimental Psychology, 34 A, 61-77. Sergent, J. (1982). The cerebral balance of power: Confrontation or cooperation? Journal of Experimental Psychology: Human Perception and Performance, 8, 253-272. Sergent, J. (1983). The role of the input in visual hemispheric asymmetries. Psychological Bulletin, 93, 481-512. Shulman, G., Sullivan, M., Gish, K., & Sakoda, W. (1986). The role of spatial frequency channels in the perception of local and global structure. Perception, 15, 259-273. Singer, W. (1985). Critical flicker fusion and double flash discrimination in differentparts of the visual field. International Journal of Neuroscience, 2 5, 225-231. Skandries, W. (1987). The upper and lower visual field of man: Electrophysiological and functional differences. In H. Autrum et al. (Eds.), Progress in Sensory Physiology (vol. 8), Springer-Verlag. Smith, A.T., & Hammond, P. (1986). Hemifield diffe~rences in perceived velocity. Perception, 15, 111 117. Spinelli, D., & Mecacci, L. i1990). Handedness and hemispheric asymmetry of pattern reversal visual-evoked potentials. Brain and Cognition, 13, 193-210. Stein,-J. (1994). Developmental dyslexia, neural timing and hemispheric lateralisation. International J. of Psychophysiology, 18, 241-249. Treisman, A. & Gelade, G. (1980). A feature-integration theory of attention. Cognitive Psychology, 12, 97-136. Tsal, Y. (1989). Attending to horizontal, diagonal, and vertical positions in space. Bulletin of the Psychonomic Society, 27, 133-134. Tyler, C. (1987). Analysis of visual modulation sensitivity III. Meridional variations in peripheral flicker sensitivity. Journal of the
Optical Society of America, 4, 1612-1619
Umilt~, C., Salmaso, D., Bagnara, S., & Simion, F. (1979). Evidence of a right hemisphere superiority and a serial search strategy in a dot detection task. Cortex, 15, 597-608. Van Kleeck, M. (1989). Hemispheric differences in global versus local processing of hierarchical visual stimuli by normal subjects: New data and a meta-analysis of previous studies. Neuropsychologia, 27, 1165-1178. Worrall, N., & Coles, P. (1976). Visual field differences in recognizing letters. Perception & Psychophysics, 20, 21-24. Yund, W., & Efron, R. (1996). Guided Search: The effects of learning.
Brain and Cognition, 31,369-386.
Yund, E., Efron, R., & Nichols, D. (1990). Detectability gradients as a function of target location. Brain & Cognition, 12, 1-16. Zeki, S. (1969). Representation of the central visual fields in prestriate cortex of monkey. Brain Research, 14, 271-291.
SECTION V: AUDITORY PROCESSING
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
299
Chapter 10
Hemispheric specialization of human auditory processing: Perception of speech and musical sounds Robert J. Zatorre Montreal Neurological Institute McGill University Montreal, Quebec, Canada Speech and music are perhaps the most interesting way that human cognition makes use of sound. It seems likely that the complex mental operations necessary for the processing of speech and music would demand a correspondingly complex set of neural computations. This paper will review studies from our laboratory aimed at exploring these issues, utilizing both the traditional behavioral-lesion approach as well as recent brain imaging techniques. Among the latter methods, we have used both functional brain imaging, with positron emission tomography (PET), as well as structural imaging, with magnetic resonance imaging (MRI). These techniques allow us to explore cerebral activation patterns associated with the performance of certain tasks in healthy volunteer subjects, and also allow us to begin to explore structure-function relationships in the brain. We have followed the strategy of adapting or developing behavioral tasks drawn from the fields of psychophysics and cognitive psychology in order to study the neural correlates of a wide range of psychological processes relevant to auditory cognition. The tasks to be described in this paper focus on aspects of phonetic percep-
300
Zatorre
tion, melodic processing, auditory working memory, and auditory imagery. Each of these areas will be discussed in turn, preceded by a brief introduction to functional imaging methods. To conclude, we present some novel findings from structural imaging measures which may have direct relevance to understanding aspects of functional results. The development of PET as a technique to measure functional activation represents a major advance for cognitive neuroscience as it permits for the first time a relatively direct way to investigate changes in cerebral activity patterns as a function of specific task performance in normal subjects. A detailed explanation of the physics of the technique and the image processing that accompanies it is beyond the scope of this paper (cf. Raichle et al., 1983, for a description), but a brief description follows. The basic idea behind the application of PET relevant to our studies is that a short-lived radioactive tracer (oxygen-15) is used to measure cerebral blood flow (CBF) during a 60-second period. A scan is then reconstructed which represents a three-dimensional map of the CBF distribution in the entire brain during that time, with a spatial resolution on the order of 14 to 18 mm. Typically, several such scans are obtained in each individual subject. Data from a group of subjects may then be averaged, after appropriate stereotaxic normalization is applied to correct for differences in brain size, shape, and orientation. Averaged CBF data from a given condition may then be compared to another condition by superimposition of the relevant scans, and application of a pixel-by-pixel subtraction algorithm which detects significantly different areas of CBF in one condition as compared to another (Worsley et al., 1992). Although numerous other approaches to image analysis and paradigm design are also possible, including regression-based techniques (cf. Paus et al., 1996), the studies to be described in the present paper all used the paired-image subtraction method, in which two different conditions are compared directly to one another. The assumption is that the difference image reflects areas of cerebral activity specifically related to the task in question, relative to the baseline condition which typically represents an attempt to control for certain aspects of the task. A final aspect of the functional imaging work presented here is that structural images (MRI) are also obtained for each subject, and are co-registered to the PET images (Evans et al., 1992). These structural images permit much improved anatomical localization, and form the basis for the structural measures of cortical volume to be described at the end of this paper. ,
Speech and Music 301 PHONETIC M E C H A N I S M S IN SPEECH P E R C E P T I O N
The studies of speech processing carried out in our laboratory have focused on understanding the neural mechanisms relevant to the extraction of phonetic units from a speech signal. The psychological literature on speech perception has long maintained that such a process would require a specialized speech module, distinct from other perceptual mechanisms; one specific model of speech processing further emphasizes the possible role of motor-articulatory codes in the perception of speech (e.g., Liberman & Mattingly, 1985). Until recently, it has proven difficult to obtain direct evidence bearing on this question, particularly in terms of specifying the neural bases of these mechanisms. A large literature exists on aphasia, of course, with some relevant findings suggesting the involvement of motor structures in phonetic perceptual disorders (e.g., Blumstein et al., 1977; Gainotti et al., 1982). However, precise anatomical-functional correlations are often difficult to establish, and there is considerable variability in the pattern of deficits observed. In our first PET study aimed at understanding speech perception (Zatorre et al., 1992), we tested ten normal volunteers using two types of stimuli: pairs of noise bursts, which had been matched acoustically to the syllables to be used subsequently, and pairs of CVC real speech syllables. The vowels in any given syllable pair were always different, but the final consonant differed in half of the pairs; in addition, the second syllable had a higher fundamental frequency in half the pairs, and a lower frequency in the other half. The study included five conditions arranged in a subtractive hierarchy (Petersen et al., 1988). The first was a silent baseline; in the noise condition subjects pressed a key to alternate pairs of noise bursts. In the passive speech condition subjects listened to the syllables without making an explicit judgment, and pressed a key to alternate stimulus pairs. In the phonetic condition subjects listened to the same speech stimuli but responded only when the stimuli ended with the same consonant sound (e.g., a positive response would be given to the syllables [bag] and [pig]). In the pitch condition subjects once again listened to the same syllables, but responded only when the second item had a higher pitch than the first. Our aim was to use the subtractive technique to distinguish between speech-specific neural processes and more general auditory processing mechanisms. In particular, we wished to clarify the role of primary versus secondary auditory regions in speech perception. We hypo-
302
Zatorre
thesized that simple auditory stimulation (noise bursts) should lead to activation of primary cortex, whereas more complex signals should lead to activity in a broader expanse of cortical areas. Second, we wanted to test a hypothesis that phonological processing depends on left temporoparietal cortex (as suggested by the earlier PET study of Petersen et al., 1988), by using a phonetic discrimination task. Finally, we attempted to dissociate linguistic from nonlinguistic processing by requiring judgments of pitch changes in the speech syllable, which we hypothesized to involve right-hemisphere mechanisms, in accord with other data from our laboratory (Zatorre, 1988; Zatorre & Samson, 1991). When the results from the silent baseline condition were subtracted from the noise condition, activation was observed bilaterally approximately within the transverse gyri of Heschl, corresponding to primary auditory cortex (Penhune et al., 1996), as predicted. Subtraction of the noise condition from the passive speech condition yielded several foci along the superior temporal gyrus bilaterally. This region contains several cytoarchitectonically distinct cortical fields responsive to auditory stimulation and receiving input from corticocortical connections and the medial geniculate nucleus (Brugge & Reale, 1985; FitzPatrick & Imig, 1982). It is therefore likely to be involved in higher-order auditory processing of complex signals. One left-lateralized focus in the posterior superior temporal gyrus was also identified in this comparison. This finding is important as it implicates this portion of the left temporal lobe, roughly falling within the classical posterior speech region, as being automatically engaged in the processing of speech signals. Perhaps the most surprising results were obtained in the two active conditions, in which subjects made specific judgments of either phonetic identity or pitch. In the subtraction of the phonetic condition minus passive speech, activity was largely confined to the left hemisphere: the largest increase was observed in part of Broca's area near the junction with the premotor cortex, and in a superior parietal area. The prediction that pitch processing would involve right-hemispheric mechanisms was confirmed in the pitch condition minus passive speech subtraction, with two foci observed in the right prefrontal cortex. In these latter two subtractions both stimuli and responses were identical; only the nature of the required cognitive processing changed as a function of the instructions. The dissociable patterns of activity observed must therefore reflect the fundamentally different nature of the neural mechanisms involved in analysis of phonetic and pitch information, respectively.
Speech and Music 303 These results were subjected to further scrutiny in a second study (Zatorre et al., 1996a) aimed at replicating and extending the first set of findings. In particular, the finding of increased CBF in a frontal region close to the conventionally defined Broca's area was of special significance, since its involvement in a purely perceptual task has implications for models of speech processing. However, data from an)" single PET activation comparison must be viewed cautiously, as many different, and perhaps uncontrolled, task dimensions may be responsible for the observed effect (e.g., comparing an active discrimination to a passive listening condition entails many differences in cognitive demands, including attention, working memory, response organization, etc.); hence the importance of replicating results under different conditions. In order to assess these issues we carried out a new task, requiring monitoring of a given target phoneme within a stream of speech syllables (e.g., pressing a key whenever a [b] was perceived). This task is slightly different in its cognitive demands from the original phonetic task (it does not require comparison of pairs of stimuli, for example), but should engage the same phonetic processing system as was recruited by the first task. A comparison of this task to passive listening of syllables yielded very similar activation within the region close to Brocafs area previously observed. Furthermore, we carried out a reanalysis of the previous study to compare the phonetic and pitch tasks to one another. Our reasoning was that both tasks require active comparison and decision processes, but differ in terms of the crucial phonetic processing component which we wished to isolate. Once again, this comparison yielded a similar left frontal cortex response. Figure 1 shows a diagram of comparisons from our studies, together with data from similar studies conducted by D6monet (1992; 1994). Each symbol represents a stereo-taxic point at which maximum CBF activity was reported. The points cluster closely together, despite coming from different studies, different subjects, and different laboratories. It is interesting that these foci consistently cluster in the most superior and posterior aspect of cytoarchitectonic area 44, rather than to the inferior aspect of the third frontal convolution or to opercular areas traditionally associated with Broca's area. Activation in the latter location has been reported in tasks requiting overt (Petrides et al., 1993) or covert (Wise et al., 1991) vocal production. The highly consistent anatomical placement of the region shown in Figure 2 may indicate the existence of a functional subregion within Broca's area related to phonetic operations.
~o o 4~
nat
~f
9 Phonetic- Speech Zatorre et al., 1992
o
Zatorre et al., 1996
f
a Phonetic- Real Words ta Phonetic Monitoring- Real Words
Phonetic- Pitch
O Phonetic- Noise
Demonet et al. 1992 1994
k P h o n e m e s - Tones A S e q A m b - Tones
F i g u r e 1. Summary diagram demonstrating the location of activation foci near Broca's area identified in PET activation studies by Zatorre et al. (1992; 1996a) placed on coronal (front view) and saggital (lateral view) projections of a left hemisphere. For comparison purposes, the location of foci described by Dtmonet et al. (1992; 1994) in phonetic discrimination tasks is also shown. The symbols represent the center (most significant voxel) of extended regions of CBF increase identified in each comparison. The brain outlines and major sulci were generated from an averaged MRI data set, transformed into stereotaxic space (Talairach & Tournoux, 1988).
EXPERIMENT I (2-mm slices, N=20)
E X P E R I M E N T II (1-mm slices, N=20)
4000 - -
4,
5000-4,
I
>
]
4000-... 3000 -
/ 4.
4.
A
E
E
+
g _~ ~oo-
_=
,
>O t=.
<E
2000-
O.
r 2000 - ..J
.,J
lOOO
"§
'*
I
.y.:
--]
! "'~J 2000
3000
Right PAC-r Volume (ram3)
4000
|_r,L I
'.
1000 - -
I 1000
......
I 1000
'
I 2000
'
I 3000
'
I 40OO
I
1 50OO
Right PAC-r volume (mm3)
Figure 2. (Left) Scatterplots for volume of Heschl's gyms, or primary auditory cortex region (PAC-r) in left and fight hemispheres for two samples of subjects. The solid line represents values of equal left and fight volume; points which fall above the line represent subjects whose left volume was greater than fight. (Right) Average volumes of gray and white matter for left and right Heschlfs gyms in second sample of twenty subjects scanned with high-resolution MRI protocol; standard errors displayed as vertical bars. Adapted from Penhune et al. (1996).
c~ :3" ta r
(t) b,,io
O
306
Zatorre
The conclusion that part of Broca's area participates in phonetic processing does not imply that other regions, particularly the left posterior temporal region, do not play a role in the phonetic analysis of speech sounds. Experimental conditions in which subjects listen to speech syllables have consistently yielded asymmetric left posterior temporal CBF increases, as well as bilateral activation in the anterior portion of both superior temporal gyri (Wise et al., 1991; Petersen et al., 1988; Price et al., 1992; Zatorre et al., 1992). Functional MRI data have also corroborated this finding (Binder et al., 1994). Thus, when passive listening is used as the baseline state, any neural activation in these regions would be subtracted away. It seems clear that "passive" listening would include an important phonetic processing component that would be engaged automatically, but that is not observable in subtractions using passive listening as a baseline. It is reasonable to assume that neural processes in the superior temporal gyri are initially responsible for perceptual analysis of the complex incoming speech stream, since neurophysiological studies of auditory cortices reveal the presence of neuronal populations sensitive to acoustic features that are present in speech sounds, such as frequency modulation (e.g. Whitfield & Evans, 1965), or onset times (e.g. Steinschneider et al., 1995). It is therefore likely that the CBF activation in the left and right anterior superior temporal area observed during "passive" speech reflects the operation of such neural systems. The posterior region of the left superior temporal plane likely plays a special role in speech processing, since this region is not activated by simple tones or noise stimuli (Zatorre et al., 1992, 1994), or by auditory tonal discrimination tasks (DEmonet et al., 1992, 1994; Zatorre et al., 1994), but is consistently activated by speech stimuli. The processing carried out within this left posterior temporal area is not fully understood, but probably involves the analysis of speech sounds leading to comprehension, and may operate at the syllabic or whole-word level. This aspect of speech processing appears to be distinct, however, from processes that engage the network that includes the portion of Broca's area identified above. In the phonetic tasks in question, a relatively abstract pattern-extraction process must take place, since individual phonetic units belonging to the same category may have very different acoustic manifestations. It is therefore apparently insufficient to rely on a whole-syllable representation to perform this type of task; rather, recourse must be made to a specialized mechanism that is able to
Speech and Music 307 compute the similarity between phonetic segments that are differently encoded acoustically by virtue of being embedded in syllables with different vowels (Liberman & Mattingly, 1985). We would argue that this type of judgment calls into play the specialized articulatory recoding system whose neural manifestation is activity in a portion of Broca's area. PROCESSING OF MELODIC PATTERNS
The apparent ease with which most people, including very small children, recognize and reproduce melodies belies the complex nature of music processing. Indeed, perceiving and encoding a melody, or pattern of pitches, entails multiple cognitive operations that include perceptual analysis, abstract pattern-matching, and working memory, among others (Deutsch, 1982; Dowling & Harwood, 1986). Many questions are raised by the ubiquitous human capacity to listen to and perform music. Over the past several years our laboratory has attempted to explore the neural correlates of pitch and melody processes, not only to localize the systems responsible, but also to address whether music may rely on specialized neural operations distinct from those used for speech, and for other auditory processes. A considerable number of studies have begun to point to the existence of a specialization within the right temporal cortex for processing of certain aspects of pitch. For example, although simple frequency discrimination is affected only slightly or not at all by unilateral cortical lesions in humans (Milner, 1962, Zatorre, 1988) or bilateral lesions in animals (Evarts, 1952; Heffner & Masterton, 1978, Jerison & Neff, 1953), if a pitch judgment requires spectral analysis, then right-hemisphere auditory cortical regions seem to play a special role. Thus, perception of the missing fundamental is affected specifically by right temporal-lobe lesions which invade portions of Heschl's gyri, and not by more restricted anterior temporal-lobe damage or by left temporal excision (Zatorre, 1988). Similar lateralization findings have been reported in tasks requiring processing of complex harmonic structure (Divenyi & Robinson, 1989; Robin et al., 1990; Sidtis & Volpe, 1988). Furthermore, timbre discrimination tasks involving changes in harmonic structure have also yielded consistent evidence favoring right-asymmetric processing, both with temporal-lobe
308
Zatorre
lesioned patients (Milner, 1962; Samson & Zatorre, 1994), as well as with commissurotomized subjects (Tramo & Gazzaniga, 1989). Short-term retention is another aspect of pitch processing that apparently requires asymmetric mechanisms. Zatorre and Samson (1991) demonstrated that right temporal-lobe excision affected shortterm memory for pitch when interfering stimuli were presented between the target and comparison items (Deutsch, 1970). Bilateral ablations in the superior temporal gyrus of the monkey also result in deficits in tonal retention (Colombo et al., 1990; Stepien et al., 1960), and this region is implicated in auditory short-term memory by single-unit data as well (Gottlieb et al., 1989). Zatorre and Samson (1991) also observed that right frontal-lobe damage significantly impaired pitch retention, a finding mirrored in certain animal studies of bilateral frontal ablation (Gross & Weiskrantz, 1%2; Iversen & Mishkin, 1973). The latter result may reflect a disruption of functional connectivity between frontal and parietal cortices (Petrides & Pandya, 1988), which may be involved in maintenance of pitch in working memory (Perry et al., 1993; Marin & Perry, in press; see also Chavis & Pandya, 1976). Despite the apparent importance of neural systems within the right cerebral hemisphere for aspects of pitch processing, additional evidence clearly indicates that there are important contributions to musical processes from the left hemisphere as well. The most dramatic evidence for this conclusion comes from studies of patients who suffer from amusia, a specific disorder of recognition of all musical information, including very familiar melodies, which has been observed to occur only after bilateral lesions involving the two superior temporal gyri (Peretz, 1996). Even unilateral left superior temporal lesions can, in some instances, result in mild or moderate melodic processing deficits (Zatorre, 1985; Samson & Zatorre, 1988), particularly for recognition memory tasks (Samson & Zatorre, 1992). The above studies formed the backdrop for a functional imaging study (Zatorre et al., 1994) whose goal was to better understand the neural basis for perception of melodic patterns and for retention of pitch information in working memory. Our goal was to test two specific hypotheses: (1) that perceiving a novel tonal melody would entail neuronal processing in both left and right superior temporal regions, with a possibly greater contribution from the right; and (2) that right frontal-lobe mechanisms would be engaged when subjects make specific judgments that require retention of pitch over a filled interval.
Speech and Music 309 We tested twelve subjects without formal musical training using two classes of stimuli: noise bursts and melodies. The noise bursts were constructed so as to approximate the acoustic characteristics of the melodies in terms of number, duration, inter-stimulus presentation rate, intensity, and onset/offset shape. Sixteen different eight-note tonal melodies were also prepared, all identical in their rhythmic configuration, with the aim of allowing pitch judgments of either the first two notes, or the first and last notes. The last note had a higher pitch than the first in half of the sixteen melodies, and in the other half the last note was higher in pitch. Four separate conditions were run during each of the four scanning periods. During the first condition, termed the "noise" condition, subjects listened to the series of noise bursts described above, and after each "noise melody" depressed a key to control for motor activity. In the second condition, termed "passive melodies," the subjects were presented with each of the sixteen tonal melodies, and depressed a key after each one, as before. No overt judgments were required, but subjects were instructed to listen carefully to each melody. In the third condition, the "2-note" pitch comparison, subjects listened to the same melodies as before, but this time were instructed to determine whether the pitch of the second note was higher or lower than that of the first note. Finally, in the "first/last" pitch judgment, subjects were asked to compare the pitch of the first and last notes, ignoring the notes in between, and to respond as before, according to whether the pitch rose or fell. Subjects kept their eyes shut throughout the scanning period. The experiment was set up to permit specific comparisons, accomplished via subtraction of relevant conditions. The first comparison, passive melodies minus noise, permits examination of the cerebral regions specifically active during listening to novel tonal melodies, as opposed to the activation that might be present with any auditory stimulus with similar acoustic characteristics. The principal result indicated a significant CBF increase in the fight superior temporal gyrus, anterior to the primary auditory cortex. A much weaker CBF increase was also visible within the left superior temporal gyrus. In addition, and unexpectedly, a significant focus was also identified in the fusiform gyrus of the right hemisphere, within area 19. The finding of an activation within the fight superior temporal gyrus while listening to melodies fits in well with our prediction, and likely reflects the specialization of neuronal networks within the right secondary auditory cortices for perceptual analysis of tonal information consistent with the
310
Zatorre
human lesion evidence reviewed above (Milner, 1962, Zatorre, 1985; 1988). Although subjects were listening "passively," it is evident that they would be extracting perceptual information during this phase, and the CBF changes we observed most likely reflect these automatically engaged processes. The weak activity in the left superior temporal area may indicate the additional but perhaps less important or less consistent participation of left temporal cortices in melody processes, which is also suggested by the evidence from lesion studies. Note that, in this subtraction, no CBF increase was present in the primary auditory cortices beyond that elicited in the control condition. This result is explained by the control condition: by using acoustically matched noise bursts, nonspecific auditory processing can be dissociated from that uniquely elicited by listening to melodies. We previously demonstrated that similar noise bursts result in primary auditory cortical stimulation when contrasted to a silent condition (Zatorre et al., 1992). These findings, together with findings from prior PET studies using speech sounds or tones (DEmonet et al., 1992; Petersen et al., 1988; Wise et al., 1991) point to differential activation of primary vs. secondary auditory areas within the superior temporal gyrus, according to the nature of the processing elicited by a given stimulus. Although the noise stimuli proved successful in demonstrating the intended dissociation, caution must be still exercised in interpreting the results, for the noise bursts are clearly not physically identical to the melodic sounds. For example, the noise stimuli contain no periodicity, whereas the tones do; their spectral composition also is quite different. It remains to be established, therefore, which specific features of the melodies may lead to the observed pattern of activation. Our finding of activation in the right fusiform gyrus is puzzling. Area 19 is typically described as extrastriate visual cortex (Diamond et al., 1985); there is scant physiological evidence for its direct participation in auditory processing. The possibility that the effect is due to some extraneous visual stimulation is excluded, since scanning was carried out with the subjects' eyes closed. This phenomenon clearly invites further investigation; the possibility that it reflects visual imagery processes elicited indirectly by the melodic patterns in an intriguing one, but this conjecture must await direct evidence before it can be accepted. The second and third comparisons in this study both used the passive melody condition as the baseline, so that any activation seen represents neural responses beyond those already present during initial
Speech and Music 3 1 1 listening to the same stimulus materials. Subtraction of the passive condition from the 2-note condition resulted in significant activation within the right frontal lobe, as predicted. Two separate foci could be distinguished within distinct cytoarchitectonic regions, including Brodmann's areas 47/1 l, and 6. The first/last-passive melodies subtraction yielded a number of cortical and subcortical activation sites in both hemispheres. Among the more relevant results were CBF increases within the fight frontal lobe, consistent with the predictions, including a focus in area 47/11 identical to that observed in the 2-note condition. Of particular interest was an area of significant CBF increase within area 21 of the right temporal lobe, indicating that this condition resulted in greater activity within the right auditory association cortex than already present during passive listening to melodies. As well, we observed an increase in CBF within the fight inferior colliculus. The pattern of results from these conditions implicates frontal-lobe mechanisms in effecting pitch comparisons, as had been predicted, with a particularly important contribution from right-frontal regions. In the first/last minus passive melody comparison, we observed a greater number of separate foci of CBF change over a wider swath of cortical and subcortical territory than was evident in the lower memory-load condition of judging the first two notes; this finding perhaps reflects the complexity and increased cognitive demands of the task, which was also manifested in increased error rate and slower reaction times. Although we are not in a position to interpret all of these foci, one may speculate that the numerous frontal-lobe sites observed might be associated with successful performance of distinct aspects of the task. For example, maintenance of pitch information in working memory might depend on a mechanism separate from that involved with the more "executive" functions required to monitor the presentation of the tones and their temporal order, and to direct the appropriate pitch comparison (see Milner & Petrides, 1984). The right inferior colliculus, known to be an important auditory processing structure (Aitkin, 1986), was also activated in this subtraction, indicating that it too is a component of a specialized distributed network involved in pitch memory. Putting the results from the various tasks together with the physiological and lesion literature discussed earlier, a preliminary outline of a model to describe the neural substrates associated with pitch processing may be suggested. We may speculate that the primary auditory cortex is chiefly involved in early stages of processing (which
312
Zatorre
might include computation of such signal parameters as pitch, duration, intensity, and spatial location), whereas more complex feature extraction, involving temporally distributed patterns of stimulation, is performed via populations of neurons within the secondary cortices. Neuronal systems located in both temporal lobes likely participate in higher-order perceptual analysis of melodies, but those on the right seem to be particularly important, perhaps because they are specialized to extract the features that are most relevant for melodic stimuli (including, for example, invariant pitch-interval relationships, and spectral characteristics important for pitch and timbre perception). The existence of temporal-lobe neurons with complex response properties (e.g., McKenna et al., 1989) would be in keeping with this idea. In both pitch judgment conditions we observed significant CBF increases within the right frontal cortex. Only in the first/last comparison, however, did we observe an additional CBF increase in the right temporal lobe, beyond that seen in passive listening. We interpret this result, together with the fight frontal activation, as evidence that the high memory load imposed by the first/last task engaged a specialized auditory working memory system, and that this system is instantiated in the brain via interaction of inferior frontal and superior temporal cortices in the right cerebral hemisphere (Matin & Perry, in press). This conclusion would be in accord with our earlier study (Zatorre & Samson, 1991), in which deficits in pitch retention were observed after fight frontal and/or temporal-lobe lesions. AUDITORY IMAGERY
Many people, musically trained or not, report a strong subjective experience of being able to imagine music or musical attributes in the absence of real sound input. But subjective reports are of limited use to assess the characteristics of cognitive representations in a scientifically rigorous manner. Therefore, in recent years, psychologists have tried to find more objective means of evaluating the nature of imagery processes. Much of this research has concentrated on the visual domain, and has yielded the conclusion that visual imagery processes operate with similar characteristics to perceptual processes (see Farah, 1988, for a review). This view leads to the hypothesis that perception and imagery may share at least partially, the same neural substrate. Perhaps the nervous system has evolved in such a way that all sensory processing
Speech and Music 313 areas, which are normally responsive to environmental input, can also be activated endogenously, i.e., in the absence of external stimulation. If so, then at least a preliminary explanation of the neural basis for imaginal processing would be at hand. In our laboratory we have examined the aforementioned hypothesis within the context of musical imagery, using both a behavioral lesion approach (Zatorre & Halpern, 1993), and via PET functional imaging (Zatorre et al., 1996b). We adapted a paradigm originally developed by Halpern (1988), in which musically untrained subjects compared the pitch of two lyrics from a familiar, imagined song. (For instance, is the pitch corresponding to "sleigh" higher or lower than that of "snow" in the song "Jingle Bells"?) She varied the distance (number of beats) between the target lyrics chosen, as well as the distance from the beginning of the song of the first lyric of the pair. Response latencies increased systematically as a function of both factors, suggesting that subjects were "mentally scanning" the tune in order to compare the imagined pitches. Thus, she concluded that the temporal pace and ordering of the notes in the real song were preserved in analogous fashion in the image of the song. This result is similar to the conclusion that real-world spatial characteristics are preserved in visual images (Kosslyn, Ball, & Reiser, 1978). In the first of our neuropsychological imagery studies (Zatorre & Halpern, 1993), we examined whether auditory imagery and perception may share similar neural mechanisms by presenting a modification of the tune scanning task to patients having undergone right or left temporal-lobe excision for the relief of intractable epilepsy. A perceptual version of the task was devised in which the listener made pitch judgments while actually hearing the song. As well, subjects participated in an imagery condition in which judgments of pitch were made to imagined tunes indexed by the lyrics. The results of that study were very clear and striking. While all subjects did better on the perception task compared to imagery, patients with left-temporal excisions showed no deficits whatsoever relative to normal controls, whereas those with damage to the right temporal area were significantly worse than the other groups on both tasks, and by about the same amount on each task. We concluded that structures in the right temporal lobe were crucial for successful performance of both imagery and perception tasks, suggesting the same kind of neuroanatomical parallelism (and by
314
Zatorre
extension functional parallelism)shown by Farah (1988), Kosslyn et al. (1993), and others for visual imagery and perception. PET methodology allows us to study the neural processes of normal subjects with greater anatomical precision compared to many other physiological techniques, including lesion studies. We therefore designed an experiment to investigate the putative similarity between perceptual and imagery mechanisms. We presented three tasks to a group of 12 normal participants: a visual baseline condition and two active tasks, one termed "perception," the other "imagery." The latter two were similar to those used by Zatorre and Halpern (1993): Two words from a familiar tune were presented on a screen, and the task was to decide if the pitch corresponding to the second word was higher or lower than the pitch corresponding to the first word. In the perceptual task, participants actually heard the song being sung, while in the imagery task they carried out the task with no auditory input. In the baseline task subjects viewed the words and performed a visual length judgment. By subtracting the activation in the visual baseline from both the perception and imagery tasks, we should, in principle, eliminate cerebral activity related to nonspecific processes shared by the two tasks, such as reading words on a screen, making a forced-choice decision, pressing a response key, etc. Thus, any CBF changes still remaining must be due to the unique demands of listening to a tune or imagining it, and making a pitch comparison. The most striking findings from these subtractions were that for nearly every region demonstrating CBF change in one condition, there was a corresponding CBF peak in the other condition, often within a few millimeters. Most importantly, CBF increases were found bilaterally in the temporal lobes, in both perceptual and imagery conditions, and in the right frontal lobe. In addition, we observed areas of activation in both tasks in the left frontal and parietal lobes, as well as in supplementary motor area (SMA) and midbrain. The similarity in CBF distribution across the two conditions supports the idea that the two processes share a similar neural substrate. Not surprisingly, highly significant CBF increases were found within the superior temporal gyrus bilaterally when subjects were processing the auditory stimuli for the perceptual task, as compared to the baseline task, in which no auditory stimulation was provided. More interesting is the finding that regions within the superior temporal gyrus were also activated, albeit at a much weaker level, when subjects imagined hearing
Speech and Music 3 15 the stimulus, again as compared to the baseline condition. Note that this latter subtraction entails two entirely silent conditions, so that positive CBF changes in the superior temporal gyri (associative auditory cortices) cannot be due to any external stimulation, but are most likely attributable to endogenous processing. It is of interest to note that the temporal-lobe activation in the perceptual-visual baseline comparison incorporated primary auditory cortex and extended well into association cortical regions along most of the length of the superior temporal cortices. In contrast, this was not the case for the imagery-baseline comparison: CBF increases in that case occurred exclusively in association cortex (and were of lower relative magnitude). This distinction may be important, and supports the idea that primary sensory regions are responsible for extracting stimulus features from the environment, whereas secondary regions are involved in higher-order processes, which might include the internal representation of complex familiar stimuli. The activation of the SMA is also of particular interest, given its role in motor processes. This region has consistently shown CBF increases during various types of motor tasks, including speech production tasks (Petersen et al; 1988; 1989). Of greatest relevance to the present study, SMA is also involved when a motor task is only imagined, rather than overtly executed (Rao et al. 1993; Wise et al., 1991). The finding of SMA activation may therefore imply that the SMA is part of a substrate for both overt and covert vocalization, and therefore supports the idea that imagery for songs includes not only an auditory component ("heating the song in one's head"), probably related to temporal cortical activity, but also a subvocal component ("singing to oneself"), reflected in SMA activity. Taking the findings of the PET study together with the behavioral lesion study of imagery (Zatorre & Halpern, 1993), we conclude that there is good evidence that perception and imagery share partially overlapping neural mechanisms, and that these include superior temporal auditory regions, as well as motor areas. Clearly, a great deal remains to be learned about the neural correlates of such a complex function as mental imagery, not the least of which is to disentangle verbal from tonal aspects of imagery. The converging evidence provided by the complimentary approaches of lesion and functional imaging appear to be quite powerful, however, and we are thus
3 16
Zatorre
optimistic about being able to provide a more complete model of this intriguing aspect of cognition. M O R P H O M E T R Y OF A U D I T O R Y C O R T E X VIA
STRUCTURAL MRI We conclude the survey of our research into the correlates of human auditory processing by presenting some recent data pertaining to structural measures of auditory cortex (Penhune et al., 1996). Unlike the PET techniques described above, the aim of this research was to characterize the shape, volume, and position of the human primary auditory cortical region in vivo. The findings were somewhat surprising, and have possibly important implications for a better understanding of functional differences, however. The approach taken was to use three-dimensional MRI scans taken from groups of normal fight-handed volunteer subjects, and to label the region of Heschl's gyrus using interactive pixel-marking software that permits simultaneous viewing in all three planes of section, which greatly facilitates accurate anatomical delineation. Heschl's gyrus has long been known to contain the highly granular koniocortex, or primary auditory receiving area (Von Economo & Horn, 1930; Galaburda & Sanides, 1980), but the gross morphology of this region is highly variable across individuals and between hemispheres. Several studies have examined the cytoarchitecture of this region in the human brain, and most agree that the primary cortical region is confined roughly to the medial two-thirds of the most anterior Heschl's gyrus (Galaburda & Sanides, 1980; Rademacher et al., 1993), a conclusion also consistent with measures of evoked responses from depth electrodes within Heschl's gyrus (Li6geoisChauvel et al., 1991). These data indicate that gyral and sulcal landmarks may serve as consistent boundaries to define the region of interest. The results from our MRI-based morphometric measures therefore represent an estimate of the position and extent of primary auditory cortex in the human brain, but it is important to note that they also necessarily include non-primary cortical fields (particularly near the lateral edge of the gyrus), since there are no gross morphological features that would permit an exclusion of such areas. Two sets of data were obtained, each on a different sample of 20 volunteer subjects; the first set underwent MR scanning using a 2-mm
Speech and Music 317 slice thickness, the second underwent a higher resolution scanning protocol in which 1-mm thick slices were obtained. T 1-weighted images were acquired, then transformed to the standardized stereotaxic space of Talairach and Tournoux (1988) using an automatic three-dimensional cross-correlation algorithm which matches each individual MRI to an average of 305 manually registered images (Collins et al., 1994). Heschl's gyrus was then identified and labeled according to gyral and sulcal features visible in the three-dimensional data set. This procedure yielded a set of points within the same standardized stereotaxic space for each subject, which may be superimposed to create a probabilistic map of the structure in question, in this case Heschl's gyrus (see Penhune et al., 1996, for further details). In addition, the second, higher-resolution sample of MRI scans also allowed for automatic segmentation of the labeled volumes into gray- and white-matter components. The gray/ white boundary was calculated from the histogram of pixel intensity values by taking the midpoint between peak values corresponding to gray and white matter. The result of greatest relevance to the present review pertains to the estimates of the total volume of Heschl's gyrus, which is corrected for any overall differences in brain size or shape by virtue of the stereotaxic normalization procedure. In the first sample, there was a significant difference between the volume of the left and right Heschl's gyrus (Fig. 2, top panel), with 17 of the 20 subjects demonstrating a difference of 10% or more favoring the left side. This finding was replicated in the second sample of subjects, in which the asymmetry was slightly less marked but still strongly significant (Fig. 2). Most interesting of all was the outcome of the gray/white segmentation analysis: the differences in volume were found to be confined to the white matter underlying Heschl's gyrus, and not to the volume of cortical tissue within the structure (Fig. 2, bottom panel). These findings reveal an anatomical asymmetry that arises from a difference in the volume of fibers that carry information to and from the primary auditory cortex, and surrounding regions. However, the asymmetry in white matter may reflect any of a number of underlying neuronal structural differences. With in-vivo MRI we cannot ascertain, for example, whether the asymmetry reflects thalamocortical or corticocortical connections. We also cannot determine if the increased white-matter volume is a consequence of a greater number of axonal elements entering and exiting the primary cortical region, or if it may
3 18
Zatorre
indicate a higher degree of myelination of these axons. Quantitative cytoarchitectonic studies in the human brain show evidence of differential cellular organization in the left and right auditory cortices. Seldon (1981, 1982), for example, found that cell columns in the left primary auditory cortex are both wider and more widely spaced than those on the right. As well, Hutsler and Gazzaniga (1996) have recently shown that the left primary auditory region has larger layer III pyramidal cells, which would be likely to form larger columns and to send out thicker or more heavily branched axons to other regions of auditory cortex. The functional significance of such structural asymmetries is not yet clearly established, but it is interesting to speculate that they may be directly related to some of the functional asymmetries observed in many of the studies described in the preceding sections of this paper. In particular, several investigators (e.g. Tallal et al., 1993) have noted that the acoustic parameters necessary to process speech sounds entail rapid acoustic changes, particularly when tracking changes in formant transitions. Conversely, musical stimuli typically involve much slower rates of frequency change. Our data from the MRI study are consistent with this general explanation if the white-matter volume measures are related to degree of myelination. That is, a greater degree of left-sided myelination could lead to faster transmission of acoustically-relevant information, thereby permitting a specialization of left auditory cortices in the fine-grained analysis of temporal aspects of the signal, which would be highly relevant for decoding of speech sounds. On the other hand we may speculate that there is a tradeoff between speed of response and spectral selectivity. Neural systems on the left would have a fast rate of response, but their spectral tuning function would therefore necessarily be fairly wide-band, as would be appropriate to speech sounds. The right-hemisphere system would have narrower tuning functions, and thus be well-suited to the processing of stimuli containing small frequency differences, but would have a slower rate of integration in the temporal domain. This model could explain why many aspects of pitch processing relevant to music might be predominantly processed by right auditory cortical mechanisms, as reviewed in previous sections, since musical stimuli generally contain relatively slower changes, but small frequency differences are important. This account of the possible relation between structural and functional asymmetries in the human auditory cortex is at this stage very
Speech and Music 3 19 preliminary and necessarily highly speculative. It is important to put forth such ideas, however, in that they should be testable and verifiable. Perhaps more important, they point to the type of integration of evidence from multiple types of studies (behavioral, functional, structural) that will become more feasible in the near future, thanks to the development of new in-vivo imaging technologies, together with more traditional physiological and anatomical knowledge. References
Aitkin, L.M. (1986). The Auditory Midbrain. Clifton NJ: Humana press. Binder, J.R., Rao, S.M., Hammeke, T.A. et al. (1994). Functional MRI of human auditory cortex. Annals of Neurology, 35, 662-672. Blumstein, S.E., Baker, H. & Goodg!ass, H. (1977). Phonological factors in auditory comprehension in aphasia. Neuropsychologia, 15, 19-30. Brugge, J.F. & Reale, R.A. (1985).. Auditory cortex, in A. Peters and E.G. Jones (Eds.), Cerebral Cortex, Vol. 4. New York: Plenum Press, pp. 229-271. Chavis, D., & Pandya, D.N. (1976). Further observations on corticofrontal pathways in the rhesus monkey. Brain Res., 117, 369386. Colombo, M., D'Amato, M.R., Rodman, H.R., & Gross, C.G. (1990). Auditory association cortex lesions impair auditory short-term memory in monkeys. Science, 247, 336-338. Collins, D.L., Neelin, P., Peters, T.M. & Evans, A.C. (1994). Automatic 3D intersubject registration of MR volumetric data in standardized Talairach space. J. Comput. Assist. Tomogr., 18, 192-205. DEmonet, J.-F., Chollet, F., Ramsay, S., Cardebat, D., Nespoulous. J.-L., Wise, R., Rascol, A. & Frackowiak, R. (1992). The anatomy of phonological and semantic processing in normal subjects. Brain, 115, 1753-1768. DEmonet, J.-F., Price, C., Wise, R., & Frackowiack, R.S.J. (1994). A PET study of cognitive strategies in normal subjects during language tasks. Brain, 117, 671-682. Deutsch, D. (1970). Tones and numbers: Specificity of interference in short-term memory. Science, 168, 1604-1605. Deutsch, D. (1982). The processing of pitch combinations. In D. Deutsch (Ed.), The Psychology of Music. New York: Academic Press. Diamond, I.T., Fitzpatrick, D., & Sprague, J.M. (1985). The extrastraite visual cortex. In-Peters A., Jones E.G. (Eds.), Cerebral Cortex, Vol. 4. New York: Plenum Press. Divenyi, P.L. & Robinson, A.J. (1989). Nonlinguistic auditory capabilities in aphasia. Brain Lang., 37, 290-326. Dowling, W.J. & Harwood, D.L. (1986). Music Cognition. Orlando, Fla." Academic Press. Evans, A.C., Marrett, S., Neelin, P., Collins, L., Worsley, K., Dai, W., Milot, S., Meyer, E., & Bub, D. (1992). Anatomical mapping of
320
Zatorre
functional activation in stereotactic coordinate space. Neuroimage, 1, 43-53. Evarts, E.V. (1952). Effect of auditory_ cortex ablation on frequency discrimination in monkey. J. Neurophysiol., 15, 443-448. Farah, M.J (1988). Is visual imagery really visual? Overlooked evidence from neuropsychology. Psychol. Rev.,95, 307-317. FitzPatrick, K.A. & [mig, T.J. (1982). Organization of auditory connections: The primate auditory cortex. I n C.N. Woolsey (Ed.), Cortical Senso.ry Organization, VoL 3. Clifton, N.J.: Humana Press. Gainotti, G., Miceh, G., Silveri, M.C., & Villa, G. (1982). Some anatomoclinical aspects of phonemic and semantic comprehension disorders in aphasia. Acta Neurol. Scandinav., 66, 652-665. Galaburda, A. & Sanides, F. (1980). Cytoarchitectonic organization of the human auditory cortex. J. Comp. Neurol., 190, 597-610. Gottlieb, Y., Vaadia, E. & Abeles, M. (1989). Single unit activity in the auditory cortex of a monkey performing a short term memory task. Exp. Brain Res., 74, 139-148. Gross, C.G. & Weiskrantz, L. (1962). Evidence for dissociation of Impairment on auditory discrimination and delayed response following lateral frontal lesions in monkeys. Exper. Neurol., 5, 453476. Halpern, A.R. (1988). Mental scanning in auditory imagery for tunes. J. Exp. Psychol.: Learn., Mem. Cognit., 14, 434-443. Heffner, H.E. & Masterton, B. (1978). Contribution of auditory cortex to hearing in the monkey (Macaca mulatta).. In Chivers, DJ, & Herbert, J. (Ed.), Recent advances in primatology: Vol 1. Behaviour. New York: Academic Press. Hutsler, J.J. & Gazzaniga, M.S. (1996). Acetylcholinesterase staining in human auditory and language cortices6regional variation of structural features. Cereb. Cortex, 6, 260-270. Iversen, S.D. & Mishkin, M. (1973). Comparison of superior temporal and inferior prefrontal lesions on auditory and non-auditory tasks in rhesus monkeys. Brain Res., 55, 355-367. Jerison, H.J. & Neff, W.D .(1953). Effect of cortical ablation in the monkey on discrimination of auditorypatterns. Fed. Proc., 12, 237. Kosslyn, S. M., Ball, T. M. & Reiser, 13. J. (1978).. Visual images preserve metric spatial information: Evidence from studies of image scanning. J. Exp. Psychol.: Hum. Percep. Perf., 4, 47-60. Kosslyn, S.M., Alpert, N.M., Thompson, W.L. Maljkovic, V., Weise S.B., Ctiabris, C.F, Hamilton, S.E. Rauch, S.L., & Buonanno, F.S. (1993). Visual mental imagery activates topographically organized visual cortex: PET investigations. J. Cog. Neurosci., 5, 263-287. Liberman, A.M. & Mattingly, I.G. (1985). The motor theory of speech perception revised. Cognition, 21, 1-36. Li6geois-Chauvel C, Musolino A, & Chauvel P (1991). Localization of the primary auditory area in man. Brain, 114, 139-153. Matin, O.S.M. & Perry, D.W. (in press). Neurological aspects of music perception and performance. In Deutsch, D. (Ed.), The Psychology of Music, 2nd Ed.-New York: Academic Press.
Speech and Music 321 McKenna, T.M., Weinberger, N.M. & Diamond, D.M. (1989). Responses of single auditory cortical neurons to tone sequences. Brain Res. , 481, 142-153. Milner, B. (1962).. Laterality effects in audition. In Mountcastle,V.B. (Ed), Interhemispheric Relations and Cerebral Dominance, pp. 177195. Baltimore: Johns Hopkins Press. Milner, B. & Petrides, M. (1984). Behavioural effects of frontal-lobe lesions in man. Trends Neurosci., 7, 403-407. Paus, T., Pe.rry, D.W., Zatorre, R.J., W orsley, K.J. & Evans, A.C. (1996). Modulation of cerebral blood flow in the human auditory cortex during speech: role of motor-to-sensory discharges. Europ. J. Neurosci., 8, 2236-2246. Penhune, V.B., Zatorre, R.J., MacDonald, J.D., & Evans, A.C. (1996). Interhemispheric anatomical differences in human primary auditory cortex: Probabilistic mapping and volume measurement from MR scans. Cereb. Cortex, 6, 661-672. Peretz, I., Kolinsky, R., Tramo, M., Labrecque, R., Hublet, C., Demeurisse, G. & Belleville, S. (1994). Functional dissociations following bilateral lesions of auditory cortex. Brain, 117, 1283-1301. Perry, D.W., Petrides, M., Alivisatos, B., Zatorre, R.J., Evans, A.C., & Meyer, E. (1993). Functional activation of human frontal cortex dun'rig tonal working memory tasks. Soc. Neurosci. Abs., 19, 843. Petersen, S.E., Fox, P.T., Posner, M.I., Mintun, M., & Raichle, M.E. (1988). Positron emission tomographic studies of the cortical anatomy/of single word j~rocessin~. Nature, 331, 585-589. Petersen, ~5.E., Fox, P.T., rosner, M.I., Mintun, M., & Raichle, M.E. (1989). Positron Emission Tomography studies of the processing of single words. J. Cog. Neurosci., 1, 153-170. Petrides, M. & Pandya, D.N. (1988). Association fiber pathways to the frontal cortex from the superior temporal region in the rhesus monkey. J. Comp. Neurol., 273, 52-66. Petrides, M., Alivisatos, B., Meyer, M. & Evans, A.C. (1993). Functional activation of the human frontal cortex during the performance of verbal working memory tasks. Proc. Natl. Acad. ScL USA, 90, 878882. Price, C., Wise, R., Ramsay, S., Friston, K., Howard, D., Patterson, K., & Frackowiak, R.S.J. (1992). Regional response differences within the human auditory cortex when listening to words. Neurosci. Letters, 146, 179-182. Rademacher, J., Caviness, V.S., Steinmetz, H. & Galaburda, A.M. (1993). Topographical variation of the human primary cortices: complications for neuroimaging, brain mapping and neurobiology. Cereb. Cortex, 3, 313-329. Raichle, M.E., Martin, W.R.W., Herscovitch, P., Mintun, M.A., & Markham, J. (1983). Brain blood flow measured with intravenous H20. II. Implementation a n d validation. Journal of Nuclear Medicine, 24, 790-798. Rao, S.M., Binder, J.R., Bandettini, P.A., Hammeke, T.A., Yetkin, F.Z., Jesmanowicz, A., Lisk, L.M., Morris, G.L., Mueller, W.M., Estkowski, L.D., Wong, E.C., Haughton, V.M., & Hyde, J.S. (1993). Functional
322
Zatorre
magnetic resonance imaging of complex human movements. Neurology, 43, 2311-2318. Robin, D.A., Tranel, D. & Damasio, H. (1990). Auditory ~rception of temporal and spectral events in patients with focal left and right cerebral lesions. Brain Lang., 39, 539-555. Samson, S. & Zatorre, R.J. (1988). Discrimination of melodic and harmonic stimuli after unilateral cerebral excisions. Brain Cognit., 7, 348-360. Samson, S. & Zatorre, R.J. (1992). Learning and retention of melodic and verbal information after unilateral temporal lobectomy. Neuropsychologia, 30, 815-826. Samson, S. & Zatorre, R.J. [1994). Contribution of the right temporal lobe to musical timbre discrimination. Neuropsychologia, 32, 23 1240. Seldon, H.L. (1981). Structure of human auditory cortex II: axon distributions and morphological correlates of speech perception. Brain Res., 2 2 9, 295-310. Seldon, H.L. (1982). Structure of human auditory cortex III: statistical analysis of dendritic trees. Brain Res., 249, 211-221. Sidtis, J.J. & Volpe, B.T. (1988). Selective loss of complex-pitch or speech discrimination after unilateral lesion. Brain Lang., 34, 235245. Stepien, L.S., Cordeau, J.P. & Rasmussen, T. (1960).. The effect of temporal lobe and hippocampal lesions on auditory and visual recent memory in monkeys. Brain, 83, 470-489. Steinschneider, M., Schroeder, C.E., Arezzo, J.C., & Vaughan, H.G. (1995). Physiologic correlates of the voice-onset time boundary in primary auditory cortex of awake monkey: temporal response patterns. Brain Lang., 48, 326-340. Talairach, J., & Tournoux, P. (1988). Co-Planar Stereotaxic Atlas of the Human Brain. New York: Thieme. Tallal, P., Miller, S. & Fitch, R.H. (1992). Neurobiological basis of ,~Pceech: a case for the preeminence of temporal processing. Ann. N.Y. ad. Sci., 682, 27-47. Tramo, M.J. & Gazzaniga, M. (1989). Discrimination and recognition of complex harmonic spectra by the cerebral hemispheres: Differential lateralization of acoustic-discriminative and semanticassociative functions in auditory pattern perception. Soc. Neurosci. Abstr., 15, 1060. VonEconomo, C. & Horn, L. (1930). Uber Windungsrelief, Masse und Rindenarchitektonik der Supratemporalflache, ihre individuellen und ihre Seitenunterschiede. Z. Neurol. Psychiat., 130, 678-757. Whitfield, I.C. & Evans, E.F. (1965). Responses of auditory cortical neurones to stimuli of changing frequency. J. Neurophysiol., 28, 655-672. Wise, R.J., Chollet, F., Hadar, U., Friston, K., Hoffner, E., & Frackowiak, R., (1991). Distribution of cortical neural networks involved in word comprehension and word retrieval. Brain, 114, 1803-1817. Worsley, K.J., Evans, A.C., Marrett, S., & Neelin, P. (1992). A threedimensional statistical analysis for CBF activation studies in human brain. J. Cereb. Blood Flow Metab., 12, 900-918.
Speech and Music 323 Zatorre, R.J. (1985). Discrimination and recognition of tonal melodies after unilateral cerebral excisions. Neuropsychologia, 23, 31-41. Zatorre, R.J. (1988). Pitch perception of complex tones and human temporal-lobe function. Z Acous. Soc. Amer., 84, 566-572. Zatorre, R.J., Evans, A.C., Meyer, E., & Gjedde. A. (1992). Lateralization of phonetic and pitch processing in speech perception Science, 256, 846-849. Zatorre, R.J., Evans, A.C., & Meyer, E. (1994). Neural mechanisms underlying melodic perception and memory for pitch. J. Neurosci., 14, 1908-1919. Zatorre, R.J. & Halpern, A.R. (1993). Effect of unilateral temporal-lobe excision on auditory perception and imagery. Neuropsychologia, 31, 221-232. Zatorre, R.J., Halpern, A.R., Perry, D.W., Meyer, E. &Evans, A.C. (1996b). Hearing in the mind's ear: A PET investigation of musical imagery and perception. J. Cognit. Neurosci., 8, 29-46. Zatorre, R.J., Meyer, E., Gjedde, A., & Evans, A.C. (1996a). PET studies of phonetic processing of speech: Review, replication, and reanalysis. Cereb. Cortex, 6, 21-30. Zatorre, R.J. & Samson, S. (1991). Role of the right temporal neocortex in retention of pitch in auditory short-term memory. Brain, 114, 2403-2417.
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
325
Chapter 11
Perceptual and cognitive development: Electrophysiological correlates. Dennis L. Molfese & Dana B. Narter Southern Illinois University
Infants at birth clearly are equipped to discriminate many aspects of their environment, including basic elements that are relevant later in development to the emergence of language (Molfese, 1972; Molfese, Freeman, & Palermo, 1975; Molfese & Molfese, 1979b; 1980; 1985; 1997). At birth, infants appear quite able to discriminate a number of speech sounds in a manner similar to that performed by adults (see Molfese & Betz, 1988, for a review). This has been confirmed in both behavioral studies (Eimas, Siqueland, Jusczyk, & Vigorito, 1971) as well as in neuroelectrical studies involving event related potentials (Molfese, 1978a; Molfese & Hess, 1978). Yet, while such abilities may be present at birth, it is equally clear that these skills and the brain mechanisms which subserve them develop further over time and are modified by the environment. Furthermore, it appears that such early skills may be directly related to later emerging cognitive skills including language. A growing body of literature now exists which suggests a strong relationship exists between phonologically based abilities and language skills such as reading (Lyon, 1994). Furthermore, there are an increasing number of reports in the science literature suggesting that the infant's early phonetic discrimination skills have direct relevance to later language development. Infants who experience difficulty in discriminating phonetic contrasts appear more at risk for lower levels of later
326
Molfese and Narter
language development (Moliese & Molfese, 1985; 1997; Molfese & Searock, 1986). Given that such relationships exist between the child and adult years for cognitive and language processes, it appears likely that such relationships also must exist between various cognitive domains during the infant years as well. Thus, one could expect to find a relationship between early phonological skills and the development of other cognitive processes such as memory and language. One such relationship that could be expected between speech perception, memory, and word processing is a temporal one. When words are heard by the infant, one might expect the auditory and cortical systems to first detect and process this acoustic information, then perhaps to compare it to remembered information, before extracting and understanding the meaning of the utterance. Thus, if we examined the infant's brain responses to a word that is known to that infant, we might expect speech sounds of the utterance to produce changes in the early portion of the brain response, followed in time by changes in later brain wave components which signal the involvement at some level of the memory system (recognition and/or recall), and finally the appearance of changes in the brain response which signal the understanding of that word. Alternatively, similar mechanisms could underlay different cognitive areas in infancy. Differences in ERP latencies and response patterns while infants are engaged in different tasks (such as speech perception, memory, or language) must argue for different brain mechanisms which generate these differences, similarities in brainwave characteristics across these different cognitive domains could argue for their reliance at some level on identical brain mechanisms. The following review focuses on studies utilizing a number of paradigms utilized over the past 3 decades with infants which indicate that event related potentials (ERPs) are sensitive to phonetic variations (Molfese and Molfese, 1979b, 1980, 1985), can detect differences in memory (Courchesne, Ganz, & Norcia, 1981; Cowan, Suomi, Morse, 1982; DeHaan & Nelson, 1997; Nelson & Salapatek, 1986), and early word acquisition (Molfese, 1989, 1990; Molfese, Morse, & Peters, 1990; Molfese, Wetzel, & Gill, 1994). While these three areas of abilities will be reviewed separately, attention will be paid to similarities in cortical event related brain potentials (ERPs) elicited during these different events with the hope of identifying whether there are patterns of similarities in their development. At the very least, we hope to identify developmental patterns which reflect changes in speech perception, memory, and early
Cognitive Development
327
language abilities over time. At the most, we hope to identify more specific relationships which might exist between these different domains of cognitive abilities during infancy. Given the prevalence of hemisphere differences noted during language related tasks (Bryden, 1982), we also anticipate that the ERPs should be distributed differentially across the two hemispheres depending upon whether they are elicited by speech or language materials versus non-language stimuli. The first part of this review will focus the infant's abilities to discriminate different speech sounds, on what abilities appear to develop first, and how the brainwave patterns change over time, appearing to reflect different patterns of brain involvement. Next, we will review the neuroelectrophysiological literature on memory, focusing more specifically on electrophysiological studies which track infants' abilities to discriminate frequent from infrequent events. From that point the review will move on to the study of the young infant's ability to discriminate between words as they are acquired. The review at this point will attempt to determine whether comparable brain responses are elicited by the infant's attempts to discriminate between speech sounds, their abilities to discriminate familiar from novel events, and by their responses to words that are thought to vary in their meaningfulness. First, however, the research method used to study developmentally these early phonetic, memory, and language in infants will be described. Event Related Potentials: Eleetrophysiologieal technique used to study Infant cognition. The ERP is a synchronized portion of the ongoing EEG pattern that is time-locked to the onset of some event in the infant's environment (Rockstroh, Elbert, Birbaumer, & Lutzenberger, 1982). It is usually represented as a complex waveform that reflects changes in the amplitude and frequency of electrical activity over time. This waveform reflects changes in brain activity via fluctuations in the amplitude or latency of various positive or negative peaks which occur at different points throughout its time course (Callaway, Tueting, & Koslow, 1978). Research over the past 70 years has demonstrated that the ERP, because of this time-locked feature, can be used to effectively study both general and specific aspects of the organism's response to specific stimuli (Molfese, 1978a, 1978b) as well as that individual's perceptions and decisions (Molfese, 1983; Nelson & Salapatek 1986; Ruchkin, Sutton, Munson, Silver, & Macar, 1981). Given that the ERP technique does not require a planned and overt response from the
328
Molfese and Narter
individuals from which it is recorded, it is particularly well-suited for the neuropsychological study of the infant's cognitive development. Phonetic Discriminations and ERPs. A large number of studies conducted across the past two decades have investigated the neuroelectrical correlates of speech perception from the infancy period into adulthood. These results suggest that some of the infant's perception of speech cues such as voice onset time (VOT) follow different developmental progressions than other cues such as place of articulation (POA). The two most notable differences are that (1) VOT discrimination along phonetic boundaries does not consistently appear in newborn infants and (2) VOT appears to elicit bilateral responses (i.e., equivalent responses from both hemispheres)earlier in the waveform than lateralized responses (Molfese & Molfese, 1979a; Simos & Molfese, 1997). POA, on the other hand, (1) does consistently appear in newborn infants and (2) evokes an initial lateralized response in the waveform which is then followed later in time by a bilateral responses (Molfese, Burger-Judisch, & Hans, 1991; Molfese & Molfese, 1979b, 1980, 1985). Building on a base of behavioral research (Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967), studies by Molfese and colleagues of VOT and POA have covered the developmental period extending from infancy (Molfese & Molfese, 1979a, 1979b,1980) into childhood and adulthood (Molfese, 1978a, 1978b, 1980a, 1980b, 1984; Molfese & Hess, 1978; Molfese & Schmidt, 1983). For the purposes of our review, we have limited our review of the adult work to a few exemplars which provide a framework for interpreting the infant research which is more relevant to the theme of this chapter. Voice Onset Time
Voice onset time (VOT), the temporal relationship between laryngeal pulsing and the onset of consonant release, is an important cue for the distinguishing voiced from voiceless forms of stop consonants such as b and p_ (Liberman, Cooper, Shankweiler, & Studdert-Kennedy, 1967). Adult listeners appear to discriminate a variety of speech sounds by the phonetic labels attached to them as evidenced by the fact that they readily discriminate between consonants from different phonetic categories, such as [ba] and [pa], while they perform at only chance levels when attempting to discriminate between different exemplars from the same phonetic category (Lisker & A bramson, 1970). This pattern of
Cognitive Development
329
discrimination for between phonetic category contrasts while chance levels of discrimination are noted for within-category contrasts is referred to as "categorical perception". Studies with infants (Eimas et al., 1971; Eilers, Wilson, & Moore, 1979), children (Streeter, 1976), and adult listeners (Lisker & Abramson, 1970) consistently have demonstrated that adults and children from many different language environments exhibit categorical perception and discrimination for a wide range of consonant contrasts such as voicing (as in the case of discriminating the initial consonant sound /b/ of the consonant vowel (CV) syllable b_aa from the initial consonant sound /p/ of p__a or discriminating ga from k__a)and place of articulation (discriminating the initial consonant sounds from each other of the consonant vowel syllables, b__aa,d__aa,and ga). While we already knew from Eimas et al. (1971) that young infants appear to have categorical-like perception for stop consonants, little was known regarding how the brain responded to these sounds and how such responses changed across development. Molfese (1978b), in a follow-up to work by Dorman (1974), provided the first evidence that speech cues such as VOT could elicit different ERP responses. To accomplish this, Molfese recorded ERPs from the left and fight temporal regions of 16 adults during a phoneme identification task. Adults listened to randomly ordered sequences of synthesized bilabial stop consonants with VOT values of +0 ms, +20 ms, +40 ms, and +60 ms. In the +0 ms case, the onset of consonant release and vocal fold vibration occurred simultaneously, whereas in the +60 ms condition the onset of laryngeal pulsing was delayed for 60 ms following consonant release. ERPs were recorded in response to each sound and then, after a brief delay, adults pressed a series of keys to identify the consonant sound. Two regions of the ERP (one component centered around 135 ms and the second occurring between 300 and 500 ms following stimulus onset) did change systematically as a function of the consonant's phonetic category. Stop consonant sounds with VOT values of +0 and +20 ms (sounds identified as b_a) were discriminated from those with VOT values of +40 and +60 ms (sounds identified as p__a).However, the ERPs did not discriminate between the sounds from within the same category. Thus, there were no differences in the waveforms between the 0 and +20 ms sounds or between the +40 and +60 ms sounds. This pattern of responding resembled what Lisker and Abramson called categorical discrimination. That is, ERPs could discriminate between sounds from
330
Molfese and Narter
different phonetic categories but not between sounds from the same phonetic category. Electrophysiological studies employing similar stimuli with various populations have replicated this finding (Molfese & Hess, 1978; Molfese & Molfese, 1979a; Molfese, 1980a Molfese & Molfese, 1988). Surprisingly, however, in all these studies at least one region of the ERP in which this categorical discrimination effect was noted across the different age groups occurred over the RIGHT temporal region. This sense of surprise comes from the expectation that such speech sound discrimination effects would be expected to be left hemisphere in origin since the left hemisphere, not the right, is usually associated with language functions (Lenneberg, 1967). Similar effects were noted with 4 year old children in a study involving the velar stop consonants, k and g:. Molfese and Hess (1978) recorded ERPs from the left and right temporal scalp regions of 12 preschool-age children (mean age = 4 years, 5 months) in response to randomly ordered series of synthesized consonant-vowel syllables in which the initial consonant varied in VOT from +0 ms, to +20 ms, to +40 ms, to +60 ms. In their analyses of the ERPs, they, like Molfese (1978b), also found a categorical discrimination effect whereby one late-occurring portion of the waveform (peak latency = 444 ms) changed systematically in response to consonants from different phonetic categories but did not respond differentially to consonants from within the same phonetic category. Also as in the case of Molfese (1978b), this effect occurred at the right hemisphere electrode site. Unlike the adult study by Molfese, however, they found a second portion of the auditory ERP that occurred earlier in the waveform (peak latencies = 198 and 342 ms), before this right-hemisphere effect, and which was detected by electrodes placed over both hemispheres. This earlier occurring bilaterally detected auditory ERP component also discriminated voiced from voiceless consonants in a categorical manner. Similar results were reported by both Molfese (1980b) and Segalowitz and Cohen (1989) with different populations of adults and by Molfese and Molfese (1988) with 3 year old children. This work has been extended to include newborn and older infants by Molfese and Molfese (1979a), who presented the four consonantvowel syllables used by Molfese (1978b) to 16 infants between 2 and 5 months of age (mean = 3 months, 25 days). ERPs were again recorded from the left and right temporal electrode sites, T3 and T4. Analyses revealed that one portion or component of the auditory ERP, recorded
Cognitive Development
331
from over the right hemisphere approximately 920 ms following stimulus onset, discriminated between the different speech sounds in a categorical manner. As in the case of Molfese and Hess, they also noted a second portion of the auditory ERP that was detected at electrode sites over both hemispheres and that also discriminated between the consonant sounds categorically. The major portion of this component occurred 528 ms following stimulus onset. Thus, these results paralleled the findings of Molfese and Hess in noting two portions of the auditory ERP that discriminated between the speech sounds categorically. These included a bilateral component that occurred first in the waveform, followed by a right-hemisphere lateralized component that occurred later in time and also discriminated between the sounds categorically. A second experiment described by Molfese and Molfese (1979a) failed to note any such bilateral or fight hemisphere lateralized effects related to VOT discrimination with 16 newborn infants under 48 hours of age. However, a subsequent study by Simos and Molfese (1997) did find such effects with a different group of 16 newborn infants using nonspeech auditory stimuli (TOT) which mimicked the temporal delays of the speech stop consonant voiced vs. voiceless distinction (see also Molfese, 1980a). These TOT stimuli were obtained from Pisoni (1977) and had previously been used to study electrophysiological correlates of temporal differences in both adults (Molfese, 1980a) and children (Molfese & Molfese, 1988). The TOT stimuli were 230 ms in duration and consisted of two simultaneously occurring tones which differed from each other in the temporal onset of the lower frequency tone (500 Hz) relative to the higher frequency tone (1500 Hz). The lower tone began at the same time as the upper tone for the 0-ms TOT stimulus but lagged behind the upper tone by 20 ms for the +20-ms TOT stimulus. This delay increased to 40 ms and 60 ms, respectively, for the +40- and +60-ms TOT stimuli. Both tones ended simultaneously. Simos and Molfese did find categorical-like discrimination effects at both the N200 and N530 negative peaks over parietal electrode sites and interpreted these results to indicate that the "temporal voicing cue used in speech perception may have an innate basis" (page 89). Notably, the latency for the bilateral effect reported for these non-speech stimuli was virtually the same as that reported by Molfese and Molfese (1979a) with a different population of infants who were listening to speech stimuli. This similarity in responses to speech and nonspeech temporal stimuli was also reported by Molfese (1980a) in adults and by Molfese and
332
Molfese and Narter
Molfese (1988) with preschool children. One other study has included speech stimuli varying in VOT with young infants (Kurtzberg, Hilpert, Kreuzer, & Vaughan, 1984). However, Kurtzberg et al. did not include an analysis of these speech contrasts so the material is not relevant for this review. An overall summary of the infant speech perception research for VOT as well as for a second speech cue, POA, that will be reviewed below, is outlined in Table 1. Table 1. Studies Using Event-Related Potential (ERP) Procedures To Study
Infant Speech Perception. Voice Onset Study/Ss
Time Electrode
Molfese & Molfese (1979)
T3, T4
Newborn infants, n= 16
Simos & Molfese (1997) Newborn infants, n= 16
Kurtzberg, Hilpert, Kreuzer, & Vaughan (1984) Term, and very low-birthweight infants tested at 40 weeks gestational age and again at 1-, 2- and 3months,
Reference= linked ears.
FL, FR T3, T4 PL, PR Reference= linked ears. Midway between Fz and Pz, midway between left mastoid and C3, midway between fight mastoid and C4. Reference --midoccipital electrode.
Task Auditory ERPs; 4 Bilabial stop consonants with VOT values of 0-, 20-, 40-, & 60msecrepeatedly presented in random orders.
Results
Auditory ERPs
Bilateral N200 & N530 Parietal Categorical-like effect discriminating 0-, 20ms from 40-, 60-ms stimuli.
0-, 20-, 40-, 60msec TOT repeatedlypresented in random orders. Audi tory ERPs /da/and/ta/, 800Hz tone
No Categorical-like VOT effects
Did not test for differencesbetween speech sounds.
Cognitive Development
333
Table 1: cont'd Study/Ss Molfese & Molfese (1979) 2- to 5- month old infants, n= 12
Electrode T3, T4
Task Auditory ERPs
Reference = linked ears.
4 Bilabial stop consonants with VOT values of 0-, 20-, 40-, & 60msecrepeatedly presented in random orders.
Place of Articulation Study Electrode Molfese & T3, T4 Molfese, 1980 Reference = Preterm infants, linked ears. 35.9 weeks, n=ll
Molfese & Molfese, 1979 Newborns, n - 16
T3, T4 Reference = linked ears.
Results RH P920 Categoricallike effect discriminating 0,20-ms from 40-, 60ms stimuli. Bilateral N528 Categorical-like effect discriminating 0-, 20ms from 40-, 60-ms stimuli.
Task Auditory ERPs. 2 phonetic and 2 nonphonetic versions of bae,gae plus 4 nonspeech controls matched to the center frequencies of each. Presented in random order with equal probabilities of occurrence.
Results LH N848 POA
Auditory ERPs. /bae/and/gae/ speech syllables and nonspeech controls matched to the center frequencies of each. Presented in random order with equal probabilities of occurrence.
LH N192 POA Bilateral N630 POA
334
Molfese and Narter
Table 1: cont'd Study/Ss Molfese, BurgerJudisch, & Hans, 1991
Electrode F L , FR T3, T4 P L , PR
Newboms, n=38
Reference= linked ears.
Molfese & Molfese, 1985
T3, T4
Auditory ERPs.
Reference = linked ears.
9 CV syllables combining b, d, g with i, ae, au
Newborns, n= 16
Task Auditory ERPs. /bi/and/gi/speech syllables and nonspeech controls matched to the center frequencies of each. Presented in random order with equal probabilities of occurrence.
Results 210 ms. LH Females: N210 Frontal & Parietal POA
LH N 168 POA Bilateral N664 POA
Presented in random order with equal probabilities of occurrence. DahaeneLambert & Dahaene (1994) 3-month old infants, n= 16
58electrodes in geodesic net
Auditory ERPs. ba,ga Two conditions: (1) REPEATED TRIALS-standard CV presented 5 times in a row (2) DEVIANT TRIALS-standard CV presented 4 times in a row followed by different CV syllable.
390 ms. ba response differed from ga LH>RH over parietal sites
Cognitive Development
335
Table 1, cont'd. Vowel Sounds Electrode Study Molfese & T3, T4 Searock, 1986 Reference= linked ears. 12-month-old infants, n= 16.
Infants divided into 2 groups, High and Low Groups, based on Median Split of McCarthy Verbal Scores at 3-years-of age.
Task Auditory ERPs.
3 Vowel sounds, I, ae, au, and their nonspeech controls matched to the center frequencies of each vowel. Presented in random order with equal probabilities of occurrence.
Results P60 RH discriminated I from au.
N200 High Group discriminated nonspeech control I from ae and ae from au. P300 High Group RH discriminated speech vowels I from ae, RH discriminated nonspeech vowels I from ae and I from au. LH discriminated ae from au. Low Group RH discriminated ae from an.
Although the right-hemisphere discrimination of the VOT cue appears paradoxical in light of arguments that language processes are carried out primarily by the left hemisphere, the fact that identical responses are elicited by both speech and nonspeech sounds which contain the same temporal cues suggests that it may be the temporal quality of the sounds, not their speech-like quality, which in fact triggers the right hemisphere response. Furthermore, studies of clinical populations suggest that the V OT cue is discriminated, if not exclusively, then at least in part, by brain mechanisms restricted to the right hemisphere (for a review of this literature, see Molfese, Molfese, & Parsons, 1983, or Simos, Molfese, & Brenden, 1997). For example, Miceli, Caltagirone, Gianotti, and Payer-Rigo (1978), using a nondichotic pair presentation task, noted that the left-brain-damaged aphasic group made fewest errors with stimuli differing in voicing but not place of articulation. Blumstein, Baker, and Goodglass (1977) also
336
Molfese and Narter
noted fewer errors for voicing contrasts than for place contrasts with left hemisphere damaged Wemicke aphasics. Finally, Perecman and Kellar (1981), based on their own findings that left-hemisphere-damaged patients continue to match sounds on the basis of voicing but not place, speculated that voicing could be processed by either hemisphere but that the POA cue was more likely to be processed by only the left hemisphere. Three general findings have emerged from this series of temporal discrimination studies involving VOT and TOT. First, the discrimination of the temporal delay cue common to voiced and voiceless stop consonants can be detected by ERPs recorded from electrodes placed on the scalp over the two hemispheres. Second, from at least 2 months of age, if not before, the infant's brain appears capable of discriminating voiced from voiceless stop consonants in a categorical manner. That is, the ERPs appear to discriminate stimuli in one phonetic category from those with VOT values which characterize a second phonetic category. At the same time, these ERPs can not discriminate between different VOT stimuli that come from the same phonetic category. Third, categorical discrimination across different ages appears to be carried out first by bilaterally represented mechanisms within both hemispheres and then, somewhat later in time, by right-hemisphere lateralized mechanisms. The bilateral effects appear to be reflected with some consistency in the negative peak that occurs with a latency of approximately 530 ms in infants from birth onward. The lateralized effect, when noted in the infant ERPs, has a markedly longer latency. The presence of several different peaks with markedly different latencies which are responsive to the same temporal cues may signal that multiple regions of the brain are responsive to and perhaps process differently these voicing or temporal contrasts.
Place of Articulation (POA) In addition to studies of VOT, a second speech cue, place of articulation or POA, has been investigated in a number of studies with infants and adults (Molfese, 1978a, 1980b, 1984; Molfese, Buhrke, & Wang, 1985; Molfese, Linnville, Wetzel, & Leicht, 1985; Molfese & Schmidt, 1983, Molfese & Molfese, 1979b, 1980, 1985). As in the case of the VOT temporal cue, these studies of the POA cue identified both lateralized and bilateral hemisphere responses that discriminated
Cognitive Development
337
between different consonant sounds. However, unlike the discrimination abilities for VOT, the ability to discriminate the POA cue consistently appeared to be present from birth. There were some important differences, however, both in the development of ERP responses to the POA cue and in the character of the lateralized responses which distinguished the perception of this cue from that for VOT. Molfese (1978a) first attempted to study POA in adults in order to obtain some reference for studying these abilities in infants. For the most part, this research focused on attempts to isolate the neuroelectrical correlates of the second formant transition, the cue to which listeners attend in order to discriminate between different consonant sounds which are formed in different portions or places within the vocal tract. In this study, Molfese presented a series of consonant-vowel syllables in which the stop consonants varied in POA, formant structure, and phonetic transition characteristics. Changes in the POA cue (i.e., changes in the second formant transition) signaled either the consonants b or g. The formant structure variable referred to two sets of sounds, one set of which consisted of nonspeech sounds that contained formants composed of sinewaves 1 Hz in bandwidth whereas a second set of speech sounds contained formants with speech-like formant bandwidths of 60, 90, and 120 Hz for formants 1 through 3, respectively. The phonetic transition cue referred to two stimulus properties in which one stimulus set contained formant transitions that normally characterize human speech patterns while the second set contained an unusual pattern not found in the initial consonant position in human speech patterns. Auditory ERP responses were recorded from the left and right temporal regions of 10 adults in response to randomly ordered series of CV syllables that varied in consonant place of articulation, bandwidth, and phonetic transition quality. Two regions of the auditory ERP that peaked at 70 and 300 ms following stimulus onset discriminated the phonetic transition and POA cues only over the left-hemisphere temporal electrode site. As in the case of Molfese (1978b) who also used only a single left-hemisphere temporal site, no bilateral place discrimination was noted. Similar left hemisphere POA discrimination effects were noted by Molfese (1980b), Molfese and Schmidt (1983), Molfese (1984), and Gelfer (1987) with the exception that, with the inclusion of auditory ERP data collected from more electrode recording sites over each hemisphere, consistent discrimination of the place cues were also noted to occur at the same time over both hemispheres (bilateral effects).
338
Molfese and Narter
Several general findings have emerged from these adult POA studies. First, when multiple electrode sites are employed, bilateral stimulus discrimination effects are usually found in addition to left hemisphere lateralized ones. Second, these bilateral effects invariably occur early in the waveform and prior to the onset of the left hemisphere lateralized POA discrimination responses. This temporal relationship between bilateral and lateralized effects was noted earlier in our review of the VOT studies. Third, in addition to stimulus related hemisphere effects, portions of the ERPs also vary between hemispheres that are unrelated to stimulus, task, or subject features. Apparently, during the discrimination of these auditory tokens, both hemispheres initially discriminate between POA and VOT/TOT stimuli at the same time in adults, somewhere approximately 100 ms following stimulus onset. Shortly afterwards, at approximately 300 ms following stimulus onset, the left hemisphere discriminates between differences in the POA cue, while the right hemisphere at approximately 400 ms discriminates the VOT or temporal offset cue. Finally, throughout this time period and afterwards there are brief periods of ERP activity during which the two hemispheres appear to be doing quite different things, which may be unrelated to the discrimination of the stimuli. In an extension of these POA findings to younger populations, Molfese and Molfese (1979b) noted similar patterns of lateralized and bilateral responses with newborn and young infants. Unlike findings for VOT, however, POA discrimination were consistently found to be present at birth. In this study, ERPs were recorded from the left and fight temporal regions (T3 and T4) referred to linked ear references of 16 full term newborn human infants within 2 days of birth. These data were recorded while the infants were presented series of consonantvowel syllables that differed in the second formant transition (F2, which signaled POA information), and formant bandwidth. As with adults, one auditory ERP component that appeared only over the left-hemisphere recording site discriminated between the two consonant sounds when they contained normal speech formant characteristics (peak latency = 192 ms). A second region of the auditory ERP varied systematically over both hemispheres and also discriminated between the two speechlike consonant sounds (peak latency = 630 ms). Notably, these latencies were markedly shorter than those found for VOT and TOT in young infants.
Cognitive Development
339
In a subsequent replication and extension of this work, Molfese and Molfese (1985) presented a series of consonant-vowel syllables that varied in POA and formant structure. Two different consonant sounds, b, g, combined with three different vowel sounds were presented with speech or nonspeech formant structures. ERPs were again recorded from the left and right temporal regions (T3, T4). As in the case of Molfese and Molfese (1979b), analyses identified two regions of the auditory ERP that discriminated POA differences. One region, with a peak latency of 168 ms, was detected only over the left-hemisphere site as discriminating between the two different consonant sounds; a second region with a peak latency of 664 ms, discriminated this POA difference and was detected by electrodes placed over both hemispheres. The lateralized effect noted for infants in the Molfese series of infant studies for the POA cue occurred prior to that for the bilateral effect, a finding opposite to that noted when adults were studied. However, the reversal of the temporal relationship between the bilateral and lateralized responses appears to be a legitimate one, given that virtually identical results were found by Molfese and Molfese (1985) and Molfese and Molfese (1979b) with different populations of infants and somewhat different stimulus sets which contained the POA variable. A replication and extension of this work which involved recorded ERPs from 6 scalp locations of 38 newborn infants to a somewhat different stimulus set reported comparable effects at similar latencies (Molfese, BurgerJudisch, & Hans, 1992). This temporal pattern of initial lateralized responses followed by bilateral responses is also opposite to that noted previously for VOT cues for adults as well as that found for infants exposed to changes in the VOT temporal cue. Clearly, such differences in the ERP effects suggest that different mechanisms subserve the perception and discrimination of the different speech related cues. The relationship between lateralized and bilateral responses is not clear at this time. It does appear, however, that bilateral responses may develop after the lateralized ones for POA, both ontogenetically as well as phylogenetically. For example, Molfese and Molfese (1980) noted only the presence of left-hemisphere lateralized responses in 11 preterm infants born on average 35.9 weeks postconception. Stimuli identical to those employed in Molfese (1978b) with adults were presented to these infants while ERPs were recorded from the left- (T3) and righthemisphere (T4) temporal regions. As was found with full-term infants (Molfese & Molfese, 1979b), a portion of the auditory ERP recorded
340
Molfese and Narter
from over the left hemisphere with a peak latency at 848 ms discriminated between speech stimuli containing different consonant transition cues. An additional left-hemisphere component with a peak latency of 608 ms differentiated only between the nonphonetic consonants, a finding similar to that reported by Molfese (1978b) with adults, with the exception that adults were sensitive to both phonetic and nonphonetic contrasts. While most studies of POA in infants involve newborns, one study by Dehaene-Lambert and Dehaene (1994) noted a POA effect in older infants at 3 months of age. They tested a group of 16 infants, recording ERP activity from a series of 58 scalp electrodes to ba and ga syllables. Two conditions were used: a repeated trials condition in which a standard sound was presented five times and a deviant trial condition in which the standard was repeated four times, followed by one instance of a different syllable. Consonant discrimination changes were noted at one ERP peak, at 390 ms, which declined in amplitude as the standard stimulus was presented and then increased in amplitude with the presentation of the different syllable. Thus, the ERPs detected a difference in the speech sounds and recovered in amplitude in contrast to repetitions of the same stimulus which resulted in further decreases in amplitude. They also noted a moderate LH asymmetry for this peak over posterior electrode sites. While this study is consistent with the neonatal research in reporting LH lateralized effects in young infants, the latency of the response differs from that reported by Molfese and colleagues. In addition, there is no report of a bilateral effect. However it is difficult to determine whether such differences result from the different paradigms used, the differences in the ages of the infants sampled across the studies, or other factors. It is clear that much more research is needed to fill in the missing gaps in our knowledge of the neuroelectrical correlates of POA from early infancy into the late adolescent years. In summary, unlike the VOT studies, the POA cue evokes a relatively stable pattern of lateralized and bilateral responses from infancy into adulthood (Gelfer, 1987; Molfese, 1978b, 1980b; Molfese, Buhrke, & Wang, 1985; Molfese, Linnville, Wetzel, & Leicht, 1985; Molfese & Schmidt, 1983; Molfese & Molfese, 1979b, 1980, 1985). These effects appear to replicate well across laboratories (Gelfer, 1987; Segalowitz & Cohen, 1989), although a great deal more research is needed to characterize the changes in both the neuroelectrical correlates of VOT and POA through the infant and child years.
Cognitive Development
341
Vowel Sounds
There is one published study investigating vowel discrimination abilities in one-year old infants (Molfese & Searock, 1986). They recorded auditory ERPs from 16 Caucasian infants tested within two weeks of their one-year birthdate. The infants heard three vowel sounds, 300 ms in duration, with normal speech formant characteristics (i,ae,au) and three nonspeech controls which contained one-hertz sinewave formant bandwidths instead of the normal speech formant bandwidths of 60, 90, and 120 Hz for formants l, 2, and 3, respectively. Auditory stimuli were presented randomly at 80 dB SPL(A) and ERPs were collected from scalp electrodes placed at the left and right temporal areas, T3 and T4, referred to linked ears (Jasper, 1958). Upon completion of the testing, the infants were divided using a median split into two groups based on their McCarthy verbal scores at three years of age. The High Group had a mean McCarthy verbal score of 77.25 (s.d.=15.5) and the Low Group had a mean McCarthy score of 20.5 (s.d.=12.6). Three regions of the ERP changed following presentation of the vowel sounds. An initial positive component (P60) changed systematically over the RH of both groups to the vowel sounds, i and au. A second region of the waveform, the N190, which characterized a negative peak at 190 ms post vowel sound onset, only changed systematically for the High Group at the RH temporal site and discriminated the nonspeech control sounds, i from ae and ae from au. Finally, a positive peak at 340 ms discriminated at the RH site the Vowel speech sounds, i from ae, and the nonspeech sounds, i from ae and i from au. The only LH discrimination of sounds for the High Group occurred in response to the nonspeech control sounds for ae versus au. Only one discrimination over the RH was noted for the Low Group, ae from au. On the basis of these electrophysiological data, it appears that brain responses to speech materials from infancy into adulthood are multidimensional and that they develop in a dynamic fashion. First, it is clear that discrimination of different speech cues emerge at different times in early development. This is true from both the standpoint of behavioral research (Eimas, et al., 1971) as well as ERP research (Molfese & Molfese, 1979a, 1979b, 1985, 1997). Relatively stable and reliable ERP correlates of consonant place of articulation (POA) discrimination have been noted in newborns. At the same time, however, discrimination of a different speech cue, voice onset time (VOT), does
342
Molfese and Narter
not appear to develop until sometime after birth, at least in the majority of the population (Molfese & Molfese, 1979a; Simos & Molfese, 1997). Second, different regions of the auditory ERP elicited by the different auditory stimuli appear to lateralize differently, depending on the evoking stimuli. The temporal cue, VOT, elicits a differential righthemisphere response, while the POA cue elicits a differential lefthemisphere response. Third, the scalp distributions for ERP effects in relation to speech sound discrimination change with development. Thus, for example, Molfese and Molfese (1979a) note temporal lobe lateralized effects in newborn infants, while more pronounced temporalparietal effects are noted in children (Molfese & Molfese, 1988) and adults (Molfese, 1978a). The fourth point is that different portions of the ERP waveform appear sensitive to phonetic speech sound contrasts at different developmental stages. Thus, shortly after birth, speech sound discriminations are noted to occur at relatively long latencies (520 - 920 ms, see Molfese & Molfese, 1979b; Simos & Molfese, 1997), while these effects shift forward in the ERP wave to 180 - 400 ms for one-year-olds (Molfese & Searock, 1986) and for preschoolers (Molfese & Hess, 1978; Molfese & Molfese, 1988), and from 50 to 350 ms for elementary school children and adults (Molfese, 1978a, 1978b, 1980a, 1980b). Fifth, and finally, at some point during the auditory ERP to virtually all stimuli tested to date using this procedure, the two hemispheres, in both infants and adults, respond differently to all stimuli. This general hemisphere difference seems most pronounced in the preterm infants, with many different regions of the ERP varying between the two hemispheres (Molfese & Molfese, 1980). However, this difference is also present in newborn (Molfese & Molfese, 1979b, 1985), one-year-old infants (Molfese & Searock, 1986), preschool age children (Molfese & Hess, 1978; Molfese & Molfese, 1988), and adults (Molfese, 1978a, 1978b, 1980a, 1980b, 1984; Molfese & Schmidt, 1983). Electrophysiological correlates of Infant memory.
While our knowledge of infants' speech perception has expanded rapidly over the past two decades as noted above (Eimas, Siqueland, Jusczyk, & Vigorito, 1971; Kuhl, 1985; Morse, 1974; Morse & Snowdon, 1975: Molfese & Molfese, 1979b; 1980b; 1985), little is known concerning how particular speech sound patterns come to be
Cognitive Development
343
recognized and discriminated as a function of the differential experience that eventually accompanies early word acquisition. Additionally, little is known regarding the role that the brain plays in the early discrimination of these novel and familiar events. The next section of this paper focuses on one aspect of this process, infant short and long term memory. A series of papers both with human and non-human primates demonstrate that long term memory is present and measurable from an early age (Gunderson & Sackett, 1984; Gunderson & Swartz, 1985; 1986). In one such study, Gunderson and Sackett (1984) examined the development of pattern recognition in 31 infant pigtailed macaques using the familiarization-novelty technique. Following a familiarization period with 2 identical black and white patterns, the infants were tested on the familiar and novel patterns. A novelty preference clearly emerged with increasing age. Younger infants (mean age 178 days postconception) did not show a reliable visual preference for either the novel or the familiar patterns while older infants (by approximately 25 days) attended longer to novel than familiar patterns. However, by 200 days postconception, infant macaques could remember some aspects of previously exposed stimuli and consistently preferred a novel stimulus. Cowan, Suomi, and Morse (1982) noted that such memory also occurs for speech sounds. Using a modification of an adult masking paradigm and a non-nutritive sucking discrimination procedure, Cowan et al. investigated preperceptual auditory storage in 8 and 9 wk old infants. Fifty-four infants and 10 adults listened to repeating pairs of brief vowels with a stimulus onset asynchrony (SOA) of 50 msec. Within each series, either the first vowel in a pair changed (backward masking), the second vowel changed (forward masking), or neither vowel changed (control). Discrimination of the change occurred only in the forwardmasking condition. In Experiment 3, with 30 infants and 10 adults, discrimination occurred in a backward-masking condition with an SOA of 400 msec, but not with an SOA of 250 msec or in a control condition. Cowan et al. interpreted their results as suggesting that echoic storage contributes to auditory perception in infancy, as in adulthood, but that the useful lifetime of an echoic trace may be longer in infancy. In an early study of more long term memory using human infants, Sullivan, Rovee-Collier, and Tynes (1979) provided some of the earliest indications of long term memory in young human infants 3-months of age, who were trained in a conjugate reinforcement paradigm in which
344
Molfese and Narter
footkicks produced conjugate activation of an overhead crib mobile. Following two training sessions, retention as measured by savings during cued recall was assessed cross-sectionally in a third session scheduled after intervals that varied from 2 to 14 days. No evidence of forgetting was observed for up to 8 days following original training, suggesting that these memories persisted for at least this long. These types of studies provide clear indications of different types of short and long term memory in infants early in development. Another area of early infant memory research that has begun to receive more extensive treatment in recent years concerns the neuropsychology of infant memory. There exists an extensive set of electrophysiological studies investigating memory functions in young infants. Most often, these studies utilize a paradigm commonly used with adults - the "odd-ball" technique. In this procedure, two different visual or auditory stimuli are presented for differing numbers of times. For example, a picture of one face is randomly presented on 80% of the trials while another face occurs on 20% of the trials (i.e., the "odd-ball" trials). Typically, in an adult study, a positive peak with a latency of approximately 300 ms is noted to occur when adults attend to the infrequent stimulus while no such peak or a greatly reduced peak occurs for the frequent stimulus. Studies which range from involving infants as young as 4-weeks of age (Karrer & Monti, 1995) to those testing 12-month-olds (Nelson & Karrer, 1992; Nelson & deRegnier, 1992) have noted ERP differences in response to such frequent versus infrequent occurrences. For the most part, these memory effects appear to produce changes in specific peak portions of the ERP that range from those occurring prior to 500 ms following stimulus onset such as the Nc (Hoffman, Salapatek, & Kuskowski, 1981; Karrer & Monti, 1995) to the late positive components (LPC) that occur as late as 1700 ms (Nelson & Karrer, 1992; Nelson & deRegnier, 1992). Usually midline frontal or occipital ERP effects occur in response to these frequent and infrequent events. However, it is important to note that these it is only these response sites which are usually selected for study across virtually all infant memory studies. Given the widespread use of midline electrode sites, it is not surprising that relatively few studies which have investigated hemisphere effects (DeHaan & Nelson, 1997; Molfese, 1989; Molfese & Wetzel, 1992). A summary of the infant memory studies which have utilized the ERP procedure is presented in Table 2.
Cognitive Development
345
Table 2. Studies Using Event-Related Potential (ERP) Procedures To Study Infant Memory. Study/Ss Karrer & Monti (1995)
Electrode Fz, Cz, Pz, Oz, C3, C4
Task Visual ERPs oddball task
4-7-week old infants, n=20
Reference= linked ears
80% and 20% probabilities Stimuli: highcontrast contours contained in a black and white checkerboard or in a set of 4 solid geometric shapes of different colors.
Hoffman, Salapatek, & Kuskowski, (1981)
Oz, Opz Reference= left mastoid.
3-month old infants, n= 13
Courchesne, Ganz, & Norcia
(I~I)
4-7-month old infants, n= 10
Fz, Pz Reference= fight mastoid
Results Latency of a frontally predominant Nc (5001000 ms) and magnitude of an NSW (100-500 ms)changed as a function of stimulus experience. Nc latencies faster and NSW magnitude larger to oddball (infrequen0 stimulus than to frequent stimulus. Latency of N378 component over occipital scalp faster to oddball stimulus.
Visual ERPs: Familiarization: Pre-exposure for 40 trails to 80% stimulus. Testing: 80% of trials a 500 ms a vertical squarewave grating with one spectral frequency while on 20% viewed stimulus with different spectral frequency.
Late positive component (300-600 ms) only to novel (20%) at Oz and Opz.
Visual ERPs: One female face on 88% trials, second face on 12% of trials
Negative Nc (latency = 700 ms) amplitudes larger, latencies longer for infrequent faces at both Fz and Pz
Also occurred when used vertical vs. horizontal gratings matched in spectral frequency.
Effects maximal at Fz
346
Molfese and Narter
Table 2: cont'd Study/Ss Hoffman, Salapatek, & Kuskowski, (1981)
Electrode Oz, Cz, Pz
Reference= left mastoid.
3-month old infants, n= 16
Karrer & Ackles (1988)
Fz, Cz, Pz, Oz, C3, C4.
6-week, 6-, 12-, and 18-month old infants
Reference= linked ears.
Nelson & Collins (1992) 4(n=14) & 8 (n= 17) month old infants
Task Visual ERPs: Familiarization: Pure tone paired with vertical squarewave grating. Testing: on 20% of trials changes made in either: 1) auditory 2) visual 3) auditory + visual.
Results Visual change: Late positive component (300-600 ms) only to novel at Oz, Cz and Pz. Auditory- visual change: Late positive component (300-600 ms) only to novel (20%) at Oz and Pz. Auditory change: No effect.
Visual ERPs
Large negative slow wave (NSW) complex and Nc negative component (latency=770 - 800 ms)
80% and 20% probabilities; 6 week: random shapes and checketl~atds, 6 month: 2 female faces; 12-, 18-month: stuffed animals and furniture.
Fz, Cz, Pz, Oz. Reference = linked ears.
Visual ERPs Experiment had 20 familiarization trials (10 each) with 2 alternating faces for 500 ms. Remaining design identical to Nelson & Collins (1991a).
Larger at Cz to infrequent events for only 6 month olds,
Familiarization trials: No differences between familiar and novel. 4 month olds: No effects. 8 month olds: Responses similar to Frequent-Familiar and Infrequent-Familiar Infrequem-Novd produced sustained negative slow wave after 400 ms.
Cognitive Development
347
Table 2: cont'd Study/Ss Nelson, Ellis, Collins, & Lang, (1990)
6-month old infants, n= 11
Thomas & Lykins (1995) 5-month-old infants. Expt. 1:n=24 Expt. 2:n=24
Electrode Fz, Cz, Pz, Oz.
Reference = linked ears.
Electro-cap International. Cz, Fz, T3. Reference= linked ears.
Nikkel & Karrer (1994)
Fz, Cz, Pz, Oz, C3, C4
6-month old infants, n=28
Reference= linkedears
Task Visual ERPs
On 80% trials a 100 ms presentation of doll face; On 20% of trials no face presented.
Audi tory ERPs. Tones (100ms, 400 Hz), dicks (5 ms burst) Expt 1: Day 1 Tones and dicks presented for 100 trials. On Day 2, 50 old stimuli presented along with 50 new stimuli. Expt 2:Two tones presented as in Experiment 1. Visual ERPs 80 trials of two female face pictures, presentedat 80%/20% probabilities, was divided into 3 blocks.
Results Differences between 80% and 20% trials at 100-150 ms, 600-700 ms, and a sustained positive slow wave (1100-1700 ms) to the next presented familiar (doll face) stimulus after deleted face trial. Effect at Fz (maximal) and Cz.
Experiment 1: Familiar stimuli produced larger N350 (N2) on Day 2. Experiment 2: N350 replicates Experiment 1. P200 larger in amplitude and faster for familiar stimuli on Day 2.
Nc amplitude decreased across blocks at all central and anterior scalp sites but significantly decreased only at Cz. Pb increased in amplitude at all sites across blocks but was significant only at C4. No changes in NSW, Pz, or Oz at any sight.
348
Molfese and Narter
Table 2: cont'd Study/Ss Karrer, Wojtascek, & Davis (1995)
Electrode Fz, Cz, Pz, Oz, C3, C4
6-month old infants,
Reference= linked ears
23 with Down syndrome
Task Modified combination of an oddball task and a habituation-novelty task using faces; 80 trials 80% with frequent stimulus and 20% for oddball.
18 without Down syndrome
Nelson & Collins (1991)
Fz, Cz, Pz, Oz.
6-month old infants, n - 12
Reference = linked ears.
Visual ERPs: 20 familiarization trials (10 each) with two alternating faces for 500 ms. Testing involved 3 conditions: Frequent-Familiar 1 face 60% InfrequentFamiliar: 2nd face on 20% of trials Infrequent-Novel: one of 12 faces seen on 20% of trials.
Results Same ERP morphology for both groups. Chronometry of information processing by infant with Down syndrome similar to or faster than that of infants without Down syndrome, depending on component. Amplitude differences between groups may implicate frontal attention processes in Down syndrome as opposed to more posterior processes. Infants with Down syndrome had an amplitude decrement in Nc over the central but not frontal cortex. Familiarization trials: No differences between familiar and novel. Infrequent-Familiar produced a LPC after 400 ms at Cz. Infrequent-Novel produced sustained negative slow wave after 400 ms.
Cognitive Development
349
Table 2: cont'd
Study/Ss Nelson & Salapatek (1986)
Electrode
Task
Results
Fz, Cz, Oz.
Visual ERPs
Reference=
Experiments 1 and 2 had 40 familiarization trials with one face for lOOms.
Experiment 1: Negative component at Cz between 550-700 ms discriminated between novel event and stimulus presented during initial familiarization;
right 6-month old infants
mastoid.
Experiment 1 n=16
Experiment 1" Familiar face on 80% of trials.
Experiment 2 n=15
Experiment 2: Familiar face on 50% of trials.
Experiment 3 n=16
Experiment 3" No familiarization, novel and familiar faces presented randomly with equal frequency.
LPC between 8501000 ms did discriminate between novel and familiar during test at Cz and Fz. Experiment 2: Negative component at Cz between 550-700 ms discriminated between novel event and stimulus presented during initial familiarization. Experiment 3: No effects.
Molfese (1989) 14-month old infants, n= 10
F L , FR T3, T4 P L , PR Referencelinked ears.
Auditory ERPs Experiment 2: Familiarization: 6 exposures for 15 min. each over 2 days. Test: 60 presentations of the novel and 60 of the familiar syllable
Large positive component at 360 ms over FL and FR for only the familiar stimulus while a large negativity occurred for the novel stimulus.
350
Molfese and Narter
Table 2: cont'd Study/Ss De Haan& Nelson (1997) 6-month old infants Experiment 1 n=22
Electrode Fz, Cz, Pz, Oz. T3, T4, TS, T6 Reference = linked ears.
Experiment 2 n=22
Task Visual ERPs
Results
Color digitized images of mother and stranger's faces which were similar or dissimilar to mother's face. Experiment 1: Infant saw own motherand stranger
Experiment 3 n=22
Experiment 2: Infant saw another mother and stranger
Experiment 4 n=22
Experiment 3: Infant saw own motherand stranger who looked similar to mother. Experiment 4: Infant saw another motherand stranger who looked similar to mother. Faces presented with equal probability. Nelson & Karrer (1992) 12- month old infants, n=24
Oz, Pz, Cz, Fz
Visual ERPs
Reference= linked ears
~ures identical to Nelson & Collins ( 1991 )
Infrequem-Familiar proOuceOa LPC between 750-1250 and between 1250-1700 ms at Fz
Cognitive Development
351
Table 2: cont'd Study/Ss Nelson, Henschel, & Collins (1993). 8-month old infants, n=58
Electrode Fz, Cz, Pz, Oz Reference = linked ears.
Task Visual ERPs following Haptic familiarization. Familiarization: 60 Sec. 3-D object placedin hand. Condition 1: 20 500 ms trials of color slide of object Then 30 slide trials of familiar object and 30 of novel. Condition 2: 20 500-ms trials of color slide of novel object; Then 30 slide trials of familiar object and 30 of novel. Condition 3: 2 10-see trials of novel and familiar object Then 30 slide trials of familiar object and 30 of novel. Condition 4: 30 slide trials of familiar object and 30 of novel.
Results Condition 1: LPC to novel stimulus. Condition 2: No effects. Condition 3: No ERP effects. I_xmking times greater to novel than to familiar stimuli. Condition 4: LPC to novel stimulus.
352
Molfese and Narter
Table 2: cont'd Study/Ss
Electrode
Task
Results
Nelson & deRegnier (1992)
Fz, Cz, Pz, Oz.
Visual ERPs.
Infrequent-Familiar produced a LPC between 750-1250 and between 1250-1700 ms at Fz.
12-month old infants, n=24 Molfese & Wetzel (1992) 14-month old infants, n=9
Reference= linked ears. F L , FR T3, T4 P L , PR Reference = linked ears.
Procgdures identical to Nelson & Collins ( 1991). Auditory ERPs: Familiarization: 6 exposures for 15 min. each over 2 days. Test 1 on day 3: 60 presentations of the novel and 60 of the familiar syllable Retest - day 10: All infants retested 1 wk later. 60 presentations of the novel and 60 of the familiar syllable
Test 1: Large positive component at 370 ms over FL and FR for only the familiar stimulus while a large negativity occurred for the novel stimulus. Retest: Large negative peak at 280 ms over F L and FR for only the familiar stimulus while a large negativity occurred for the novel stimulus. Large negative peak at 550 ms at T3 and PL sites for familiar stimuli.
LPC = Late Positive Component (Wave); NSW = Negative Slow Wave (Component) In a study conducted with the youngest group of infants to study memory, Karrer and Monti (1995) tested a group of 20 infants, four to seven weeks of age. They placed electrodes over midline positions from front to back (Fz - the midline frontal position, Cz - the midline central position, P z - the midline posterior parietal site, and Oz - the midline occipital position, see Jasper, 1958) as well as over left and right hemisphere central positions (approximately midway between the top, center of the head at Cz and the left and fight ear meatus at C3 and C4, respectively. All electrodes were referred to linked ears. The stimuli consisted of slides of high-contrast contours contained in a black and white checkerboard or in a set of four solid geometric shapes of
Cognitive Development
353
different colors (i.e., circle, cloverleaf, triangle, inverted T). Karrer and Monti used a modified oddball task in which stimuli were presented in an 80-trial sequence with an occurrence probability of 80% for the frequent stimulus and 20% for the oddball. Two second epochs beginning with a baseline period 390 ms before stimulus onset were collected. No differences in ERPs were noted based on which specific stimulus was frequent or not, so the data were combined in all subsequent analyses across stimuli. Karrer and Monti noted that the Nc negative area between 500-1000 ms at Fz was larger than at Cz. There was a significant effect of stimulus probability (frequency) on Nc latency with a later Nc peak for frequent than for oddball stimuli. Nc latency at C3 and C4 was significantly later for frequent than for oddball stimuli. The NSW area, a negative area between 100 and 500 ms preceding Nc, at Fz was significantly larger than at Cz. NSW had significantly larger areas at C3 and C4 for novel trials than for familiar trials. Finally, the latency of N378 at Oz was significantly faster for oddball trials than for frequent trials. Karrer and Monti concluded on the basis of these findings that Nc latency and the magnitude of the NSW demonstrated that these two c o m p o n e n t s are functionally independent and most predominant over anterior and right central scalp areas overlying brain regions known to be associated with attentional processes. They felt that the latencies of these ERP components permitted them to gain some insight into the infant's ability to process information. Infants from 4 to 7 weeks of age require about 40 ms (N378 latency differences) to perceive the physical difference between stimuli, 800 ms (Nc latency) to search memory to encode simple visual stimuli, and about 80 ms (Nc latency differences) to search memory to discriminate between them. In a related study with somewhat older infants, Hoffman, Salapatek, and Kuskowski (1981) noted that when an infrequent or unexpected stimulus is presented to an adult, a characteristic enhancement of the late positive component (LPC) of the averaged evoked cortical potential is observed. To test whether this effect occurs during early infancy, they presented low and high probability visual stimuli to 29 infants 3-months of age in a series of two experiments, Initially, infants viewed a single visual stimulus for 40 trials. Next, during testing, the pre-exposed stimulus was presented for 500 ms each time on 80% of the trials. Visual ERPs recorded from posterior parietal (Pz), and occipital (Oz) scalp sites contained a clear LPC effect between 300 and 600 ms following the
354
Molfese and Narter
onset of the infrequent stimulus that was presented on 20% of the trials. Hofmann et al. interpreted these findings as a demonstration that the LPC effect in infants reflects cognitive processing involving memory. A similar visual ERP effect in terms of polarity and latency was noted with another group of 3 month olds by Hoffman and Salapatek (1981) at Cz, Pz, and Oz, as well as by Nelson and Collins (1991) at the Cz site with 6 month olds. In a related study, Nelson and Salapatek (1986) noted that visual ERPs recorded over a midline electrode position (Cz) to a novel face were significantly more positive than responses to a familiar face during the interval between 551 and 700 msec following stimulus onset. Negative peak effects have also been found to index differences in familiar and novel events. Karrer and Ackles (1987) recorded visual ERPs from 6-month olds in response to two different female faces that were presented on 80% and 20% of the trails. They noted that what they labeled as the Nc component was larger at Cz to the infrequently presented face. Nelson and Salapatek (1986) also noted a negative component at Cz between 500 and 700 ms that discriminated between a novel face and one that had earlier been presented during the initial familiarization trials. Nelson and Collins (1991), in a study with 6 month olds, reported a sustained negative slow wave after 400 ms that discriminated between an infrequently presented face and a familiar face as did Nelson and Collins (1992) 8 month olds. While most of these ERP studies have focused on infants at 6 months of age or younger, there are a growing number of studies conducted with older infants between 8 and 14 months of age as illustrated in Table 2 that indicate such effects continue and even elaborate further during later infancy (Nelson & Collins, 1992; Nelson, Henschel, & Collins, 1991; Nelson & deRegnier, 1992; Molfese, 1989; Molfese & Wetzel, 1992). Only two studies investigating infant memory have been conducted using auditory stimuli (Molfese, 1989; Molfese & Wetzel, 1992). Interestingly, these were the only two studies to also include more than a single pair of lateralized electrode placements which permitted some examination of hemisphere related differences. While scientists have speculated that the left hemisphere plays a major role in early sound discrimination (Molfese, Freeman, & Palermo, 1975; Molfese & Molfese, 1979b; 1980; 1985; Molfese & Betz, 1988), few studies directly assessed early memory for these speech sounds in infants which might indicate the role that different areas of the brain play in the recognition of familiar versus novel speech sounds at this stage of
Cognitive Development 355 development. Molfese (1989) recorded auditory ERPs from frontal, temporal, and parietal scalp locations over the left and right hemispheres of 10 infants, 14 months in age, who listened to a series of repeated consonant-vowel-consonant-vowel (CVCV) syllables over a 2 day period prior to testing. On the third day, when ERPs were recorded to the familiar CVCV and to a novel one, differences in the ERPs were noted only over the left and right frontal electrode sites. These effects were most marked at approximately 360 ms following stimulus onset and were characterized by a large positive peak at this latency over both the left and right frontal regions for the familiar CVCVs but not for the novel CVCVs. Molfese and Wetzel (1992), in a follow-up to this study, replicated these effects and also noted that after one week, a retest showed in the ERPs both a large bilaterally distributed frontal negative peak (N 1) 280 ms after onset of the novel stimulus and a subsequent larger bilaterally distributed frontal positive peak (P2) for the familiar stimulus as well as larger left hemisphere lateralized temporal and parietal responses at approximately 550 ms for familiar stimuli. Thus, the memory effects noted in much younger infants appear to become more complex in that they are reflected in both lateralized and bilateral responses with varying peak latencies after some period of time. In a related study using visual stimuli, Thomas and Lykins (1995) conducted two experiments using ERPs to investigate 24-hour recognition memory in infants. ERPs were recorded to 100 identical stimuli in 5-month old infants. After 24 hours, 50 of these familiar stimuli and 50 novel stimuli were then presented. The amplitude of a large negative peak (N2; approximate latency = 350 msec) of the auditory ERP was larger on Day 2 for the familiar stimuli compared with ERPs for both Day 1familiar stimuli and Day 2-novel stimuli. Trial-to-trial latency variability of N2 decreased from Day 1 to Day 2 for the familiar stimuli. A second experiment replicated the results of Experiment 1 and also noted an additional effect, an earlier positive peak (P2; approximate latency = 200 msec) characterized by larger amplitudes and smaller latency variability to the familiar stimulus on Day 2. All of these infant memory studies have produced results supporting the suggestion that infants respond differently to two stimulus sets if one is made familiar by first exposing the infant to this stimulus set prior to the novel-familiar comparison test or else indicating that infants can discriminate between frequent and infrequent events. Consequently, it
356
Molfese and Narter
appears that these infants are able to respond to events or stimuli that are through experience differentially encoded in memory.
Electrophysiologicai correlates of Infant word acquisition While our knowledge of infants' speech perception has expanded rapidly over the past decade (Eimas, et al., 1971; Kuhl, 1985; Morse, 1974; Molfese & Molfese, 1979b; 1980; 1985), little remains known about the infant's beginning comprehension of "names" for objects/ events (Bates, 1979). While some investigations have documented and catalogued the words first comprehended by infants, beginning around 8 months of age (Benedict, 1975; Kamhi, 1986; Miller, & Chapman, 1981; Macnamara, 1982), only recently have investigations probed the nature of the older infant's early word meanings (Bloom, Lahey, Hood, Lifter, & Fiess, 1980; Clark, 1983; Snyder, Bates, & Bretherton, 1981; Retherford, Schwartz, & Chapman, 1981) and to study the very beginning stages of the infant's ability to perceive and remember the names for objects and events (Bates, Benigni, Bretherton, Camaioni, & Volterra, 1979; Bates,Bretherton, Snyder, Shore, & Volterra, 1980; Golinkoff, Hirsh-Pasek, Cauley, & Gordon, 1987; Hirsh-Pasek & Golinkoff, 1996; Kamhi, 1986). Moreover, virtually nothing is known about the role that the brain plays in the early acquisition of such word meanings (Molfese, 1989, 1990; Molfese, Morse, & Peters, 1990). Furthermore, while scientists have speculated that the left hemisphere plays a major role in early language acquisition (Best, 1988; Lenneberg, 1967), little actual work has been conducted to address this issue. Indeed, five recent papers indicate that such procedures can be successfully used to study the developmental neuropsychology of early word comprehension in infants from 12- to 20-months of age (Mills, Coffey-Corina, & Neville, 1993, Molfese, 1989, 1990; Molfese, Morse, & Peters, 1990; Molfese, Wetzel, & Gill, 1993). The youngest group of infants to be studied using ERPs was a population of 12-month-old infants (Molfese, Wetzel, & Gill, 1993). This study represented a direct attempt to determine whether ERPs recorded from 12-month-old infants could discriminate between words thought by infants' parents to be known to these young infants from those words that parents strongly believed were not known to the infant. It was hoped that multiple electrode sites over various areas of each hemisphere would provide more information concerning the involve-
Cognitive Development
357
ment of different brain regions in early word discrimination. Because details of this study are relevant to subsequent studies to be reviewed here, this review of Molfese et al. (1993) will be more extensive. A group of nine infants, three females and six males, (mean age = 12.2, s.d. = .36) were tested. Initial screening indicated that each of the parents were strongly fight handed as indicated by their responses to the Edinburgh Handedness Inventory (Oldfield, 1971) which yielded a mean Laterality Quotient greater than +0.7. Unique stimulus tapes were constructed for each infant, based upon the parental ratings obtained during a telephone interview during which parents were asked to identify all of the words from the original list of ten (i.e., "bottle," "book," "cookie," "key," "kitty," "ball," "dog," "baby," "duck," and "cat.") which they believed that their infant understood. Next, they were asked to rate their confidence in their identification using a five-point scale. Parents were told that a rating of "5" indicated that they were "very confident" that the infant did or did not know the word, while a rating of "1" signified that the parents were "not confident at all" about their decision. Following the interview, parent ratings were converted to a range from "1" to "10", with "1" signifying high confidence that the infant did not know the word, and "10" signifying high confidence that the infant knew the word. The stimuli which were used as the "known" words in the present study had a mean rating of 9.7 out of 10.0 (s.d. = .4). For the "unknown" words, there was a mean rating of 1.9 (s.d. = .5). Each tape contained stimulus repetitions of two spoken words produced by an adult male speaker using flat intonation. Each word began with a voiced, stop consonant to minimize E R P variations due to acoustic factors such as voicing or rise time. One of the two words was identified by that infant's parent as known to the infant while a second word was believed by the parent to be unknown to the infant. The known and unknown words were arranged on the tape in a block random order, with 54 occurrences of each and a randomly varied interstimulus interval. Six silver cup scalp electrodes were placed over the left and right sides of each infant's head. These placements included two electrodes placed respectively over the left (T3) and fight (T4) temporal areas of the Ten-Twenty System (Jasper, 1958); a third electrode placed a t FL, a point midway between the external meatus of the left ear and Fz; a fourth electrode placed at FR, a position midway between the right external meatus and Fz; a fifth electrode placed at PL, a point midway
358
Molfese and Narter
between the left external meatas and Pz; and a sixth electrode placed at PR, a point on the fight side of the head midway between the right ear's external meatus and Pz. These electrode placements were positioned on the scalp over the left frontal (FL), temporal (T3), and parietal (PL) areas of the brain and the corresponding areas of the fight hemisphere (FR, T4, and PR, respectively). Such placements, it was hoped, would provide information concerning not only left versus right hemisphere responses to the known and unknown words, but in addition, information within each hemisphere concerning general language perception areas commonly thought to be localized to the left temporal and parietal language receptive regions of the brain as well as the language production areas of the frontal lobe. The electrical activity recorded from these scalp electrodes was obtained in response to a randomly arranged series of words matched in duration and peak intensity levels. As in previous studies by Molfese and his collaborators, scalp electrodes were referred to electrodes placed on each earlobe and linked together (A1, A2). The words were presented auditorily through a speaker positioned approximately one meter over the midline of the infant's head at 80 dB SPL (A) as measured at the infant's ears and occurred while the infant was in a quiet awake state. Continuous monitoring of the infant's ongoing EEG and EMG, as well as behavioral observation, were used to determine when stimulus presentation should occur. During periods of motor activity, stimulus presentation was suspended and the infant was shown various toys and pictures until quieting. Testing was resumed when the infant's motor activity declined to an acceptable level. Data reduction and analysis procedures first involved digitizing 70 data points over a 700 ms period beginning at stimulus onset for each electrode site, stimulus event, and infant. Next, the ERPs were subjected to artifact rejection for each electrode to eliminate from further analyses those contaminated by motor or eye movements. This resulted in rejecting less than 10% of the trials for each infant. Rejection rates were comparable across the two stimulus conditions. Following artifact rejection, the single trial data were then averaged separately for each electrode site and stimulus condition. Thus, 12 averages were obtained for each infant,. These included averages for the known and unknown words for each of the six scalp electrode sites. As in the case of the speech perception studies, the average ERPs then were submitted to a two step analysis procedure (Brown, Marsh, & Smith, 1979; Chapman,
Cognitive Development
359
McCrary, Bragdon, & Chapman, 1979; Donchin, Teuting, Ritter, Kutas, & Heffley, 1975; Gelfer, 1987; Molfese, 1978a, 1978b; Molfese & Molfese, 1979b, 1980, 1985; Ruchkin, Sutton, Munson, Silver, & Macar, 1981; Segalowitz & Cohen, 1989). This procedure first involved the use of a Principal Components Analysis (PCA) and then an Analysis of Variance. Factor scores or weights were generated by the PCA for each of the 108 averaged ERPs for each of the six rotated factors accounting for 77.84% of the total variance. The ANOVAs were based on the design of Subjects (9) X Word Understanding (2) X Electrode Sites Within Hemispheres (3)X Hemispheres (2). These were conducted to determine if any of the regions of the ERPs identified by the six factors varied systematically as a function of the specific levels of the independent variables in this study. Two ERP regions were noted to vary as a function of whether parents believed that the infant could recognize the meaning of a word or not. The first region which reflected variations in the initial portion of the waveform until approximately 140 ms following stimulus onset indicated that ERP activity recorded from over all left hemisphere sites discriminated the known from the unknown words. However, only electrical activity from the frontal and temporal regions of the right hemisphere made similar discriminations. The ERPs elicited by the known words were characterized by a larger negative (or downward) peak prior to 140 ms while a markedly smaller negative peak characterized the ERPs evoked by the unknown words. Thus, the vertical amplitude appears larger for the known words than the unknown words. A second region of the averaged ERPs between 210 and 300 ms post stimulus onset also varied systematically as a function of whether the word was thought to be understood by the infant. This effect was reflected by amplitude differences in the second major negative component of the ERP where the overall negative peak-to-peak amplitude in the region between 210 and 300 ms was generally larger for ERPs elicited by the known than by the unknown words. These results were interpreted by Molfese and colleagues to indicate that auditory evoked ERPs successfully discriminated between words that parents believed their 12-month-old infants knew from those that the infants were thought not to understand. Moreover, Molfese et al. noted that even at this young age the process of word comprehension appeared to be dynamic in that different regions of the brain responded differently over time following the onset of the word that was known to
360
Molfese and Narter
the infant. Initially, differential electrical activity was generated early in the waveform for approximately 200 ms over the entire left hemisphere and most of the right hemisphere except the right parietal area Such early responses suggest that the infant may begin to process words as meaningful from virtually the time they begin to hear the auditory signal. This initial response period then was followed for a 100 ms period by a spreading of the discrimination to all ERP scalp regions. Thus, the differential response of the ERPs to words appears to continue through much of the time that the infant hears the word, although it appears to be carried out by different brain regions. In Experiment 1 of a study with older infants, Molfese (1989) recorded auditory evoked responses from frontal, temporal, and parietal scalp locations over the left and right hemispheres of 10 infants, 14 months in age, who also listened to a series of words, half of which were determined to be known to the infants (based on behavioral testing and parental report) and half of which were believed not to be known to the infant. A behavioral test was used to confirm the parents' ratings of their infant's word comprehension. As in the case of the 12-month infant study (Molfese, et al., 1993), parents rated words from a list as either words the infant knew or words they did not know. In addition, however, each infant received four behavioral trials, with two independent observers rating whether or not the infant knew the word presented. In order to assess the infant's comprehension, a specially constructed cabinet was used. The cabinet was 1 meter in height and contained four shelves, each .4 meters in length, with two shelves to the left and two shelves to the fight of midline. The object representing the known or the unknown word (as appropriate) was placed in one of the four compartments of a test cabinet. Two compartments of the test box each contained distracter items randomly selected for each trial from a sack of toys while the fourth compartment remained empty. The parent then instructed the infant to look at or retrieve various toys using instructions to the infant such as, "Go get the book." or "Look at the duck." The compartments that contained the test object, the empty space, and the distracters were randomized for each trial for each infant based on a randomly generated list derived by computer prior to the test session. On each trial the raters independently determined whether they believed that the infant responded to the instructions correctly and recorded their confidence in these judgments on a 5-point scale identical to that previously used by the parents. For the children in this study, both the
Cognitive Development
361
parents (across the two interviews) and the raters reliably rated the words they believed were known to the infant as different from those that they believed the infant did not understand. Analyses of the ERP data isolated three regions of the evoked potential waveform that discriminated known from unknown words in this population. Initially, ERP activity across both hemispheres (with the exception of the fight parietal region) between 30 and 220 ms following stimulus onset discriminated known from unknown words. This effect appeared as a positive peak for the known words and a negative peak in this same region for the unknown words. This activity was followed shortly by a large positive to negative shift in the waveform between 270 and 380 ms across all electrode sites for both the left and right hemispheres that was larger for the known than for the unknown words. Finally, a late negative peak between 380 and 500 ms that was detected only by electrodes placed over the left and right parietal regions was larger for the known than for the unknown words. In Experiment 2 of this study, Molfese attempted to determine whether familiarity with speech stimuli produced brain responses similar to those found for the known vs. unknown word materials. In this procedure, a different set of ten 14-month-old infants first listened to a nonsense bisyllable (CVCV)over a 2 day period. Parents encouraged their infant to play with a box on which a large orange Frisbee was mounted and connected to a series of switches. Infants played with the device for three times on each of the 2 days designated for training, with 15 minutes allowed for each of the six play sessions. Five infants heard "toto" during the familiarization process while five children heard "gigi" to decrease the likelihood that any experimental effects were due to acoustic differences between the stimuli instead of due to differences in amount of previous exposure to the different stimuli. On the third day, ERPs were recorded to this now familiar CVCV and to the novel CVCV. Electrode placements were identical to those used in Experiment 1. If the latencies and scalp distributions of the brain responses found in this study were identical to those found in the known - unknown word study, it was felt that the familiarity hypothesis could not be rejected. In fact, however, results indicated that only the brain responses recorded over both the left- and right-hemisphere frontal areas discriminated between the familiar and nonfamiliar CVCVs. In addition, the major peak in the ERP that discriminated these differences occurred at 360 ms, not at the 630 ms previously found for the known - unknown word distinction.
362
Molfese and Narter
Consequently, Molfese concluded that the earlier ERP findings discussed in the first experiment did indeed reflect meaning differences and not differences in familiarity. A study with 16-month-old infants (Molfese, 1990) which utilized procedures similar to those employed by Molfese, Wetzel, and Gill (1993) and Expt. 1 of Molfese (1989), found comparable differences in response to known and unknown words. Molfese tested 18 infants (9 females, 9 males with a mean age of 16.57 mo. (s.d.=.6, range=15.417.5). As in his previous studies, parent handedness was measured and found to indicate strong right hand preferences across parents. In addition, parents were asked to rate a set of 10 words during a telephone interview and a subsequent lab visit in order to identify at least one word that parents confidently believed their infant knew and another word they believed the infant did not know. As in the case of Experiment 1 of Molfese (1989), two independent raters evaluated infants' word knowledge using the four shelf cabinet. Only infants were tested whose parents and raters agreed on the same set of known and unknown words, and who displayed high confidence that the infant did or did not know specific words. Following these rating procedures, electrodes were applied to six electrode sites, three over each hemisphere using the same scalp locations and references used previously by Molfese in this series of studies. ERP testing then commenced with 54 repetitions of each of the two words presented auditorily in random order, separated from each other by a varied ISI random (2.5-4 Sec.). A principal components analysis of the averaged ERPs yielded five factors (scree) accounting for 74.33% of the total variance. Two factors identified ERP regions which varied as a function of the KNOWN vs. UNKNOWN distinction. The first region, between 180 and 340 ms with a peak latency of 270 ms contained a larger N180 - P340 complex for UNKNOWN words at only the T3 site for females while the left and fight frontal regions as well as T3 showed a similar effect for males. A second region, between 580 and 700, with a peak latency of 650 ms, also discriminated KNOWN vs. UNKNOWN words for females but this time at all LH sites while for males, both the LH and RH sites discriminated KNOWN vs. UNKNOWN words. In all cases, this discrimination was reflected by larger negative shifts for UNKNOWN words. A fourth study by Molfese, Morse, and Peters (1990) also investigated aspects of the infant's early word comprehension, but this time in a training situation. Fourteen infants, seven females and seven
Cognitive Development
363
males, at 14 month of age (mean = 14.72 months, s.d. = .63) participated in a training study in which specific CVCV nonsense syllables were systematically paired with specific objects of specific shapes and colors over a five day training period. Handedness questionnaires administered to all parents (Edinburgh Inventory, Oldfield, 1971) indicated that both parents of each infant were strongly fight handed (group mean Laterality Quotient = .67, s.d. = . 19). On the day before the training period, infants first were tested in a Match Mismatch task in which on half of the trials each object was paired with its CVCV label (i.e., the label that the infant would later learn during the training period was the "name" of that object) while on the other half of the trials the objects were mispaired with the CVCV "names" of other objects. The parents and infants then returned home for 5 days of training. On the sixth day and immediately prior to the post-training session, parent were asked to indicate whether or not their infants knew the name of the object in question and then to rate their own confidence in that judgment using a 5 point scale with "1" as "completely not confident" and "5" as " very confident". The confidence ratings were then used to transform parents' ratings from a "1" of "confidently unknown" to a "unknown but completely not confident" rating of "5" to a "known but completely not confident" rating of "6" to a "confidently known" rating of "10". Using this rating system all parent rated the terms as "known" by the infant with a mean confidence rating for "bidu" at 8.71 (s.d. = .88) and for "gibu at 8.79 (s.d. = .94), indicating that the parents as a group were confident that their infants understood which terms labeled which objects. Next, during the post-training test, infants again were tested in a Match - Mismatch task in which on half of the trials each object was paired with its CVCV label (Match condition) while on the other half of the trials the objects were mispaired with the CVCV "names" of other objects (Mismatch condition). The electrophysiological techniques used during this phase were identical to those employed during the pretest phase of this study. Using artifact rejection and analysis procedures comparable to those employed in the earlier studies conducted by Molfese and colleagues, the ERPs were averaged separately for the pre- and post-training tests for each of the six electrode sites and each of the two stimulus conditions (match vs. mismatch). Twelve averages were obtained for each data set for each infant. Each average was based on 80 samples
364
Molfese and Narter
combining responses to "bidu" and "gibu" for the Match condition and 80 samples across "bidu" and "gibu" for the Mismatch condition. Two regions of the ERP waveform reliably reflected Match related effects during this task - an early component of the ERP which changed bilaterally over the frontal regions of both hemispheres and a late occurring lateralized response which was restricted to only the left hemisphere electrode sites. The first, which occurred between 30 and 120 ms post stimulus onset, was characterized by a marked negativity for the Mismatch condition. A second region, which began 520 ms after stimulus onset, reached its peak at 580 ms, and then diminished by 600 ms, produced a positive going wave for the Mismatch condition over all of the left hemisphere electrode sites. Since no such Match or Mismatch effects were noted in the pretraining ERP session, it is clear that the ERPs detected changes which occurred as a function of training. When a correct match occurred between the auditorily presented word and the object that the infant held, both the left and right frontal regions of the brain emitted brain responses which contained an initial positive deflection or peak between 20 and 100 ms following the auditory onset of the object name. If a mismatch occurred, however, this early positive deflection inverted 180 degrees and became a negative deflection. Later in time, between 520 and 600 ms, just before the conclusion of the ERP, a large positive going wave occurred over only the three left hemisphere electrode sites when the infant listened to a stimulus which did not name the object that the infant held. Given the short latency of the initial changes in the ERP waveshape across the frontal regions, it appears that the young infant must recognize almost immediately if there is agreement between something that it hears and something that it sees and touches. These early word acquisition studies are summarized in Table 3. Perception of Coartieulated Cues: While the early ERP response during the first 100 ms following stimulus onset might superficially appear to have occurred before the infant could process the acoustic information of the CVCVs, such early discrimination is not without precedence. Both behavioral and electrophysiological investigations have indicated that coarticulated speech cues can lead to comprehension long before the entire word or phrase has been articulated by the speaker and heard by the listener (Ali, Gallagher, Goldstein, & Daniloff, 1971; Daniloff & Moll, 1968; MacNeilage & DeClerk, 1969; Molfese,
Cognitive Development
365
Table 3. Studies Using Event-Related Potential (ERP) Procedures To Study Infant Word Acquisition. Study/Ss Molfese (1990) 14-month old infants, n= 16 (8 males, 8 females)
Electrode F L , FR T3, T4 P L , PR Reference= linked ears.
Task Auditory ERPs Experiment 1" Randomly ordered words rated as known to the infants and words rated as unknown. Equal number of known and unknown words.
Molfese, Wetzel, & Gill (1994) 12-month old infants, n= 12
Molfese (1989) 14-month old infants, n= 14
F L , FR T3, T4 P L , PR Reference= linked ears.
F L , FR T3, T4 P L , PR Reference = linked ears.
Auditory ERPs Randomly ordered words rated as known to the infants and words rated as unknown.
Results Known vs. Unknown word effects: Females: T3 Males: FL,T3, FR 180-340 ms. Females: LH frontal, temporal, parietal Males: All electrode sites. 580-700 ms. Known vs. Unknown word effects: LH frontal, temporal, parietal RH frontal, temporal 210-300 ms.
Equal number of known and unknown words.
All electrode sites. 210-300 ms.
Auditory ERPs
Known vs. Unknown word effects:
Experiment 1: Randomly ordered words rated as known to the infants and words rated as unknown. Equal number of known and unknown words.
LH frontal, temporal, parietal RH frontal, temporal 30 - 220 ms. All electrode sites 270-380 ms. LH and RH parietal 380-500 ms.
366
Molfese and Narter
Table 3: cont'd Study/Ss Mills, CoffeyCorina, & Neville (1993) 20.5-month old infants, n=24 (16 females, 8 males)
Electrode F7, F8, 33% distance from T3/4 to C3/4, 50% of distance from T3/4 to P3/4, O1, 02. Reference= linked mastoids.
Task AuditoryERPs Ten known words, unknown words, andbackward presented words All presented 6 times each for total of 180 trials.
Results N200: 90% showed larger N200 known vs. unknown words at LH temporal & parietal sites. 80% showed known > unknown words at LH frontal and RH temporal & parietal sites. Known: Temporal, parietal > frontal, occipital Unknown: RH amplitude > LH. Larger to known than backward at frontal, temporal & parietal sites. Larger to unknown than backward at RH sites. Backward ERPs more positive over LH anterior than LH posterior sites. N350 latency: Known < backward (by 20 ms). N350 amplitude: Known > unknown at LH temporal & parietal sites. Known > backward at LH sites. Unknown > backward at RH sites. Unknown larger for RH than LH sites. N600-900: For all stimuli, anterior RH>LH. No sex related effects.
Cognitive Development
367
T a b l e 3: cont'd
Study/Ss Molfese, Morse, & Peters (1990) 14-month old infants, n= 14
Electrode
Task
Results
F L , FR T3, T4 P L , PR
Auditory ERPs
Pretraining test: No ERP differences for Match vs. Mismatch
Reference= linked ears.
Pretraining test: ERPs to meaningless CVCVs. 5 days training, 2 blocks of 15 minutes per day training one CVCV to one novel object and another CVCV to another novel
Post training test: Match vs. Mismatch wordeffects: LH and RH frontal 30 - 120 ms.
LH frontal, temporal, parietal 530-600 ms
object. Post training test: Parents rate training effectiveness; ERPs to same CVCV s used in pretraining test. RH = Right hemisphere LH = Left hemisphere
1979). It is possible that infants used such information to discriminate between the Match and Mismatch conditions of the Molfese, Morse, and Peters (1990) study as well as the known versus unknown discrimination studies of Molfese (1989, 1990). Coarticulation refers to the finding that the shape of the vocal tract during the production of a speech sound will be altered by the place and manner of articulation for later speech sounds. MacNeilage and DeClerk, in one study, made cineflurograms and electromyograms of individuals producing a series of 36 CVC syllables and found that the articulation of the initial consonant sounds changed as a function of the identify of the following sounds. Ali, et al. noted a perceptual counterpart of coarticulation. They constructed series
368
Molfese and Narter
of CVC and CVVC syllables in which final sounds were either nasal [m, n] or non-nasal consonants. After the final vowel-consonant and consonant transitions were removed, the resulting CV and CVV syllables were presented to a group of adults who were able to discriminate between the nasal and non-nasal sequences at well above chance levels. Ali et al. and others have argued that such coarticulated information allows the listener to perceive and process some or all of the utterance at a more rapid pace and before it has been completely articulated. In fact, there appears to be an neuroelectrical correlate of perception for coarticulated cues. Molfese (1979) in a study with adults recorded ERPs to duration matched CVC words and nonsense syllables which differed from each other only in the final consonant sound. Adults listened to each CVC and then after a brief delay pressed one of two keys to indicate whether they had heard a word or a nonsense syllable. Three regions of the ERP, including one that peaked 60 ms following stimulus onset, changed systematically as a function of the meaningfulness of the CVC syllables. Molfese interpreted this component as well as later negative peaks at 260 and 400 ms as sensitive to the coarticulated speech cues which carried information concerning the later occurring (after 650 ms) final consonant sound. In the Molfese, Morse, and Peters (1990) study, given that the consonant burst and frequency transition information that discriminated one C V C V from the other occurred during approximately the first 50 ms of each stimulus (MacNeilage & DeClerk, 1969), it is possible that infants utilized this coarticulated perceptual information to rapidly identify and discriminate early in time between the auditory tokens that matched or failed to match the object the infant was holding throughout a block of trials. If infants can indeed process such acoustic (and consequently linguistic) information this rapidly so early in development, it would seem quite likely that we have been underestimating significantly the infant early language processing abilities. Such findings suggest that ERP studies with young infants can be used to both successfully study early word acquisition and the processes they use to acquire and recognize words. Mills, Coffey-Corina, and Neville (1993), in a more recent study of early infant word perception, recorded auditory ERPs from 24 children, 16 females and 8 males (mean age = 20.5 mo.) to a series of 10 comprehended (known), unknown, and backward presented words. A language assessment test, the Early Language Inventory (ELI), was administered one week before ERP testing. Parents rated 120 words on
Cognitive Development
369
1-4 scale of confidence that child did or did not comprehend word. Also a Comprehension Book with 50 object names, 9 verbs and modifiers was presented to the child who then pointed to each picture verbalized by experimenter. The child's language production was tested separately. These procedures allowed investigators to identify stimuli for the ERP test. For the electrophysiological part of the study, each child was seated in a parent's lap, opposite an audio speaker which was positioned behind hand puppets used to focus the child's attention. ERPs were collected to 10 known, unknown, and backward words presented 6 times each for 180 trials. An electrode cap was used with electrodes placed at F7, F8, 33% distance from T3 to C3 and from T4 to C4, 50% of distance from T3 to P3, and from T4 to P4, O1, 02, all referenced to linked mastoids. Prior to analyses the children were divided into two groups based on whether their ELI was above or below the 50th percentile. Subsequent tests then confirmed that these groups differed on productive vocabulary and comprehension. In general, Mills et al reported larger temporal and parietal responses than for frontal or occipital sites for known words, while there were larger RH responses overall to unknown words. A number of known vs. unknown word effects were noted. For the N200 region (the most negative point between 125-250 ms) 90% of the children produced a larger N200 to known versus unknown words at LH temporal and parietal sites while 80% showed a larger N200 to known versus unknown words at LH frontal and RH temporal and parietal sites. When comparisons included backward speech, larger responses were noted to known and unknown words than to backward speech, with generally more positive responses over LH anterior than LH posterior sites. At the next major peak measured (N350 - the most negative point between 275 and 450 ms), the known words elicited a faster N350 than backward speech (by 20 ms). Amplitude measures of this peak found larger responses to known than unknown words at LH temporal and parietal sites. Unknown words overall elicited larger responses over RH than LH sites. Finally, responses to known words were larger than to backward stimuli at LH sites while unknown words elicited larger responses than backward stimuli at RH sites. Measures of the negative wave in the region between 600 and 900 ms noted that all stimuli elicited larger anterior RH than LH responses.
370
Molfese and Narter
When ERP sensitivity to language abilities was compared between the two language groups, a number of additional effects were noted. Overall, the N200 amplitude was greater for Low language producers for both known and unknown words. N350 latency was fastest to known words for the High production group (by 28 ms). Additionally, High production group had larger N350 amplitudes at LH temporal and parietal sites. Other later effects were also noted within the 600-900 region (mean negative amplitude between 600-900 ms). The negative amplitude was largest for known words for the High production group at RH frontal site, and the Low production group at bilateral anterior sites. Known words also produced larger responses for the Low than High producers at LH temporal and parietal sites while the unknown words produced larger responses for the Low than for the High producers at RH parietal site. Unlike Molfese (1990), who tested 16-month-olds, they did not note any sex effects. Summary of early word acquisition studies. Based on this review, it is clear that electrophysiological measures involving the auditory event related potential can be used successfully to discriminate between ERPs elicited by words thought to be known to an infant versus words identified as unknown. As argued elsewhere, such procedures open up a number of possibilities, both for exploring further the semantic development of the young infant and for detecting developmental problems in children who are slow in acquiring their first words. There are remarkably similarities in terms of scalp electrode effects and k n o w n - unknown word discrimination effects across studies. For example, Molfese (1989, Experiment 1) noted that three regions of the ERP waveform discriminated known from unknown words. Initially, ERP activity across both hemispheres (with the exception of the right parietal region) between 30 and 220 ms following stimulus onset discriminated between known and unknown words. Thus, in two different ages of infants, 12-month-olds and 14-month-olds, similar regions of the ERP waveform distributed over the same electrode recording areas discriminated known from unknown words. A similar effect was also reported by Molfese, Morse, and Peters (1990). Furthermore, in all three studies, this activity was followed shortly by a large positive to negative change in amplitude across all electrode sites for both the left and right hemispheres that was larger for known than for unknown words or match versus mismatched labels. Molfese (1989) reported that this effect
Cognitive Development
371
occurred between 270 and 380 ms while Molfese et al. (1990) identified this area between 210 and 300 ms. These studies differ, however, in that Molfese (1989) reported a third, late negative peak between 380 and 500 ms, that was detected only by electrodes placed over the left and fight parietal regions, an effect which was larger for the known than for the unknown words. No such effect was noted by Molfese et al. The absence of such a late effect in the ERP responses of 12-month-old infants could reflect differences in the developmental stages between the younger infants tested by Molfese, Wetzel, and Gill (1993) and the older infants tested by Molfese (1989, 1990). Molfese, Wetzel, and Gill (1993), as in the case of both Molfese (1989) and Molfese, Morse, and Peters (1990), observed an effect in the initial portion of the ERP that discriminated known from unknown words. Given other behavioral and electrophysiological investigations of coarticulated speech cues as noted above (Ali, Gallagher, Goldstein, & Daniloff, 197 l, Daniloff & Moll, 1968; MacNeilage & DeClerk, 1969; Molfese, 1979), it is possible that the infants can use acoustic correlates of articulatory information in the initial portion of words to identify words. If so, this suggests that such perceptual strategies are mastered by the infant at a very early stage of the language learning process. It is interesting to note that, although a general belief exists that language perception is carried out by mechanisms within the left hemisphere (Lenneberg, 1967), none of the known versus unknown word related effects were exclusively restricted to only the left hemisphere electrode sites in infants younger than 16 months of age. Even the Mills et al. study with 20+ month-old infants indicate that a large percentage of children (80%) showed larger N200 responses to known than unknown words at both LH frontal and RH temporal and parietal sites. The N350 latency difference between known and unknown words also occurred across both hemispheres, rather than for only LH sites. These data can be used to argue that, at least in the early stages of language acquisition, both hemispheres of the brain are dynamically involved in the process of learning to relate word speech sound sequences to word meanings. If this is indeed the case, then perhaps the reason that young infants experiencing left hemisphere brain damage during the initial stages of language acquisition are able to recover language skills more quickly than those injured later may be due to the duplicated mechanisms subserving language abilities of the right hemisphere at that stage. Consequently, after the loss of the left
372
Molfese and Narter
hemisphere, the infant's language functions may continue to be served by these fight hemisphere mechanisms. If, on the other hand, the child experiences left hemisphere damage later in development, the outcome could be quite different. Either due to changes in brain plasticity or further specialization of the brain with development, the ability of the non-language specialized hemisphere in the normally developing child becomes more restricted with age. Consequently, following injury to the left hemisphere at this later stage of development, there may not be a right hemisphere that is capable of performing these functions and therefore language performance would be impaired because the right hemisphere is unable to continue this process in the absence of a fully functioning left hemisphere. Instead, the right hemisphere would only have residual abilities that reflected its involvement at a much earlier stage of development. The initial tenet of this review was to discern whether similar mechanisms might underlay different cognitive areas in infancy. A working hypothesis was that similarities in peak latency, amplitude, and scalp distribution effects across different cognitive domains would suggest that these different cognitive domains depend at some level on similar mechanisms. Alternatively, if one accepts Karrer and Monti's (1995) argument that differences in latency and response pattern must argue for different mechanisms which generate these differences, one is struck by marked differences between these different cognitive domains. Infant speech perception responses generally appear to occur at approximately 200, 530, and 900 ms post stimulus onset for VOT and at approximately 200, 400, 630, and 850 ms for POA. The major discrepancy between studies for these two cues appears to be centered on variations in the ERP's middle latencies between 400 and 630 ms, as well as the scalp distributions for these effects. This suggests that different mechanisms must subserve these two different speech cues. With age, responses to both speech cues appear to occur earlier in time as indicated by shorter latency responses in the ERP waveforms. In addition, ERPs to speech materials appear to generate multiple responses in the infant waveform with an earlier ERP peak varying bilaterally while a second, later peak occurs in a lateralized fashion. Across both speech cues bilateral responses appear to occur with similar latencies. The lateralized responses for these two cues, however, show much more variability. POA and VOT thus elicit different lateral patterns with
Cognitive Development
373
different latencies. This is a further indication that different mechanisms must underlay these two different speech cues. Studies of infant memory, on the other hand, report many bilateral effects, often restricted to frontal areas. Latencies for these responses initially appear to occur between 100 - 500 ms and between 500 - 1000 ms. A later response beyond 1000 ms appears to emerge by 6-months of age (Nelson, Ellis, Collins, & Lang, 1990) while the response prior to 500 ms is less well noted. Finally, word related effects appear even at the earliest ages tested (12 months of age) to elicit both left and right hemisphere responses throughout the first 500 ms of the wave. Even with older infants at 16 and 22 months of age, discrimination of known from unknown words continues to occur between 200 and nearly 700 ms following the onset of the word. Thus, there is a marked pattern of overlap between ERP components which (1) signal the reception of acoustic and visual information with (2) those components which indicate recognition and recall, to (3) those peak changes which reflect the understanding of word meanings. Instead of the simplistic serial model of processing described at the outset of this chapter, the infant brain appears to process different types of information (i.e., speech sounds, memory, and meaning) in an overlapping fashion and not necessarily in the expected temporal order. There are, of course, some marked differences in the ages of the infants tested for each cognitive domain and such differences could contribute to the overlapping latency results. Younger infants might simply produce longer latency responses regardless of the cognitive area of test because of their more immature nervous system and the additional time needed for information to travel along incompletely myelinated pathways to dendritic trees which are still relatively early in their developmental life. The bulk of the infant speech perception studies focus on infants from birth through 5 months of age while the memory studies report findings from infants ranging in age from four weeks to 12 months of age. The earliest work with word discrimination appears to be have the least overlap in subject ages tested with these other cognitive areas since testing did not commence until 12 months of age (Molfese, Wetzel, & Gill, 1993)and then extended upwards to 20 months of age (Mills et al, 1993). Nevertheless, from the existing data with vowel perception (Molfese & Searock, 1986), memory (Nelson & Karrer, 1992; Nelson & deRegnier, 1992; Molfese, 1989, Molfese & Wetzel, 1992), and word discrimination studies (Molfese, 1989, 1990,
374
Molfese and Narter
Molfese, Morse, & Peters, 1990; Molfese, Wetzel, & Gill, 1993) we can still discern that by one-year of age some speech discrimination effects occur slightly earlier in time than some word related effects, which all occur earlier than the memory effects reported for older infants. The one question we started out with at the beginning of this chapter now becomes many questions. Are the memory studies which produce such late responses relative to these other domains measuring factors which simply are unrelated to speech perception and word meaning? Are speech perception and word discrimination more automated at this stage and consequently require less time for processing or are there some innately specified mechanisms which subserve at least some aspects of speech perception which contribute to those faster response times? Does the infant's early knowledge of perception for coarticulated speech information tap such mechanisms and consequently results in such faster processing time for word discrimination? It is obvious from this review that there are a large number of gaps in our knowledge about each of these three domains of infant cognition. In fact, we still know very little about the neurophysiological development of mechanisms underlying not only speech perception, but memory, and early language development as well. Clearly, there is a great deal of work that still needs to be done before questions concerning the integration of infant cognition can be adequately addressed.
Acknowledgements Support for this work was provided by the National Science Foundation (BNS8004429, BNS 8210846), and the National Institutes of Health (R01-HD 17860).
References Ali, L., Gallagher, T., Goldstein, J., & Daniloff, R. (1971). Perception of coarticulated nasality. Journal of the Acoustical Society of America, 49, 538-540. Bates, E. (1979). Emergence of symbols. New York: Academic Press. Bates, E., Benigni, L., Bretherton, I., Camaioni, L., and Volterra, V. (1979). The emergence of symbols: Cognition and communication in infancy. New York: Academic Press.
Cognitive Development 375 Bates, E., Bretherton, I., Snyder, L., Shore, C., & Volterra, V. (1980). Vocal and gestural symbols at 13 months. Merrill-Palmer Quarterly, 26, 407- 423. Benedict, H. (1975, April). The role of repetition in early language comprehension. Paper presented at the meeting of the Society for Research in Child Development, Denver. Best, C.T. (1988). The emergence of cerebral asymmetries in early human development: A literature review and a neuroembryological model. In D.L. Molfese & S.J. Segalowitz (Eds.), Brain lateralization in children: Developmental implications., 5-34. New York: Guilford Press. Bloom, L., Lahey, M., Hood, L., Lifter, K., & Fiess, K. (1980). Complex sentences: Acquisition of syntactic connectives and the semantic relations they encode. Journal of Child Language, 7, 235-261. Blumstein, S., Baker, E., & Goodglass, H. (1977). Phonological factors in auditory comprehension in aphasia. Neuropsychologia, 15, 19-30. Brown, W. S., Marsh, J. T., & Smith, J. C. (1979). Principal component analysis of ERP differences related to the meaning of an ambiguous word. Journal of Electroencephalography and Clinical Neurophysiology, 46,706-714. Bryden, M.P. (1982). Laterality: Functional asymmetry in the brain,(pp 160-180). New York: Academic Press. Callaway, E., Tueting, P., & Koslow, S. H. (1978). Event-related brain potentials and behavior. New York: Academic Press. Chapman, R. M., McCrary, J. W., Bragdon, H. R., & Chapman, J. A. (1979). Latent components of event-related potentials functionally related to information processing. In J. E. Desmedt (Ed.), Progress in Clinical Neuropsychology, Vol. 6: Cognitive Components in Cerebral Event-Related Potentials and Selective Attention. Basel: Karger. Clark, E. V. (1983). Meanings and concepts. In Paul H. Mussen (Ed.), Handbook of Child Psychology, Vol. 3, (pp 787-840). (4th ed.). New York: Wiley. Courchesne, E., Ganz, L., & Norcia, A.M. (1981). Event related brain potentials to human faces in infants. Child Development, 52, 804811. Cowan, N., Suomi, K., & Morse, P. A. (1982). Echoic storage in infant perception. Child Development, 53, 984-990. Daniloff, R., & Moll, K. (1968). Coarticulation of lip rounding. Journal of Speech and Hearing Research, 11, 707-721. DeHaan, M., & Nelson, C. A. (1997). Recognition of the mother's face by six-month-old infants: A neurobehavioral study. Child Development, 68(2), 187-210. Dehaene-Lambert, G. & Dehaene, S. (1994). Speed and cerebral correlates of syllable discrimination in infants. Nature, 370, 292-295. Donchin, E., Tueting, P., Ritter, W., Kutas, M., & Heffley, E. (1975). On the independence of the CNV and the P300 components of the
376
Molfese and Narter
human averaged evoked potential. Journal of Electroencephalography and Clinical Neurophysiology , 38, 449-461. Dorman, M. (1974). Auditory evoked potential correlates of speech sound discrimination. Perception & Psychophysics, 15, 215-220. Eilers, R., Wilson, W., and Moore, J. (1979). Speech discrimination in the language-innocent and language-wise: A study in the perception of voice onset time. Journal of Child Language, 6, 1-18. Eimas, P.D., Siqueland, E., Jusczyk, P., & Vigorito, J. (1971). Speech perception in infants. Science, 171,303-306. Gelfer, M.P. (1987). An AER study of stop-consonant discrimination. Perception & Psychophysics, 42, 318-327. Golinkoff, R.M., Hirsh-Pasek, K., Cauley, K.M., & Gordon, L. (1987). The eyes have it: Lexical and syntactic comprehension in a new paradigm. Journal of Child Language, 14, 23-45. Gunderson, V. M., & Sackett, G. P. (1984). Development of pattern recognition in infant pigtailed macaques (Macaca nemestrina). Developmental Psychology, 20, 418-426. Gunderson, V. M., Swartz, K. B. (1985). Visual recognition in infant pigtailed macaques after a 24-hour delay. American Journal of Primatology, 8, 259-264. Gunderson, V. M., Swartz, K. B. (1986). Effects of familiarization time on visual recognition memory in infant pigtailed macaques (Macaca nemestrina ). Developmental Psychology, 22, 477-480. Hirsh-Pasek, K. & Golinkoff, R.M. (1996). The origins of grammar: Evidence from early language comprehension. Cambridge, MA: The MIT Press. Hoffman, M. & Salapatek, P. (1981). Young infants' event-related potentials (ERPs) to familiar and unfamiliar visual and auditory events in a recognition memory task. Journal of Electroencephalography and Clinical Neurophysiology, 52, 405-417. Hoffman, M., Salapatek, P., & Kuskowski, M. (1981). Evidence for visual memory in the averaged and single-trial evoked potentials in human infants. Infant Behavior and Development, 4, 401-421. Jasper, H. H. (1958). The ten-twenty electrode system of the International Federation of Societies for Electroencephalography: Appendix to report of the committee on methods of clinical examination in electroencephalography. Journal of Electroencephalography and Clinical Neurophysiology ,10, 371-375. Kamhi, A.G. (1986). The elusive first word: the importance of the naming insight for the development of referential speech. Journal of Child Language, 13, 155-161. Karrer, R. & Ackles, P. (1988). Visual event-related potentials of infants during a modified oddball procedure, (pp. 603-608). In R. Johnson, Jr., J. Rohrbaugh, & R. Parasuraman. (Eds.), Current trends in eventrelated potential research (EEG Suppl. 40).
Cognitive Development 377 Karrer, R., & Monti, L. A. (1995). Event-related potentials of 4-7-weekold infants in a visual recognition memory task. Elctroencephalography and Clinical Neurophysiology, 94, 414-424. Karrer, R., Wojtascek, Z., & Davis, M. G. (1995). Event-related potentials and information processing in infants with and without Down Syndrome. American Journal on Mental Retardation, 100, 146-159. Kuhl, P.K. (1985). Constancy, categorization, and perceptual organization for speech and sound in early infancy. In J. Mehler and R. Rox (Eds.), Neonate cognition: Beyond the blooming, buzzing confusion. Hillsdale, NJ: Erlbaum. Kurtzberg, D., Hilpert, P. L., Kreuzer, J. A., & Vaughan, H. G., Jr. (1984). Differential maturation of cortical auditory evoked potentials to speech sounds in normal fullterm and very low-birthweight infants. Developmental Medicine and Child Neurology, 26, 466-475. Lenneberg, E. (1967). Biological Foundations of Language. New York: Wiley & Sons. Liberman, A. M., Cooper, F. S., Shankweiler, D., & Studdert-Kennedy, M. (1%7). Perception of the speech code. Psychological Review, 74, 431-461. Lisker, L. & Abramson, A.S. (1970). The voicing dimension: Some experiments in comparative phonetics. In: Proceedings of the 6th International Congress of Phonetic Sciences. Prague: Academia. 563 -567. Lyon, G.R. (1994). Frames of reference for the assessment of learning disabilities: New views on measurement issues. Baltimore: Brookes. Macnamara, J. (1982). Names for things. Cambridge, MA: MIT Press. MacNeilage, P. F. & DeClerk, J. L. (1969). On the motor control of coarticulation in CVC monosyllables. Journal of the Acoustical Society of America, 45, 1217-1233. Miceli, G., Caltagirone, C., Gianotti, G., & Payer-Rigo, P. (1978). Discrimination of voice versus place contrasts in aphasia. Brain and Language, 6, 47-51. Miller, J. F. & Chapman, R. S. (1981). The relation between age and mean length of utterance in morphemes. Journal of Speech and Hearing Research, 24, 154-161. Mills, D. L., Coffey-Corina, S. A., & Neville, H. J. (1993). Language acquisition and cerebral specialization in 20-month-old infants. Journal of Cognitive Neuroscience, 5, 317-334. Molfese, D. L. (1972). Cerebral asymmetry in infants, children and adults: Auditory evoked responses to speech and music stimuli. Unpublished doctoral dissertation, The Pennsylvania State University. Molfese, D. L. (1978a). Electrophysiological correlates of categorical speech perception in adults. Brain and Language, 5, 25-35.
378
Molfese and Narter
Molfese, D. L. (1978b). Left and right hemispheric involvement in the speech perception: Electrophysiological correlates. Perception and Psychophysics, 23, 237-243. Molfese, D. L. (1979). Cortical and subcortical involvement in the processing of coarticulated cues. Brain and Language, 7, 86-100. Molfese, D. L. (1980a). Hemispheric specialization for temporal information: Implications for the processing of voicing cues during speech perception. Brain and Language, 11, 285-300. Molfese, D. L. (1980b). The phoneme and the engram: Electrophysiological evidence for the acoustic invariant in stop consonants. Brain and Language, 9, 372-376. Molfese, D. L. (1983). Event related potentials and language processes. In A.W.K. Gaillard & W. Ritter (Eds.), Tutorials in ERP Research: Endogenous Components, (pp 345-368). The Netherlands: North Holland Publishing Co. Molfese, D. L. (1984). Left hemisphere sensitivity to consonant sounds not displayed by the right hemisphere: Electrophysiological correlates. Brain and Language, 22, 109-127. Molfese, D. L. (1989). Electrophysiological correlates of word meanings in 14-month-old human infants. Developmental Neuropsychology, 5, 79-103. Molfese, D. L. (1990). Auditory evoked responses recorded from 16month-old human infants to words they did and did not know. Brain and Language, 38, 345-363. Molfese, D. L. & Betz, J. C. (1988). Electrophysiological indices of the early development of lateralization for language and cognition and their implications for predicting later development. In D.L. Molfese and S.J. Segalowitz (Eds.), Developmental Implications of Brain Lateralization, (pp. 171-190). New York: Guilford Press. Molfese, D. L. & Hess, R. M. (1978). Speech perception in nursery school age children" Sex and hemispheric differences. Journal of Experimental Child Psychology, 26, 71-84. Molfese, D. L. & Molfese, V. J. (1979a). Infant speech perception: Learned or innate. In H. A. Whitaker and H. Whitaker (Eds.), Advances in Neurolinguistics, Vol. 4. New York: Academic Press. Molfese, D. L. & Molfese, V. J. (1979b). Hemisphere and stimulus differences as reflected in the cortical responses of newborn infants to speech stimuli. Developmental Psychology, 15, 505-511. Molfese, D. L. & Molfese, V. J. (1980). Cortical responses of preterm infants to phonetic and nonphonetic speech stimuli. Developmental Psychology, 16, 574- 581. Molfese, D. L. & Molfese, V. J. (1997). Discrimination of language skills at five years of age using event related potentials recorded at birth. Developmental Neuropsychology,13, 135-156. Molfese, D. L. & Molfese, V. J. (1985). Electrophysiological indices of auditory discrimination in newborn infants: The basis for predicting
Cognitive Development 379 later language development. Infant Behavior and Development, 8, 197-211. Molfese, D. L. & Molfese, V. J. (1988). Right hemisphere responses from preschool children to temporal cues contained in speech and nonspeech materials: Electrophysiological correlates. Brain and Language, 33, 245-259. Molfese, D. L. & Schmidt, A. L. (1983). An auditory evoked potential study of consonant perception. Brain and Language, 18, 57-70. Molfese, D. L. & Searock, K. (1986). The use of auditory evoked responses at one year of age to predict language skills at 3 years. Australian Journal of Communication Disorders, 14, 35-46. Molfese, D. L. & Wetzel, W. F. (1992). Short and long term memory in 14 month old infants: Electrophysiological correlates. Developmental
Neuropsychology, 8, 135-160. Molfese, D. L., Buhrke, R. A., & Wang, S. L. (1985). The right hemisphere and temporal processing of consonant transition durations" Electrophysiological correlates. Brain and Language, 26, 289-299. Molfese, D. L., Burger-Judisch, L. M., & Hans, L. L. (1991). Consonant discrimination by newborn infants: Electrophysiological differences. Developmental Neuropsychology, 7, 177-195. Molfese, D. L., Freeman, R. B., Jr., & Palermo, D. S. (1975). The ontogeny of lateralization for speech and nonspeech stimuli. Brain
and Language, 2, 356- 368. Molfese, D. L., Linnville, S. E., Wetzel, W. F., & Leicht, D. (1985). Electrophysiological correlates of handedness and speech perception contrasts. Neuropsychologia, 23, 77-86. Molfese, D. L., Morse, P. A., & Peters, C. J. (1990). Auditory evoked responses from infants to names for different objects: Cross modal processing as a basis for early language acquisition. Developmental Psycho logy, 26, 780-795. Molfese, D. L., Wetzel, W. F., & Gill, L. A. (1993). Known versus unknown word discrimination in 12-month-old human infants: Electrophysiological correlates. Developmental Neuropsychology, 34, 241-258. Molfese, V. J., Molfese, D. L., & Parsons, C. (1983). Hemispheric involvement in phonological perception. In S. Segalowitz (Ed.), Language Functions and Brain Organization, (pp. 29-50). New York: Academic Press. Morse, P. A. (1974). Infant speech perception: A preliminary model and review of the literature. In R. Schiefelbusch and L. Lloyd (Eds.),
Language perspectives: Acquisition, retardation, and intervention, (pp. 19-53). Baltimore: University Park Press. Morse, P. A. & Snowdon, C. (1975). An investigation of categorical speech discrimination by rhesus monkeys. Perception &
Psychophysics, 17, 9-16.
380
Molfese and Narter
Nelson, C. A. & Collins, P. (1991). An event related potential and looking time analysis of infants' responses to familiar and novel events" Implications for visual recognition memory. Developmental Psychology, 27, 50-58. Nelson, C. A. & Collins, P. (1992). Neural and behavioral correlates of recognition memory in 4- and 8-month olds infants. Brain and Cognition, 19, 105-121. Nelson, C. A. & deRegnier, R. (1992). Neural correlates of attention and memory in the first year of life. Developmental Neuropsychology, 8, 119-134. Nelson, C. A. & Salapatek, P. (1986). Electrophysiological correlates Of infant recognition memory. Child Development, 57, 1483-1497. Nelson, C. A., Ellis, A., Collins, P., & Lang, S. (1990). Infants' neuroelectric responses to missing stimuli: Can the missing stimuli be novel stimuli? Developmental Neuropsychology, 6, 339-349. Nelson, C. A., Henschel, M. & Collins, P. (1993). Neural correlates of cross-modal recognition memory in 8-month-old infants.
Developmental Psychology,. Nikkel, L., & Karrer, R. (1994). Differential effects of experience on the ERP and behavior of 6-month-old infants: Trends during repeated stimulus presentations. Developmental Neuropsychology, 10, 1-11. Oldfield, R. L. (1971). The assessment of handedness: The Edinburgh Inventory. Neuropsychologia, 9, 97-113. Perecman, E., & Kellar, L. (1981). The effect of voice and place among aphasic, nonaphasic right-damaged and normal subjects on a metalinguistic task. Brain and Language, 12, 213-223. Pisoni, D. B. (1977). Identification and discrimination of the relative onset time of two component tones" Implications for voicing perception in stops. Journal of the Acoustical Society of America, 61, 1352-1361. Retherford, K. S., Schwartz, B. C., & Chapman, R. S. (1981). Semantic roles and residual grammatical categories in mother and child speech. Journal of Child Language, 8, 583-608. Rockstroh, B., Elbert, T., Birbaumer, N., & Lutzenberger, W. (1982). Slow brain potentials and behavior. Baltimore: Urban & Schwarzenberg. Ruchkin, D., Sutton, S., Munson, R., Silver, K., & Macar, F. (1981). P300 and feedback provided by absence of the stimulus. Psychophysiology, 18, 271-282. Segalowitz, S. J. & Cohen, H. (1989). Right hemisphere EEG sensitivity to speech. Brain and Language, 37, 220-231. Simos, P. G., & Molfese, D. L. (1997). Electrophysiological responses from a temporal order continuum in the newborn infant. Neuropsychologia, 35, 89-98.
Cognitive Development 381 Simos, P. G., Molfese, D. L., Brenden, R. A. (1997). Behavioral and electrophysiological indices of voicing cue discrimination: Laterality patterns and development. Brain and Cognition, Snyder, L., Bates, E., and Bretherton, I. (1981). Content and context in early lexical development. Journal of Child Language, 8, 565-582. Streeter, L. A. (1976). Language perception of two-month-old infants shows effects of both innate mechanisms and experience. Nature, 259, 39-41. Sullivan, M., Rovee-Collier, C., Tynes, D. (1979). A conditioning analysis of infant long-term memory. Child Development, 50, 152-162. Thomas, D. G. and Lykins, M. S. (1995). Event-related potential measures of 24-hour retention in 5-month-old infants. Developmental Psychology, 31,946-957.
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
383
Chapter 12
The lpsilateral Auditory Pathway: A Psychobiological Perspective. Kendall A. Hutson University of Toledo
Since the time of Ferrier (1876a, 1890) the ascending auditory system has been viewed as "contralateralized". That is, auditory information arriving at one cochlea traverses the central auditory pathways destined for the contralateral auditory cortex. For nearly a century, the pathway ultimately responsible for "hearing" was thought to be the projection of the dorsal cochlear nucleus to the contralateral medial geniculate body of the thalamus and from there to the auditory cortex (e.g., Winkler, 1911, 1921). Not all investigators agreed with the conclusion that only pathways from the dorsal cochlear nucleus subserved heating. Indeed, the view of a contralateralized auditory system was not universally accepted. One of the earliest investigations into the function of auditory cortex was conducted by Luciani (1884). From his studies, Luciani provided some very important insights as to the function of the ascending auditory system. He noted that subsequent to unilateral ablation of auditory cortex, an animals behavior could only be explained by "each ear having connections with both auditory spheres [cortex] ... In fact, every unilateral extirpation of sufficient extent in the province of the auditory sphere causes a bilateral disorder of hearing, more marked on the opposite side." He concluded that in the auditory system, as in the visual system "we must distinguish a crossed and a direct fasciculus."
384
Hutson
Thompson (1878) had expressed a similar view, and further stated the true nature of the problem: "to account for a blending of the sensations [derived from the two ears] ... this point deserves the attention of anatomists and physiologists." By the turn of the century, the anatomical bilaterality of the central auditory pathways had been well documented (e.g., Bechterew, 1885; Baginski, 1886, 1890; Monakow, 1887, 1890; Bumm, 1889; Held, 1891, 1893; Kolliker, 1896; Ferrier & Turner, 1898; Tschermak, 1899). Indeed, the findings of these Continental anatomists had even made their way into American textbooks, e.g., Gordinier (1899). Physiological evidence for a functional ipsilateral component of the ascending auditory pathway was provided by Kreidl (1914). Here, after section of the decussating contralateral pathways experimental animals "showed no difference in hearing ability from normal control animals." Thus, it seemed an inescapable conclusion that the ascending auditory system was composed of both ipsilateral and contralateral pathways. Unfortunately, as noted by Rosenzweig (1961), "After these early achievements, progress not only lagged but some of the findings were even forgotten by workers in the field. Thus, for example, it was taken as surprising in 1928 when removal of one hemisphere of a patient did not destroy hearing in the opposite ear." Again the view of a contralateralized auditory system prevailed, and over the years the notion of a functional ipsilateral pathway was subjected to relative obscurity. Textbook descriptions of the central auditory system referred only vaguely to the existence of ipsilateral components of the system "which is known more on a clinical and physiological than an anatomical basis" (Fulton, 1946; Ranson & Clark, 1947). Further, the underlying pathway for heating was once again reduced only to the contralateral projections of the dorsal cochlear nucleus (e.g., Strong & Elwyn, 1953). From the results of modern investigations of the central auditory system, the concept of a functional ipsilateral pathway has re-emerged. Moreover, it is clear that ipsilateral pathways have an important role in the so-called "acoustic chiasm" of Glendenning and Masterton (1983). Therefore, it seems timely to re-evaluate the nature of the ipsilateral auditory system and its role in audition. The general outline of this paper will be to discuss evidence favoring a functional ipsilateral auditory pathway. Results from anatomical, physiological, and behavioral investigations will be introduced as they pertain to the question of the potential significance of ipsilateral
Auditory Pathway
385
pathways. Particular attention will be paid to physiological mechanisms of the auditory system in which a functional ipsilateral pathway may help to explain.
I. Anatomy of the Ascending Auditory System Briefly, the ascending central auditory system begins with the termination of the eighth nerve onto the secondary sensory neurons of the cochlear nuclear complex. The cochlear nuclei then give rise to three separate, yet not completely isolated central pathways known as the acoustic striae. The striae traverse the brainstem sending axons, in various amounts, to the superior olivary complex, the nuclei of the lateral lemniscus, and the inferior colliculus. Neurons of the superior olivary complex send their axons to the nuclei of the lateral lemniscus and to the inferior colliculus. The nuclei of the lateral lemniscus project mainly to the inferior colliculus. The inferior colliculus then originates fibers destined for the medial geniculate body, which in turn issues fibers of the auditory radiations to innervate the auditory cortex. Various commissural systems exist which interconnect the right and the left halves of the central pathways. In the sections that follow, details of the central auditory system will be presented. These will be concerned primarily with the pathways from the cochlear nucleus to the inferior colliculus, accepting the established fact that projections from inferior colliculus through auditory cortex remain, for all practical purposes, on the same or homolateral side of the brain (see Hutson, 1988; Hutson et al., 1991; 1993). Therefore, above the level of the inferior colliculus there is no need to distinguish a substantial ipsilateral versus contralateral fiber pathway.
Nuclei of the Central Auditory System Cochlear Nucleus. This complex of cells is composed of two major divisions, one dorsal (dorsal cochlear nucleus, DCN), the other ventral (ventral cochlear nucleus, VCN). The entering eighth nerve further subdivides the VCN into an anterior portion (anterior ventral cochlear nucleus, AVCN) and a posterior portion (posterior ventral cochlear nucleus, PVCN). Having entered the cochlear nucleus, the eighth nerve bifurcates giving rise to an ascending branch and a descending branch. The ascending branch terminates in the AVCN, while the descending
386
Hutson
branch innervates both the PVCN and the DCN (Ramon y Cajal, 1909; Lorente De No, 1933; Rose, 1960; Jones et al., 1992, Ryugo and Rouiller, 1988). Although the cochlear nucleus is made up of at least 14 different cell types based on Golgi material (e.g., Lorente De No, 1933; Brawer et al., 1974), the most useful cell classifications are those of Osen (1969a), Warr (1982), and Morest et al., (1990) which are based on Nissl preparations. Following this scheme, there are five major cell types in the cochlear nucleus. Three of these cell types are unique to one of the three divisions of the cochlear nucleus: pyramidal (or fusiform) cells in the DCN; spherical cells in the AVCN; and octopus cells in the PVCN. The remaining two cell types, globular and multipolar, are distributed around the entrance of the eighth nerve and therefore lie in both the AVCN and PVCN. The VCN can then be visualized as having spherical cells rostral, globular and multipolar cells in middle regions, and octopus cells caudally (Osen, 1969a). Along its margins, the cochlear nucleus is surrounded or encapsulated by a small cell shell or cap, composed predominantly of granule cells (Mugnaini et al., 1980; Hutson and Morest, 1996). Superior Olivary Complex. The superior olivary complex consists of four major sub-nuclei, all of which are embedded in the fibers of the trapezoid body. These are the lateral superior olive (LSO), the medial superior olive (MSO), the medial nucleus of the trapezoid body (MTB), and a group of cells collectively termed the peri-olivary nuclei. The peri-olivary nuclei are cell groups which surround the LSO, MSO, and MTB (Morest, 1968), and for the most part participate in descending connections of the auditory system (though some exceptions may exist, e.g., see Adams, 1983; Spangler et al., 1985, Schofield & Cant, 1992; Ostapoff, et al., 1997; Warr & Beck, 1996). Since the peri-olivary nuclei probably do not contribute greatly to the ascending auditory system, they will not be discussed in detail. Lateral Superior Olive. The predominant cell type of the LSO is fusiform in shape, possessing large polar dendrites extending toward the margins of the nucleus. The distinguishing characteristic of the LSO is its convoluted shape. Typically the nuclear contour follows an S-shape, as seen in the cat, although there are some species differences (e.g., West, 1970; Moore, 1987; Heffner & Masterton, 1990). Despite species variation in shape, the neurons and their appendages always orient themselves in such a way to be approximately perpendicular to the
Auditory Pathway
387
margins of the nucleus (Held, 1893; Hofmann, 1908; Stotler, 1953; Scheibel & Scheibel, 1974; Cant, 1984; Helfert & Schwartz, 1987). Medial Superior Olive. The MSO has certain striking morphological qualities that make it easily discernible from other nuclei of the superior olivary complex. It is composed of a cluster of cells arranged in a dorsal-ventral stack or column, (often appearing comma-shaped as in the cat), bounded laterally by the LSO and medially by the MTB. Individual cells of MSO are quite unique in that radiating from their cell bodies are two large diametrically opposed dendrites, one pointing laterally the other medially. These singular properties of MSO have been well documented across a wide variety of mammals (La Villa, 1898; Hofmann, 1908, Ramon y Cajal, 1909; Shaner, 1934; Stotler, 1953; Verhaart, 1970; Moore & Moore, 1971, Harrison & Howe, 1974; Willard & Martin, 1983; Henkel & Brunso-Bechtold, 1990). Medial Nucleus of the Trapezoid Body. This nucleus resides ventrally in the medulla within the fibers of the trapezoid body, located between the MSO laterally and the exiting sixth nerve medially. The nucleus generally appears roughly triangular in shape, although it's cells do not conform to any distinctive geometric outline as do cells of the LSO and MSO. The main cellular component of MTB is the principal cell which is large and round to ovoid in shape (e.g., Morest, 1968; Harrison & Feldman, 1970; Willard & Martin, 1983; Kuwabara & Zook, 1991). Nuclei of the Lateral Lemniscus. The nuclei of the lateral lemniscus lie among the fiber fascicles of the lateral lemniscus, collectively bridging the pons between the medullary superior olivary complex and the inferior colliculus at midbrain levels. Nuclei of the lateral lemniscus have classically been divided into two divisions, one ventral (ventral nucleus of the lateral lemniscus, VLL) the other dorsal (dorsal nucleus of the lateral lemniscus, DLL). The VLL is the larger of the two, and can be divided further into three zones, ventral, middle, and dorsal. Recently, the dorsal region of VLL has been referred to as the intermediate nucleus of the lateral lemniscus (ILL; e.g., Glendenning et al., 1981; Zook & Casseday, 1982a; Willard & Ryugo, 1983; Hutson et al., 1991). The VLL is composed of several cell types, with multipolar cells scattered throughout its length. There is a ventro-dorsal gradient of distinguishing cell types. Ventrally, large oval cells give way to a higher density of elongate cells in the middle zone, which become less packed in the dorsal zone (Whitley & Henkel, 1984). The DLL also contains a
388
Hutson
variety of cell types, the most distinguishing of which are elongate cells with their dendrites oriented horizontally (Kane & Barone, 1980). Inferior Colliculus. In Nissl preparations, the inferior colliculus is divisible into four sub-nuclei: the central nucleus, the dorsomedial nucleus, the pericentral nucleus, and the external nucleus. With respect to the ascending auditory system, the central nucleus of the inferior colliculus is of primary importance. The central nucleus is encapsulated by the pericentral nucleus (dorsally and caudally), the external nucleus (laterally), the dorsomedial nucleus (dorsomedially, as the name implies), and the peri-aquaductal grey (ventromedially, see Rockel & Jones, 1973a; Harrison, 1978). Further subdivision is possible in Golgi stained material (e.g., Morest & Oliver, 1984). Two cell types characterize the central nucleus of the inferior colliculus, primary cells with disc shaped dendritic fields and multipolar or stellate cells with spherical dendritic fields (Rockel & Jones, 1973a; Oliver & Morest, 1984; Oliver et al., 1991). For simplicity, in the remaining sections of this paper the term IC will be used to denote the central nucleus of the inferior colliculus.
Basic Physiological Properties There are three basic physiological characteristics that define the neurons of the auditory brainstem: the unit discharge properties, the binaural response classification, and the tonotopic arrangement within the individual nuclei. Unit Discharge. Neurons below the level of the inferior colliculus have, as a general characteristic, unit discharge patterns that are tonic in nature. Although variations in discharge patterns have been observed (i.e. "choppers" or "primary-notch" or "complex") all have the common characteristic of a maintained discharge throughout the stimulus period (e.g., Goldberg & Brown, 1968; Boudreau & Tsuchitani, 1970; Aitkin et a1.,1970, 1981; Guinan et al., 1972a; Tsuchitani, 1977; Brugge & Geisler, 1978; Feng et al., 1994). At the IC, the situation is quite different. Here, the tonic or sustained discharge type unit response is seen, but the much more prevalent response type is that of a phasic or "onset" type of discharge (e.g., Erulkar, 1959, 197.5; Rose et al., 1963). Binaural Response Classification. Within the central auditory system, neurons can be classified based on their response to stimuli presented to both ears (e.g., Goldberg & Brown, 1968). For example,
Auditory Pathway
389
(P C
F i g u r e 1. Tonotopic organization in auditory brainstem nuclei of the cat. Note the preservation of the arrangement of frequencies (H, high; L, low) in each nucleus. A similar arrangement is present in all other auditory structures. Also shown are the unique neuronal and dendritic arrangements which characterize the medial superior olive (MSO) and the lateral superior olive (LSO). AVCN, anterior ventral cochlear nucleus; DCN, dorsal cochlear nucleus; DLL, dorsal nucleus of the lateral lemniscus; IC, inferior colliculus; MTB, medial nucleus of the trapezoid body; PVCN, posterior ventral cochlear nucleus; VLL, ventral nucleus of the lateral lemniscus. Based on Stotler, 1953; Rose et al., 1959; Aitldn et al., 1970; Guinan et al., 1972b; Merzenich & Reid, 1974.
neurons of the LSO respond to acoustic stimuli presented to either ear, typically stimulation of the ear ipsilateral to LSO causes excitation while stimulation of the contralateral ear results in inhibition (e.g., Goldberg et al., 1963). Thus, LSO units may be classified binaurally as El, (meaning ipsilateral = Excitatory, contralateral = Inhibitory; Goldberg & Brown, 1968; Tsuchitani, 1977; Caird & Klinke, 1983)1. Similarly MSO can be classified as binaurally excitatory, or EE; in contrast, MTB is classified as OE meaning no response to ipsilateral ear stimulation but
390
Hutson
Right SOC MSO
MTB
Figure 2. Details of tonotopic organization of the right superior olivary complex (SOC) of the cat. Numbers reflect range of best frequencies in kHz of units located within the SOC. Note that high frequencies are represented medially in the medial nucleus of the trapezoid body (MTB), ventrally in the medial superior olive (MSO), and medially (medial "limb") in the lateral superior olive (LSO). Low frequencies are represemed laterally in MTB, dorsally in MSO, and laterally (lateral "limb") in LSO. Also note that both MTB and LSO have a larger cross-sectional area devoted to the representation of higher frequencies (e.g., greater than 4 kHz) while MSO has a larger cross-sectional area devoted to the representation of lower frequencies (e.g., less than 4 kHz). Adapted from Boudreau and Tsuchitani, 1970; Guinan et al., 1972b.
excited by contralateral ear stimulation (Goldberg & Brown, 1968; Guinan et al., 1972a; Caird & Klinke, 1983; Spitzer & Semple, 1995). Units within VLL are OE; while the DLL appears to have two types of binaural units, EE cells are found dorsally in DLL and IE cells ventrally in DLL (Aitkin et al., 1970; Brugge et al., 1970; Markovitz and Pollak, 1994). The IC has a mixed population of binaurally classified cells, most of which are IE, with EE and OE units also being quite common (e.g., Roth et al., 1978; Semple & Aitkin, 1979; Aitkin & Martin, 1987). Tonotopicity. A common organizational feature amongst the various auditory nuclei is that they are arranged in an orderly tonotopic sequence. Within the cochlea, moving from the apical to basal turns,
Auditory Pathway
3 91
there is an orderly progression of sensitivity from low to high frequencies. This orderly progression is followed through the eighth nerve onto the cochlear nucleus (e.g., Rose, 1960; Arnesen & Osen, 1978). Electrophysiological studies have reported that within each auditory nucleus a similar orderly progression of frequency sensitivity can be demonstrated (e.g., Rose et al., 1959, 1963; Rose, 1960; Boudreau & Tsuchitani, 1970; Aitkin et al., 1970; Guinan et al., 1972b; Merzenich & Reid, 1974; Bourk et al., 1981; Clarey et al., 1992). Thus, there is a tonotopic or more correctly a cochleotopic, organization topograghically mapped onto each auditory nucleus (see Figure 1). Particular attention has been given to demonstrating the frequency distribution within the superior olivary complex (Figure 2). The LSO and MTB appear to have more nuclear volume devoted to the representation of middle to high range frequencies, whereas MSO has most of its nuclear mass devoted to low to middle range frequencies (Boudreau & Tsuchitani, 1970, 1973; Guinan et al., 1972b; Tsuchitani, 1977, 1982; Yin & C h a n , 1990). Other auditory nuclei do not show such a distinct frequency specificity, rather they show a more even representation of frequency sensitivity (e.g., Rose, 1960; Merzenich & Reid, 1974).
The Central Auditory Pathways The central auditory pathways will first be described in terms of an integrated system, in which both crossed (contralateral) and uncrossed (ipsilateral) connections will be included. A discussion of the crossed pathway is necessary for arguments that will be made in later sections of this paper. The final anatomy section will detail the ipsilateral pathway. Connections of the Cochlear Nucleus. Neurons of the cochlear nucleus give rise to all second order auditory afferent fibers, which are collectively termed the acoustic striae. The acoustic striae compose, in part, the medullary decussation of the auditory system and were well known to the Classical anatomists (e.g., Flechsig, 1876; Monakow, 1890; Held, 1891, 1893; Kolliker, 1896; Van Gehuchten, 1902, 1903; Ramon y Cajal, 1909). The striae can be divided into three major pathways: the dorsal acoustic stria of Monakow (DAS); the intermediate acoustic stria of Held (IAS); and the ventral acoustic stria of Flechsig, better known as the trapezoid body. Axons of the three stria distribute themselves to the superior olivary complex, the nuclei of the lateral lemniscus, and the inferior colliculus in various quantities.
392
Hutson
The Dorsal Acoustic Stria contains fibers that originate primarily from the fusiform (or pyramidal) cells of the DCN. From the DCN axons pass dorsally over the restiform body, then sweep ventrally to cross the midline just dorsal to the superior olivary complex, where the fibers enter the medial aspects of the contralateral lateral lemniscus (Fernandez & Karapas, 1967; Van Noort, 1969). During their course through the brainstem fibers of the DAS may give off collaterals to the VLL, however virtually all of these fibers terminate throughout the contralateral IC (Osen, 1972, Adams & Warr, 1976; Beyerl, 1978; Adams, 1979; Brunso-Bechtold et al., 1981; Nordeen et al., 1983; Willard & Martin, 1983, Oliver, 1984). The Intermediate Acoustic Stria originates primarily from the octopus cells of the PVCN (though some fusiform cells of DCN may contribute axons to this pathway as well). Fibers of the IAS pass dorsally over the restiform body in association with the DAS. At this point the IAS separates and descends medial to the restiform body to a level just above the ipsilateral superior olivary complex where the fibers turn medially to pass through the nuclei and tract of the spinal trigeminal complex. The IAS then continues on a medial course just above the trapezoid body, crossing the midline and joining with the DAS to enter the medial aspects of the contralateral lateral lemniscus. Fibers of the IAS terminate primarily in the contralateral VLL, though contributing a scant projection to the ipsilateral LSO and contralateral IC (Fernandez and Karapas, 1967; Oliver, 1984). Unless specifically referred to by name (dorsal or intermediate acoustic stria), these two pathways will collectively be designated as the dorsal striae (DS; see Figure 3). The Trapezoid Body contains fibers originating from every cell group of the VCN (e.g., Warr, 1982), and is by far the largest fiber component of the medullary auditory system. Trapezoid body fibers exit the VCN ventral to the restiform body and course medially to innervate the ipsilateral superior olivary complex, the contralateral superior olivary complex, and then continue within the lateral aspects of the lateral lemniscus to reach the contralateral nuclei of the lateral lemniscus and IC (e.g., Flechsig, 1876; Held, 1891, 1893; Sabin, 1897; Ferrier & Turner, 1898; Van Gehuchten, 1902, 1903; Ramon y Cajal, 1909; Fuse, 1919; Poljak, 192.5; Barnes et al., 1943; Warr, 1966, 1969, 1972, 1982; Van Noort, 1969; Browner & Webster, 1975; Strominger et al., 1977; Brunso-Bechtold et al., 1981; Glendenning et al., 1981, 1985, Thompson & Thompson, 1991).
Auditory Pathway
393
ClC
,-q I I I I I I
C~ i . . , - , , . /
I
I I
!R
. i .
i m
-)'1_.i I
I
I
I I i
Lt
(
D8
I I
'sol TB
midline
Figure 3. Schematic diagram summarizing the major connections of auditory brainstem nuclei. Note that this is a simplified diagram, and each projection shown has a "mirror-image" on the opposite side of the midline. Solid lines major pathways, dashed lines minor pathways. Also for convenience the dorsal and intermediate acoustic striae have been depicted simply as the dorsal striae (DS). See text for details of connections. CIC, commissure of inferior colliculus; CPr, commissure of Probst; DCN; dorsal cochlear nucleus; DLL, dorsal nucleus of the lateral lemniscus; IC, inferior colliculus; LL, lateral lemniscus; LSO, lateral superior olive; MSO, medial superior olive; MTB, medial nucleus of the trapezoid body; TB, trapezoid body; VCN, ventral cochlear nucleus; VLL, ventral nucleus of the lateral lemniscus.
In addition to the three striae just described, axons of cells from all divisions of the cochlear nucleus enter the ipsilateral lateral lemniscus directly, ascending to the ipsilateral nuclei of the lateral lemniscus and ipsilateral IC. Although in terms of the number of cells involved, these ipsilateral cochlear nucleus projections are not as large as the contralateral projections, they do exist, as demonstrated by a variety of anato-
394
Hutson
mical 1902, 1982; 1979;
techniques (e.g., Baginski, 1886; Held, 1893; Van Gehuchten, 1903; Poljak, 1925; Woollard & Harpman, 1940; Warr, 1972, Brunso-Bechtold & Thompson, 1976; Roth et al, 1978; Adams, Glendenning et al, 1981; Nordeen et al., 1983; Oliver, 1987).
Connections of the Superior Olivary Complex The Medial Nucleus of the Trapezoid Body. One of the most salient characteristics of the MTB is in the mode of termination of its afferent supply. Since before the turn of the century it was known that each principal cell of MTB received a large nerve ending, the calyx of Held (Held, 1893; Ramon y Cajal, 1896; Meyer, 1896; Kolliker, 1896; La Villa, 1898; Turner & Hunter, 1899; Vincenzi, 1900; Veratti, 1900). It was assumed, on the basis of histological inspection of normal material, that it was the contralateral VCN which originated the fibers ending in calyces on MTB cells (e.g., Held, 1893), but the connection was proven by degeneration studies (e.g., Lewy, 1909). Ablation of the cochlear nucleus causes dense degeneration within the contralateral MTB, demonstrable by degeneration stains or by the loss of calyces in normal stains (Barnes et al., 1943; Stotler, 1953; Harrison & Warr, 1962; Powell & Erulkar, 1962; Warr, 1972; Jean-Baptiste & Morest, 1975). While unilateral cochlear nucleus destruction only affects the contralateral MTB, bilateral destruction of the cochlear nucleus or section of the trapezoid body near the midline causes total loss of calyces in the MTB of both sides, therefore the evidence is quite substantial that the source of afferent innervation to MTB is the contralateral cochlear nucleus (Tschermak, 1899; Fuse, 1916, 1919; Stotler, 1953; Harrison & Warr, 1962; Harrison & Irving, 1964). More recent investigations using restricted lesion placement, anterograde or retrograde axonal transport methods have demonstrated that it is the globular cells of the contralateral VCN, whose large axons traverse the trapezoid body, that terminate in calyces on MTB principal cells (e.g., Warr, 1972, 1982; Tolbert et al., 1982; Glendenning et al., 1985, Spirou et al., 1990; Smith et al., 1991; Kuwabara et al., 1991). The principal cells of MTB send their axons to three major sites, all of which are located homolateral to MTB; these are the LSO, the nuclei of the lateral lemniscus, and the IC. Axons of these cells pass over, under, and through MSO to reach their targets in LSO. Other axons, or more likely, collaterals enter the lateral aspects of the lateral lemniscus
Auditory Pathway
3 95
ascending primarily to the ILL, and to a much lesser extent to VLL, DLL and IC (Browner & Webster, 1975; Brunso-Bechtold & Thompson, 1976; Elverland, 1978; Glendenning et al., 1981, 1985; Held, 1893; Lewy, 1909; Ramon y Cajal, 1909; Rasmussen, 1946, 1967; Spangler et al., 1985; Van Noort, 1969). In summary, the MTB is a monaural nucleus receiving afferents from the contralateral cochlear nucleus, and distributing its axons to other auditory nuclei of the same side. The Medial Superior Olive. In contrast to MTB, the MSO is a binaural nucleus, receiving its primary afferent supply from the cochlear nucleus of both sides of the brain (Ferrier & Turner, 1898; Poljak, 1925; Stotler, 1949, 1953; Warr, 1966; Glendenning et al, 1985). Afferents to MSO terminate in a unique and specific manner. Destruction of one cochlea or cochlear nucleus will cause the loss of terminals or degeneration in MSO on both sides of the brain, but only on one dendrite, i.e., if the left cochlear nucleus is destroyed, degeneration will be seen on the laterally directed dendrites of the left (ipsilateral) MSO and on the medially directed dendrites of the fight (contralateral) MSO (Stotler, 1953). Thus, the medial dendrites receive afferents from the contralateral cochlear nucleus, while the lateral dendrites receive afferents from the ipsilateral cochlear nucleus. That afferents to MSO travel within the trapezoid body can be demonstrated by the various tract-tracing methods of degeneration (Lewy, 1909; Poljak, 1925; Woollard & Harpman, 1940; Warr, 1966; Harrison & Irving, 1966; Goldberg & Brown, 1968; Van Noort, 1969; Strominger & Strominger, 1971; Glendenning et al., 1981, 1985), or axoplasmic transport (e.g., Glendenning et al., 1985; Zook & Casseday, 1985; Smith et al., 1993). More specifically, section of the trapezoid body at the midline results in afferent degeneration on the medial dendrites of MSO on both sides of the brain, while section of the trapezoid body as it exits the V CN produces the same degeneration pattern as cochlear nucleus ablation (Stotler, 1953; Harrison & Warr, 1962). Restricted lesions of V CN, or axonal transport studies reveal that the cells of origin of MSO afferents are the spherical cells of the AVCN (Warr, 1966; Kiss & Majorossy, 1983; Zook & Casseday, 1985; Cant & Casseday, 1986; Smith et al., 1993). The MSO also receives indirect inputs from both cochlear nuclei via collateral innervation from other cell groups of the ipsilateral superior olivary complex. Fibers from MTB occasionally send collaterals to MSO while en route to other structures, and thus indirectly supply information from the contralateral cochlear nucleus, while some
396
Hutson
indirect information from the ipsilateral cochlear nucleus is provided by one of the peri-olivary nuclei, i.e., the lateral nucleus of the trapezoid body (e.g., Spangler et al., 1985; Banks & Smith, 1992; Cant & Hyson, 1992; Smith, 1995). Axons of MSO cells project within the medial aspects of the ipsilateral lateral lemniscus to terminate in the ipsilateral DLL (Rasmussen, 1946; Niemer & Cheng, 1949; Van Noort, 1969; Elverland, 1978; Glendenning et al., 1981; Henkel & Spangler, 1983) and ipsilateral IC (Monakow, 1890; Fuse, 1911; Niemer & Cheng, 1949; Roth et al., 1978; Adams, 1979; Brunso-Bechtold et al., 1981; Zook & Casseday, 1982b; Henkel & Spangler, 1983). Minor projection targets of MSO may include the ipsilateral VLL (Browner & Webster, 1975) and possibly the contralateral IC (Rasmussen, 1946; Adams, 1979; BrunsoBechtold et al., 1981; Kudo et al., 1988). However, the ipsilateral DLL and ipsilateral IC are by far the principal efferent targets of MSO. Thus, the MSO receives direct bilateral afferent innervation, while axons of its constituent cells ascend almost exclusively to the ipsilateral DLL and IC. The Lateral Superior Olive. Afferents to LSO are derived from two sources, the ipsilateral cochlear nucleus and the ipsilateral MTB. The origin of LSO's afferent innervation was a puzzle for quite some time. The projection from the ipsilateral cochlear nucleus was known quite early (e.g., Held, 1893; Ferrier & Turner, 1898), but the projection of MTB to LSO though suspected (Held, 1893), remained elusive until much later (Rasmussen, 1967). It could easily be demonstrated that destruction of one cochlear nucleus would result in massive degeneration throughout the ipsilateral LSO (e.g., Ferrier & Turner, 1898; Van Gehuchten, 1902; Woollard & Harpman, 1940), yet electrophysiologists could also demonstrate that LSO neurons responded to stimulation of both ears (e.g., Goldberg et al., 1963). Therefore, there must be a functional connection from the contralateral cochlear nucleus. Evidence of a functional connection came from Rasmussen's (1967) studies of trapezoid body transections. Selective sectioning of the trapezoid body medial to the MTB produced degeneration in MTB and MSO as expected, while cuts that included MTB produced degeneration in MSO and LSO. This was the proof needed to substantiate not only the previous Golgi observations (Held, 1893), but also his own experimental results (Rasmussen, 1946). It also proved correct the suspicions of Harrison and Warr (1962) who found that sectioning the trapezoid body immediately below LSO resulted in
Auditory Pathway
3 97
retrograde chromatolysis of principal cells in MTB. Thus the LSO receives binaural afferent innervation directly from the ipsilateral cochlear nucleus and in a two-step manner from the contralateral cochlear nucleus via the MTB. There are some details about the afferent innervation of LSO that warrant further discussion. The excitatory ipsilateral supply to LSO has been found to originate from the spherical cells of the AVCN (Harrison & Warr, 1962; Harrison & Irving, 1966; Osen, 1969b; Rouiller & Ryugo, 1984f Glendenning et al., 1985; Shneiderman & Henkel, 1987, Cant & Casseday, 1986), and are distributed to the LSO in a precise tonotopic arrangement (e.g., Warr, 1966, 1982). The MTB also projects to LSO in a precise tonotopic manner (Glendenning et al., 1985; Spangler et al., 1985), yet this innervation of LSO by MTB is not equally distributed to all parts of LSO. Electrophysiologically, the ipsilateral innervation of LSO is excitatory throughout the nucleus, while the contralateral inhibition (via the inhibitory interneurons of MTB) does not extend throughout the nucleus. Although the excitatory and inhibitory inputs to the same area of LSO are well matched in terms of best-frequency (e.g., Boudreau & Tsuchitani, 1968; Guinan et al., 1972b; Banks & Smith, 1992), low frequency LSO neurons, i.e., those of the lateral limb responding to 2000 Hz or less, are not inhibited by contralateral stimulation (Boudreau & Tsuchitani, 1968, 1970; Tsuchitani & Boudreau, 1969; Tsuchitani, 1977, 1982). This same point has been demonstrated anatomically in that MTB lesions result in anterograde degeneration that is more dense in the medial limb of LSO (high frequency) and very sparse degeneration in the lateral limb of LSO (low frequency; Goldberg, 1975; Glendenning et al., 1985). This graded projection of MTB to LSO has been verified by anterograde transport studies and by 14C-2deoxyglucose techniques (e.g., Glendenning et al., 1985). Thus the entire LSO receives direct excitatory innervation from the ipsilateral AVCN, and an indirect inhibitory innervation from the contralateral VCN via synapse in the MTB which ultimately exerts its influence more on high frequency units and then grades down to no influence on low frequency units. Two other lines of investigation directly address the point of MTB inhibition of LSO. Microiontophoretic application of glycine, a potent inhibitory neurotransmitter, effectively mimics the action of MTB on LSO neurons actively responding to stimulation of the ipsilateral ear. More importantly strychnine, a specific glycine receptor antagonist,
398
Hutson
blocks the inhibitory effects of binaural stimulation. Thus glycine is most likely the neurotransmitter used by MTB to inhibit neuronal activity in LSO (Moore & Caspary, 1983). Even more striking was the demonstration that 3H-strychnine (which specifically binds to glycine receptors) marked a receptor distribution gradient in LSO essentially identical to what would be predicted from the anatomical results, i.e., highest concentration of receptors in the medial limb and lowest concentration of receptors in the lateral limb (Sanes et al., 1985). Thus there is ample reason to believe that MTB inhibits LSO non-uniformly across the nucleus, and that the neurotransmitter acting at the MTB-LSO synapse is probably glycine (Wenthold et al., 1987; Glendenning & Baker, 1988; Bledsoe et al., 1990; Wu & Kelly, 1992). However, the low frequency lateral limb of LSO is not totally devoid of inhibitory inputs (Brownell et al., 1979; Wu & Kelly, 1994). The source of these inputs appear to arise from the ipsilateral lateral nucleus of the trapezoid body or from the ipsilateral cochlear nucleus itself (Glendenning et al., 1991). Neurons of LSO project axons into the lateral lemniscus of both sides in approximately equal numbers (Monakow, 1890; Yoshida, 1925; Poljak, 1925; Papez, 1930; Ohnisi, 1932; Niemer & Cheng, 1949; Stotler, 1953; Glendenning & Masterton, 1983). Efferent axons of LSO terminate predominantly in the IC (Roth et al., 1978; Brunso-Bechtold et al., 1981; Glendenning & Masterton, 1983; Casseday & Covey, 1983; Shneiderman & Henkel, 1987; Hutson, 1988; Vater et al., 1995), in DLL (Elverland, 1978; Glendenning et al., 1981; Shneiderman et al., 1988; Hutson, 1988; Vater et al., 1995), and only meagerly in VLL (Browner and Webster, 1975). An important feature of the efferent projections of LSO is that they are not random, instead specific portions of LSO project either ipsilaterally or contralaterally. The high frequency medial limb projects contralaterally, crossing the midline above the trapezoid body to enter the contralateral lateral lemniscus medially, while the low frequency lateral limb projects to the ipsilateral lateral lemniscus ascending in both its medial and lateral aspects (Van Noort, 1969; Glendenning et al., 1981). Cells of LSO representing intermediate frequencies project to either lateral lemniscus, but very few project bilaterally (Glendenning & Masterton, 1983). Nuclei of the Lateral Lemniscus. The VLL receives its afferent supply almost exclusively from the contralateral VCN (Glendenning et al., 1981; Warr, 1982; Vater & Feng, 1990). One exception to this is a small area of the ventral division of VLL that receives, apparently
Auditory Pathway
3 99
exclusive, innervation from the ipsilateral VCN (Glendenning et al., 1981; Warr, 1972). Other minor projections to VLL arise from both the ipsilateral and contralateral superior olivary complex, the most substantial of these originate from the ipsilateral MTB (Lewy, 1909; Zook & Casseday, 1979, Glendenning et al., 1981, Spangler et al., 1985). The ILL (or dorsal division of VLL), while receiving its major afferent supply from the contralateral VCN, also receives a substantial projection from the ipsilateral MTB. This is the heaviest MTB projection to any structure other than LSO (Glendenning et al., 1981; Spangler et al., 1985). When the ILL is injected with retrogradely transported tracer substances, the majority of principal cells of MTB are labeled throughout its rostro-caudal length. Therefore it would seem likely that the same MTB cells that innervate the LSO also send a collateral to ILL or at least some portion of the VLL (Spangler et al., 1985). Despite small projections from LSO and MSO, the VLL and ILL can be collectively considered as monaural nuclei under the influence of the contralateral cochlear nucleus directly or indirectly via the MTB. The only exception is a very small area of the ventral VLL that is innervated exclusively by the ipsilateral VCN, and therefore also monaural in nature (Warr, 1982, Glendenning et al., 1985, Spangler et al., 1985). Neurons of VLL, together with ILL, project principally to the IC (Zook & Casseday, 1979, 1982b; Brunso-Bechtold et al., 1981, Whitley & Henkel, 1984; Hutson, 1988; Glendenning et al., 1990; Hutson et al., 1991), though there are some meager VLL fiber terminations in the ILL and DLL, again probably collaterals of fibers en route to IC (Kudo, 1981; Whitley & Henkel, 1984). A tiny, yet demonstrable projection of ILL is to the medial division of the ipsilateral and contralateral medial geniculate body (Papez, 1929a,b; Kudo, 1981; Whitley & Henkel, 1984; Aitkin & Phillips, 1984a, Hutson, 1988; Hutson et al., 1991). These few axons that bypass the IC have been termed the central acoustic tract and travel medial to the brachium of the inferior colliculus (Ramon y Cajal, 1909; Papez, 1929a,b; Whitley & Henkel, 1984, Hutson et al., 1991). In contrast to VLL, the DLL receives bilateral innervation through the convergence of a variety of inputs, many of which are presumed to be collaterals of fibers destined for IC (Warr, 1966; Glcndenning et al., 1981). The most meager of the afferents to DLL arise from the ipsilateral MTB and the cochlear nucleus of both sides (Held, 1893; Kolliker, 1896; Fernandez & Karapas, 1967; Glendenning et al., 1981; Van Noort, 1969, Shneiderman et al., 1988). The few fibers from the
400
Hutson
contralateral DCN enter DLL medially (Fernandez & Karapas, 1967), while those of the ipsilateral and contralateral ventral cochlear nucleus enter DLL from its lateral edge (Warr, 1966; Van Noort, 1969; Browner & Webster, 1975). The contribution from the contralateral ventral cochlear nucleus is somewhat more substantial than the ipsilateral VCN (Glendenning et al., 1981; Shneiderman et al., 1988). The major afferents to DLL originate in the ipsilateral LSO and MSO, contralateral LSO and DLL (Monakow, 1890; Held, 1893; Glendenning et al., 1981; Woollard & Harpman, 1940; Barnes et al., 1943; Elverland, 1978; Zook & Casseday, 1979; Shneiderman & Henkel, 1987; Hutson, 1988; Hutson et al., 1991). Of all the afferents to DLL, the most prominent arise from the ipsilateral superior olivary complex (Glendenning et al., 1981). The principal efferent targets of DLL are the ipsilateral IC, contralateral IC and DLL (Kolliker, 1896; Woollard & Harpman, 1940; Kudo, 1981; Brunso-Bechtold et al., 1981; Shneiderman et al., 1988; Hutson, 1988; Hutson et al., 1991). To reach the ipsilateral IC, DLL efferent fibers ascend directly within the lateral lemniscus, while efferent fibers destined to innervate contralateral structures course through the commissure of Probst (Probst, 1902) bound for the opposite DLL and the contralateral IC (Held, 1893; Van Gehuchten, 1903; Stokes, 1912; Castaldi, 1926; Ariens Kappers et al., 1936; Woollard & Harpman, 1940; Goldberg & Moore, 1967; Brunso-Bechtold et al., 1981; Glendenning et al., 1981; Kudo, 1981; Whitley & Henkel, 1984; Hutson et al., 1991). The Inferior Colliculus. As mentioned above, very few fibers of the ascending auditory system bypass the IC, and as such the IC may be considered an obligatory synapse for all ascending afferents (Ferrier & Turner, 1898; Aitkin & Phillips, 1984a; Rouiller & Ribaupierre, 1985; Hutson et al., 1991; Hutson et al., 1993). The discussion so far has revealed the multitude of projections to the IC, therefore a brief list of these afferents should suffice. The major afferents originate in the contralateral cochlear nucleus, contralateral and ipsilateral LSO, ipsilateral MSO, ipsilateral VLL, ipsilateral and contralateral DLL (e.g., Adams, 1979; Brunso-Bechtold et al., 1981). Lesser afferents arise from the ipsilateral cochlear nucleus, and ipsilateral MTB (Brunso-Bechtold & Thompson, 1976; Adams, 1979; Brunso-Bechtold et al., 1981; Warr, 1982; Spangler et al, 1985).
Auditory Pathway
401
Neurons of the IC have two major efferent routes, the brachium of the inferior colliculus and the commissure of the inferior colliculus. The main efferent pathway of the IC is through its brachium to reach the ipsilateral medial geniculate body (MG). Terminals of IC fibers can be found in all divisions of the MG, but they are most concentrated in the ventral division of MG, which in turn projects to primary auditory cortex (Monakow, 1895; Woollard & Harpman, 1940; Anderson et al., 1980; Moore & Goldberg, 1963; Winer et al., 1977; Kudo & Niimi, 1978, 1980; Ravizza & Belmore, 1978; Morel & Imig, 1987; Hutson, 1988; Hutson et al., 1991, 1993). Axons projecting through the commissure of the inferior colliculus terminate in the dorsomedial division of the contralateral inferior colliculus, probably to participate in the descending system, though a modest number of fibers continue on to reach the MG of the opposite side (Diamond et al., 1969; Kudo & Niimi, 1980; Aitkin & Phillips, 1984b; Hutson, 1988; Hutson et al., 1991, 1993). Commissures. In the preceding sections the various commissural fibers of the auditory system were discussed. Since these are the points where the two sides of the auditory system interconnect, i.e., the points of decussation, they are of obvious importance in any attempt to separate the ipsilateral pathways from the contralateral pathways. To reiterate, the major commissures are the three acoustic striae, the commissure of Probst, and the commissure of the IC (see Hutson et al., 1991). There is of course, the corpus callosum, interconnecting the auditory cortices of the two hemispheres (e.g., Diamond et al., 1968; Imig & Brugge, 1978; Code &Winer, 1986).
The Ipsilateral Auditory Pathway From the preceding discussion it can be seen that there are many routes ipsilateral information can travel. This can be direct, e.g., cochlear nucleus to ipsilateral IC, or indirect, e.g., cochlear nucleus to contralateral DLL and then back to ipsilateral IC through the commissure of Probst. For the remaining discussion of ipsilateral pathways, emphasis will be given to those pathways that are of direct origin. That is to say, nuclei whose afferent innervation and efferent projections do not cross the midline. This will be presented in terms of how the ascending auditory system would be affected by severing the decussating (contralateral) pathways at the midline (see Figure 4).
402
Hutson
The Cochlear Nucleus. Degeneration and axonal transport studies have reported a direct ipsilateral projection from DCN to IC, yet each commented on the sparsity of the pathway (Baginski, 1886; Strominger, 1973; Roth et al., 1978; Adams, 1979; Brunso-Bechtold et al., 1981; Nordeen et al., 1983). Therefore it is the VCN that gives rise to major afferents of the ascending ipsilateral auditory system. The VCN distributes its fibers to directly innervate the ipsilateral (laterally directed) dendrites of MSO, the entire LSO, a small portion of ventral VLL, the DLL, and finally the IC (e.g., Ferrier & Turner, 1898; Woollard & Harpman, 1940; Warr, 1966, 1969, 1972, 1982; Van Noort, 1969; Strominger & Strominger, 1971; Harrison & Howe, 1974; Strominger et al., 1977; Brunso-Bechtold et al., 1981; Glendenning et al., 1981, 1985, 1991, 1992; Ryugo et al., 1981; Moore & Kitzes, 1985; Cant & Casseday, 1986; Oliver, 1987). Although cochlear nucleus projections to the ipsilateral superior olivary complex arise from all regions of the VCN (high, middle, and low frequency), the projections to the ipsilateral IC originate primarily, though not exclusively, from ventral (low frequency) regions of both the DCN and VCN (Nordeen et al., 1983; Oliver, 1984, 1987). Therefore, the cochlear nucleus projections to the ipsilateral IC may have a bias toward low frequencies. The Superior Olivary Complex. It is important to note here that the influence of MTB would be completely eliminated, due to its exclusive innervation by the contralateral VCN. Carrying this one step further, MTB's efferents to ipsilateral nuclei (LSO, IC, and nuclei of the lateral lemniscus) would also be eliminated from consideration in the ipsilateral pathway. Therefore, although MTB projections are to ipsilatera! structures, its own afferent source lies in the contralateral VCN. By the restriction posed above (no axons crossing midline), the MTB is for all practical purposes de-afferented and a non-participant in the ascending ipsilateral system. The MSO and LSO would continue to be innervated by the spherical cells of AVCN, but both MSO and LSO would function without their contralateral inputs. MSO would lack innervation of its medially directed dendrites, LSO would lack contralateral inhibitory innervation. MSO would retain all of its efferent projections, i.e., a tiny projection to VLL, and huge projections to DLL and IC. In contrast LSO, under the no crossing the midline restriction, would have its efferents as well as afferents reduced by one-half. Thus, neurons medially located in LSO
Auditory Pathway
403
IC
! !
I
,
VLL
ti
VCN
Figure 4. Summary diagram of the major connections of the ipsilateral auditory pathway. See text for details. Abbreviations as in Figure 3.
(high frequency, and projecting across the midline) would be eliminated, though neurons located laterally in LSO (low frequency) would
404
Hutson
still participate in the ascending ipsilateral pathway. LSO, like MSO, would continue to supply a sparse projection to the VLL, and more substantial projections to DLL and IC. In terms of frequency representation, the ipsilateral pathways ascending from the superior olivary complex originate chiefly from neurons which respond best at middle to low frequency stimulation (i.e., MSO, lateral LSO). High frequency ipsilateral projections from the superior olivary complex would then originate primarily from neurons located ventrally in MSO (see Figure 2). The Nuclei of the Lateral Lemniscus. The major afferent sources of VLL and ILL are the contralateral cochlear nucleus and ipsilateral MTB, but as concerns the ipsilateral system, these nuclei would be almost totally eliminated following the same argument as for MTB. The only exception would be the relatively small ipsilateral VCN projection to ventral VLL, the remaining ipsilateral afferents from MSO and LSO being almost negligible by comparison (Browner & Webster, 1975, Glendenning et al., 1981). The VLL would have a minor, essentially nonexistent, role in the ipsilateral pathway contributing only its minute projections to DLL and IC from the small area of ventral VLL which receives ipsilateral cochlear nucleus inputs. On the other hand, DLL would still retain many of its afferent connections, i.e. from the cochlear nucleus, MSO, and LSO. It would lose afferents from the contralateral cochlear nucleus and contralateral DLL, but more importantly it would lose inputs from the contralateral LSO. Thus DLL would remain a highly innervated structure, receiving direct afferents from the cochlear nucleus and also from the superior olivary complex. DLL efferents would be reduced to its projections to the ipsilateral IC. The Inferior Colliculus. As can be seen, the IC would be devoid of all inputs derived from contralateral sources both directly, i.e. cochlear nucleus, and indirectly, e.g., ipsilateral MTB, VLL, contralateral LSO and DLL. It would receive those afferents of the ipsilateral pathway described above. The major efferents of IC would be, as always, to the ventral division of the MG. The connections through the commissure of the IC would be eliminated, even though these are normally probably not of great importance to the ascending auditory system. Thus, the connections above the IC, while lacking contralaterally derived information, would not be significantly reduced in terms of their projection
Auditory Pathway
405
fiber compliment since there are no more opportunities for their axons to cross over to the opposite side of the brain until reaching cortex. In summary, the most substantial ipsilateral pathways remaining after a midline section of the brainstem would be as shown in Figure 4. The largest components being cochlear nucleus to MSO, LSO, DLL, and IC; MSO and LSO to DLL and IC; and DLL to IC. Minor components being cochlear nucleus to VLL, and VLL to DLL and IC. Having now reviewed properties of the ascending auditory system and defined the ipsilateral components, the following sections of this paper will discuss the evidence in support of potential functions of this ipsilateral pathway. II. Role of Ipsilateral Pathway in Behavior Clearly, the essential connections of the auditory system had been described by the turn of the century (see Flechsig, 1876; Monakow, 1890; Held, 1891; Van Gehuchten, 1902, 1903; Ramon y Cajal, 1909). In particular, anatomists knew of the decussations that take place in the medulla (the acoustic striae and contralateral projections of LSO), as well as those at pontine levels (commissure of Probst and commissure of the inferior colliculus). Further, it had been shown that the medullary decussations were the primary source of any contralateral projections to higher nervous structures. The anatomical course of the auditory pathway was known to originate in the cochlear nucleus and ultimately connect with the auditory cortex of the temporal lobe. However, the physiological properties of auditory elements, other than the cochlea and auditory cortex were not known (e.g., Ferrier, 1890). This lack of physiological information did not last long, and again prior to the turn of the century physiologists were well underway into investigating the properties of the deep seated auditory structures. Some of the earliest attempts to experimentally derive the functional importance of the different auditory pathways were conducted by Hammerschlag (1899, 1901). On the basis of the following types of experiment, Hammerschlag concluded that the dorsal pathways (dorsal striae) differed from the ventral pathways (trapezoid body). Cuts were made at various positions in the medulla of cats or dogs, and the immediate effects on the tensor tympany muscle noted. From this it was concluded that only cuts of the ventral pathways had an effect on this reflex, and therefore the "hearing" pathway was located in the ventral bundle of the trapezoid body. Furthermore, it was claimed that this
406
Hutson
trigeminal reflex was mediated through the MSO. Even though this line of evidence is reflexive in nature and not a true test of "heating" per se, it was probably the first attempt to delineate the possible functions of the acoustic pathways. The question of the relative contributions of the medullary decussations to hearing was re-opened by Winkler (1911), who found that severing the dorsally decussating pathways (dorsal striae) rendered his two experimental cats unresponsive to sound. On this evidence, Winkler concluded that the dorsal decussations carried the largest number of auditory impulses, and therefore it was the secondary fibers arising from DCN that are ultimately responsible for the transmission of auditory information to cortex. This interpretation achieved wide acceptance, and can be found in textbooks well into the mid-1950's (e.g., Strong & Elwyn, 1953). Contrary to Winkler's findings were those of Kreidl (1914). Kreidl's experiments were important for two reasons, they were the first experimental demonstration of a functional ipsilateral auditory system and they were the only clear cut demonstration of the ipsilateral auditory system for 30 years, enduring into the mid-1940's as the only accepted evidence for the "possible" existence of ipsilateral pathways (e.g., Ranson & Clark, 1947). Unfortunately, reference to Kreidl's contribution to our knowledge of the auditory system disappeared after the late-1940's (e.g., Ranson & Clark, 1947). For historical reasons, and for the fact that they still remain a very convincing example of a functioning ipsilateral auditory system, Kreidl's experiments will be outlined in some detail. Prior to publication of Winkler's (1911) results, Kreidl had been involved in experiments on auditory cortex and medullary respiratory centers. During his respiratory research, he had divided the dorsal medulla mid-sagittally in two animals (dogs) and found it convenient to simultaneously test the hearing capacity of these animals. The animals survived 3 and 7 weeks respectively, and "in the time between surgery and death they reacted to being called by name and showed no difference in heating ability from normal control animals." The brainstem of these animals, subjected to Marchi staining, revealed that the medulla had been divided at the midline to about 1/3 of its normal thickness and that the dorsal striae were degenerated, while the trapezoid body was undamaged. Noting the difference between his casual observations and those of Winkler (1911), and being of the opinion that the dorsal
Auditory Pathway
407
pathways were not the important pathways for audition, he undertook a more systematic investigation of the secondary auditory pathways. From a dorsal exposure, mid-sagittal sections of the medulla were made in six more dogs. One dog again received a cut that only sectioned the dorsal striae, leaving the trapezoid body intact. One received a section that left the caudal-most fibers of the trapezoid body intact while severing the middle and rostral trapezoid body and the entire dorsal striae. The remaining four dogs had the dorsal striae and trapezoid body completely transected. In each case, the animals "hearing hardly differs from a normal hearing dog", e.g., "reacts to sounds, barks when hears other dogs barking, wags tail and approaches when called by name, when lying down raises head when called by name." Clearly, these forms of observational evaluations are subjective at best and are far from being well controlled auditory tasks. Indeed no description of the testing situation is given. Yet these results are hard to dismiss even on the grounds of being uncontrolled observations. Moreover, their credibility is enhanced by the results from an additional experiment. In this case, following a 14 day post-operative survival period, a dog demonstrated "hearing" reactions similar to those described above. In addition, this dog was tested in a Pavlovian' type experiment. Here, when put in a room, a knock on the door from the outside would initiate barking in this animal. Thus a crude unconditioned stimulus, the knock, would elicit an unconditioned response, barking. Histological examination of this animals brainstem revealed that it had received the most extensive longitudinal cut of all the animals. The cut not only completely severed the dorsal striae and trapezoid body, but also a large part of the commissure of Probst. Thus, despite complete transection of the secondary auditory pathways that cross the midline, these animals were not rendered deaf or unresponsive to sounds. Since the anatomical pathways are essentially identical in dogs and cats, Kreidl attributes the discrepancy between his results and Winkler's as being due to post-surgical trauma. Winkler's cats survived only 5 and 8 days, whereas Kreidl found that it required 5-8 days of post-operative recovery before he could elicit clear reactions to sound, suggesting that if the cats had survived longer they could have been shown capable of hearing. During the initial post-operative recovery period there was no reaction to sound, but hearing returned over time. This observation gives further credence to Kreidl's results.
408
Hutson
Kreidl also completely transected both the dorsal striae and the trapezoid body in one Rhesus monkey. Again, in this animal hearing returned after approximately one week. In conclusion, Kreidl states: "One can cut the entire dorsal and ventral paths at the midline and find no difference from normal animals. The anatomy agrees that there are secondary crossed and uncrossed paths...behavior shows the uncrossed paths are important for heating. Under normal conditions crossed and uncrossed paths are used for transmitting hearing impulses. But in what way these pathways differ needs more experimentation." Even though these conclusions predict the course of over 80 years of auditory research, Kreidl's contributions have unfortunately been overlooked. The question of the relative importance of the ipsilateral (uncrossed) pathways versus the contralateral (crossed) pathways was re-examined with more quantitative methods by Brogden et al., (1936). The rationale for the experiment was as follows. Information derived from each cochlea eventually courses through the central pathways to be distributed to the MG of both sides, and from there to auditory cortex. Thus, information from the left cochlea will reach both the left and right auditory cortex. Employing a conditioned response paradigm, absolute auditory thresholds were obtained from cats at three test frequencies of 125, 1000, and 8000 Hz. The experimental procedure (outlined in Figure 5), consisted of systematic destruction of either the cochlea or auditory cortex to isolate the crossed or uncrossed pathways. Thus the degree of hearing loss, resulting from interruption of various components of the auditory system, could be determined. Destruction of one cochlea produced an average hearing loss of 3-4 dB at each test frequency, destruction of both cochleas producing total deafness (e.g., no response to tones 125 dB above pre-operative threshold). Unilateral ablation of cortex resulted in a threshold increase of 3-5 dB (or 3-5 dB hearing loss), similar to unilateral cochlea destruction. To ascertain the relative contribution of the crossed vs. uncrossed pathways, unilateral cortex ablation was combined with either ipsilateral or contralateral cochlear destruction. Contralateral cochlear destruction (eliminating uncrossed pathways) increased thresholds an average of 15 dB, and ipsilateral cochlear destruction (eliminating crossed pathways) increased thresholds an average of 13-14 dB. It was concluded that, at least for absolute thresholds, the ipsilateral pathways were functionally equivalent to contralateral pathways. Similar results were obtained for dogs (Mettler et al., 1934).
Auditory Pathway
409
AC Ablation
,
cb
A Threshold 3-4dB
MG 1+2
IC
. . . . . . .
I
. . . . . .
!
II
3
> 125dB 3-4dB
1+3
13 - 14 dB
2+3
15 dB
"11
I ~
=1 L .......
',
',
I
I
I Coch
I'
:L . . . . __2j
i
I
Figure 5. Summary diagram of the experimental paradigm employed by Brogden et al., (1936) to demonstrate the functional equivalence of ipsilateral and contralateral pathways to auditory cortex (AC). Absolute auditory thresholds were obtained before and after selective ablation of auditory structures. Note that following ablation 1+3 hearing was dependent upon an ipsilateral pathway, while after ablation 2+3 hearing was dependent upon a contralateral pathway. In either case hearing loss, as measured by change in absolute threshold (A threshold), was minor and essentially equivalent for the two pathways. Coch, cochlea; IC, inferior colliculus; MG, medial geniculate body.
A more intensive investigation, again using a conditioned response paradigm to test absolute intensity thresholds was conducted by Kryter and Ades (1943). Absolute thresholds were obtained for cats at test frequencies of 125, 1000, and 8000 Hz. By pairing a unilateral lateral lemniscus section with destruction of the cochlea on the same side, a single ipsilateral pathway could effectively be isolated, and in this case thresholds remained within normal limits. Isolation of a single contralateral pathway by pairing a lateral lemniscus section with destruction of the contralateral cochlea produced a slight hearing loss on the order of 11 dB. Interestingly, this loss was much greater for low frequencies than for high frequencies (e.g., 125 Hz, 18.7 dB; 1000 Hz, 10 dB; 8000 Hz, 5.6 dB). This experiment shows the conditioned
410
Hutson
response to be resilient to any manipulation that reduces the auditory pathway to only its ipsilateral or contralateral components. Bilateral destruction of the cochleas resulted in a permanent loss of the conditioned response. Similarly, Dixon (1973) also isolated a single ipsilateral or contralateral pathway by combined lateral lemniscus section and cochlear destruction, and measured absolute pure-tone thresholds for 250, 1000, 4000, 16000, and 32000 Hz. Again the ipsilateral pathway was equally sensitive as the contralateral pathway. Neither manipulation resulted in significant changes in pure-tone thresholds, although bilateral section of the lateral lemniscus did cause an increase in threshold of 80 dB or more. Masterton et al. (1992) repeated these experiments (unilateral cochlear destruction combined with ipsilateral or contralateral lateral lemniscus section) using a conditioned avoidance procedure and a battery of 26 auditory detection tasks. Again, in no case was heating or the conditioned response abolished for any task, regardless of whether the ascending system was reduced to a single ipsilateral or a single contralateral pathway. Although the results suggest that on 24 of the 26 tasks, the ipsilateral pathway makes no unique contribution to the ascending auditory system, the results do suggest that the ipsilateral pathway has an advantage over the contralateral pathway for detection of low frequency tones (below 4000 Hz) where thresholds remain near normal, and also for detection of a low frequency amplitude modulated tone (500 Hz). When only a contralateral pathway remained the average hearing loss, across all frequencies tested, was 18 dB. When only an ipsilateral pathway remained, the loss in sensitivity for frequencies above 4000 Hz averaged 17 dB. Surprisingly similar results have been reported for humans after surgical transection of one lateral lemniscus (e.g., Walker, 1942; Drake & McKenzie, 1953). A udiograms from these patients show a diminished ability to detect high frequency tones delivered to the contralateral ear, i.e., to the intact ipsilateral pathway. Even at the level of auditory cortex, unilateral ablation results in a significantly elevated threshold for detecting high frequency tones in the ear contralateral to the ablation, that is, the ear ipsilateral to the undamaged cortex (Heffner & Heffner, 1989). Thus there is relatively good agreement across studies that the ipsilateral pathway has a lower threshold for detecting low frequency tones, while deficient for detection of high frequency tones.
Auditory Pathway
41 1
Changing the behavioral task from absolute threshold detection to active conditioned avoidance of an auditory stimulus, cats can still maintain avoidance behavior with a single, isolated ipsilateral pathway (Jane et al., 1965). For example, in one case despite total transection of the medullary decussations (DAS, IAS, trapezoid body, contralateral projections from LSO) and the commissure of Probst, plus interruption of one lateral lemniscus and partial transection of the commissure of the inferior colliculus, this animal did not lose the learned avoidance response. Further, this manipulation also caused no detectable decrement in threshold to a 300 Hz tone. This one example shows that a single functional ipsilateral pathway can maintain not only normal thresholds, but also avoidance behavior. The behavioral studies reviewed above essentially dealt with the question of whether or not can you hear with your ipsilateral pathways. The answer is obviously you can. Further, it can be stated that not only can ipsilateral pathways maintain hearing, but also conditioned responses to auditory stimuli, conditioned avoidance to auditory stimuli, and intensity thresholds. Given that the ipsilateral pathway can maintain these fundamental aspects of audition, the question now arises as to whether it is capable of supporting more demanding auditory tasks. In other words, is there anything that the ipsilateral pathways can n o t do? Investigations e x a m i n i n g the contributions of the auditory commissures to sound localization discovered that ipsilateral pathways alone can not sustain an animals ability to localize sounds in space (e.g., Moore et al, 1974; Casseday & Neff, 197:5). While transection of the corpus callosum or the commissure of the inferior colliculus have no effect on sound localization (Moore et al, 1974), section of the medullary decussations (dorsal striae, trapezoid body, decussating fibers from the LSO) produces a profound deficit in sound localization (Moore et al, 1974; Casseday & Neff, 1975). Furthermore, transection of the medullary decussations not only destroys the pre-operative localization habit, but also renders these animals unable to re-acquire the habit despite massive training. Nonetheless, these animals displayed normal intensity thresholds (at 250, 1000, and 8000 Hz) and were capable of learning auditory pattern discriminations (Casseday & Neff, 1975). Thus ipsilateral pathways alone are sufficient to sustain a great many auditory functions, but sound localization is not one of them. Accepting this conclusion, one would reasonably suspect that sound localization is a function that relies upon the contralateral pathways.
412
Hutson
Turning to the sound localization literature to verify this suspicion leads to quite a different conclusion. This is a difficult question to answer directly, owing to the paucity of investigations specifically addressing this question (e.g., Neff & Casseday, 1977). However, many points concerning the role of contralateral pathways may be abstracted from the sound localization literature. To begin, subtotal transection of the medullary decussations (i.e., experiments severing only the trapezoid body or that destroy the region of the superior olivary complex) which leave the dorsal decussations intact (dorsal striae) produce deficits in sound localization equal in magnitude to total transection of all the medullary decussations. Therefore the contralateral pathways via the DAS and IAS apparently can not maintain sound localization behaviors, even in the presence of intact ipsilateral pathways (Masterton et al, 1967; Moore et al, 1974; Casseday & Neff, 1975). Thus, by elimination this leaves only the pathways originating in the VCN (which decussate in the trapezoid body) as the source of information usable for sound localization. Fibers emanating from VCN cross the midline to innervate the contralateral superior olivary complex, or continue on to enter the lateral lemniscus and terminate in the contralateral IC. Behavioral experiments manipulating this pathway demonstrate that these fibers, while necessary for accurate sound localization are not sufficient in and of themselves. Evidence for this conclusion results from unilateral ablations at various levels of the auditory system. Recall that once past the medulla (i.e. at the level of the lateral lemniscus) the central auditory pathways are essentially ipsilateral (lateral lemniscus to IC to MG to auditory cortex), though composed of fiber tracts transmitting information derived from bilateral, and therefore binaural, sources. Unilateral ablation of auditory structures anywhere from the lateral lemniscus to auditory cortex result in localization deficits confined to the contralateral auditory hemifield (Penfield & Evans, 1934; Wortis & Pfeffer, 1948; Sanchez-Longo & Forster, 1958; Strominger, 1969; Strominger & Oesterreich, 1970; Neff et al, 1975; Masterton et al, 1981; Jenkins & Masterton, 1982; Jenkins & Merzenich, 1984; Kavanagh & Kelly, 1987; Heffner et al., 1992; Poirier et al., 1994). In contrast, ablation of the auditory system anywhere from cochlea to superior olivary complex results in localization deficits within the ipsilateral auditory hemifield, or deficits in both the ipsilateral and contralateral hemifields (Masterton et al, 1967; Casseday & Neff, 1975;
Auditory Pathway
413
Casseday & Smoak, 1981; Jenkins & Masterton, 1982; Kavanagh & Kelly, 1992). Specifically, destruction of the ventral cochlear nucleus results in localization deficits ipsilateral to the side of the lesion (e.g., Casseday & Smoak, 1981) while ablation of the superior olivary complex results in predominantly bilateral localization deficits (e.g., Jenkins & Masterton, 1982; Kavanagh & Kelly, 1992). On the basis of these and other studies it is apparent that VCN fibers crossing the midline are necessary, but not sufficient for sound localization. Also implicated as vital for sound localization are the nuclei of the superior olivary complex, the point where fibers from both the ipsilateral and contralateral VCN first converge (Masterton et al., 1981; Jenkins & Masterton, 1982; Phillips & Brugge, 1985). Indeed, subtotal transection of the trapezoid body sparing the anterodorsal region (the area through which VCN fibers course to reach the contralateral MSO; Wart, 1966, 1982) results in sound localization deficits that are less severe or non-existent in comparison to total trapezoid body section (Casseday & Neff, 1975). From this it can be concluded that although the contralateral pathways coursing through the trapezoid body are of comparatively greater value than the ipsilateral pathways in terms of sound localization, ipsilateral pathways most certainly participate in sound localization. That accurate localization of sounds in space requires both ears is a recognized tenet of audiology (e.g., Rosenzweig, 1961; Phillips & Brugge, 1985). Interaction of information derived from the two ears can occur at many places along the central auditory pathway from cochlear nucleus to auditory cortex. The studies previously cited demonstrate the importance of the trapezoid body as a commissure interconnecting the ipsilateral with the contralateral pathways. Ablation of other potential sites of commissural interconnections, i.e., commissure of the inferior colliculus, or the corpus callosum, have no measurable effect on sound localization (Neff & Diamond, 1958; Moore et al, 1974), however, isolated transection of the commissure of Probst in rats does result in reduced localization acuity near the midline (Ito et al., 1995). Thus, the primary points of convergence between the ipsilateral and contralateral systems occur from the level of the superior olivary complex through the inferior colliculus. Ascending beyond the inferior colliculus, there is no evidence for binaural convergence taking place in any other structure (e.g., Phillips & Brugge, 1985).
414
Hutson
The results from the behavioral investigations of sound localization, taken together show that: (1) midline sections that reduce fibers ascending in the lateral lemniscus to only ipsilateral pathways or ipsilateral pathways plus the dorsal striae, can not support sound localization; (2) destruction of one lateral lemniscus, leaving the other lateral lemniscus intact (with its complement of ipsilateral and contralateral pathway fibers), can maintain sound localization at least in the hemifield opposite the intact lateral lemniscus; (3) unilateral ablation of structures above the lateral lemniscus results in the same syndrome; and (4) on the basis of anatomy and physiology the IC is the last point where clearly separable ipsilateral and contralateral pathways interact. III. Evoked Potential Studies
Another line of inquiry that is useful in supporting the functional importance of the ipsilateral pathways comes from evoked potential studies. Early evoked potential studies demonstrated that fibers coursing through the lateral lemniscus were responsive to stimulation of either the ipsilateral or the contralateral ear (Davis & Saul, 1931, Saul & Davis, 1932). Subsequent investigations used the evoked potential method to trace functional auditory pathways through the brain by an "ablationevoked potential" paradigm (e.g., Ades, 1944; Ades & Brookhart, 1950; Jungert, 1958; Fullerton & Kiang, 1990; Kelly & Li, 1997). Two experiments performed by Ades and Brookhart (1950) are of particular interest in terms of verifying the anatomical and behavioral results. In the first experiment electrodes were placed over both the right and the left auditory cortex of cats, and potentials recorded in response to click stimulation. Then the left auditory nerve was severed, without obvious effect to either cortical evoked potential. A second cut transecting the commissure of the inferior colliculus also had little or no effect on either cortical evoked potential. Finally, section of the decussating fibers in the medulla (dorsal striae, trapezoid body, decussating fibers of LSO) totally eliminated the left cortical evoked potential, while the right cortical evoked potential was only reduced in amplitude. In this experiment after transecting the left auditory nerve, the commissure of the inferior colliculus, and the medullary decussations, only the ipsilateral pathways from the right cochlear nucleus to right auditory cortex remained and they were capable of eliciting a strong cortical evoked potential.
Auditory Pathway
415
The second experiment was similar to the one just described except that the recording electrodes were placed on the two inferior colliculi. Section of the left auditory nerve and commissure of the inferior colliculus produced only minor changes in the evoked potentials, whereas medullary transection eliminated the left evoked potential, and again only reduced the amplitude of the right evoked potential. Thus ipsilateral pathways alone can elicit a strong collicular evoked potential. Other auditory evoked potential investigations demonstrated the ipsilateral and contralateral pathways to be of approximately equal value, and that binaural interactions occur at the superior olivary complex, the nuclei of the lateral lemniscus, and the IC (Ades, 1944; Rosenzweig & Amon, 1955; Rosenzweig & Sutton, 1958; Galambos et al., 1959; Kelly & Li, 1997). Further, transection of the commissure of the inferior colliculus does not eliminate or in any way interfere with the interactions (Rosenzweig & Wyers, 1955). From all the lines of evidence discussed thus far, there is certainly abundant grounds for accepting not only the existence of ipsilateral pathways but also their functional significance. Ablation-behavior experiments indicate that ipsilateral and contralateral pathways can operate alone, although they normally interact with one another. This interaction appears to be necessary for accurate localization of sounds in space. The following section of this paper will explore some of the of the physiological properties of the ipsilateral (and contralateral) pathways at the last place where they can be separately distinguished, the IC. IV. Role of Ascending Pathways in the Physiology of the IC Before drawing conclusions as to the nature of the ipsilateral pathways, this section will discuss the physiological influence of the ascending pathways on neurons of the IC. The IC is not only a major relay center for the ascending auditory system but it is also the last opportunity for direct binaural interactions to occur (Ferrier & Turner, 1898; Semple & Aitkin, 1981; Aitkin & Phillips, 1984a; Phillips & Brugge, 1985; Hutson et al, 1991). The IC integrates acoustic information arising from the cochlear nucleus, superior olivary complex, and nuclei of the lateral lemniscus. Although the IC receives pre-processed binaural information from the superior olivary complex and DLL, there is clear anatomical, electrophysiological, and behavioral evidence indicating further binaural processing at the IC (Chan & Y in, 1984;
416
Hutson
Kuwada et al., 1984; Phillips & Brugge, 1985; Carney & Yin, 1989; Yin & Chart, 1990; Batra et al., 1993; Covey et al., 1996). Since the basic physiological responses have been addressed in a previous section, the aim here will be to focus on synaptic mechanisms and synaptic relationships among the ipsilateral and contralateral inputs to IC neurons. Briefly, it will be recalled that IC neuronal responses may be typed as EE, IE, and OE based on their binaural response properties, which reflects their laterality preference from afferent inputs (e.g., Semple & Aitkin, 1979). Additional descriptive terms for IC unit discharge characteristics include "onset" or phasic and "sustained" or tonic (e.g., Erulkar, 1959; Rose et al., 1963, 1966; Hind et al., 1963; Kuwada et al., 1984). Units displaying an onset response show spike activity at the initiation of the stimulus with no further response, while sustained type units respond throughout the stimulus presentation. Onset type neurons predominate over sustained response units, and have been described as the result of intricate excitatory-inhibitory events (Hind et al., 1963). Onset type responses are of particular interest since they are found in greater numbers at the IC and higher levels, while at lower levels, e.g., cochlear nucleus or superior olivary complex, sustained responses predominate (e.g., Rose et al., 1963). The remainder of this section will be devoted to the examination of the excitatory-inhibitory interplay at the IC. Inhibition of IC Neurons. Intracellular recordings from IC neurons by Nelson and Erulkar (1963) demonstrated that acoustic stimulation could elicit depolarization, hyperpolarization, or both in combination. Furthermore, these experiments revealed the probable synaptic mechanisms underlying the onset type of response. Many IC neurons respond to auditory stimulation with an initial depolarizing excitatory potential eliciting spike discharges which are followed by a powerful inhibitory input cessating the discharge. Additional experiments were undertaken to examine the properties of the inhibitory inputs by current injection through the recording electrode in the presence or absence of acoustic stimuli. The results suggest that active inhibition may come about by two synaptic mechanisms. The most convincing evidence is for post-synaptic inhibition, arising either directly from the ascending pathways or perhaps mediated by local inhibitory interneurons. The other form of observed inhibition appeared to act by a reduction of background excitation (i.e., reduced excitatory post-synaptic potentials), evidence suggestive of the involvement of pre-synaptic mechanisms.
Auditory Pathway
417
Ultrastructural evidence from normal material supports the interpretation of post-synaptic mechanisms. Electron microscopic studies have demonstrated the presence, in significant numbers, of terminal boutons containing flattened vesicles apposed to IC neurons, indicative of inhibitory synapses (Rockel & Jones, 1973b). Unfortunately, these authors could not determine the origin of these boutons, but suggested the possibility that they could arise intrinsically from multipolar cells. Although not specifically discussed, this same electron microscopic study also yields some ultrastructural evidence for considering the presence of pre-synaptic inhibition. Examining the photomicrographs, one is struck by the abundance of terminals apposed to IC cell bodies and dendrites. In many instances astrocytic "fingers" can be seen separating the boutons, but in many other cases no such "fingers" are visible, in addition there appears to be thickening of the terminal membranes when they are in close proximity. Furthermore, apparent "stacking" of terminals can be seen, all of which are suggestive of, but certainly not proof of pre-synaptic interactions or synaptic modulation. Neurochemical and pharmacological studies have also addressed the question of inhibitory processes occurring at the level of the IC, again evidence obtained by these methods indicate that both pre- and postsynaptic inhibitory mechanisms are characteristic of inputs to the IC. Of the long list of putative neurotransmitter substances, GABA (gammaaminobutyric acid) and glycine are considered to be the best candidates as inhibitory neurotransmitters (e.g., Curtis, 1968; Johnston, 1976; McGeer & McGeer, 1981; Fagg & Foster, 1983; Davidoff & Hackman, 1985). Moreover, GABA has been associated with inhibitory mechanisms presumably pre-synaptic in nature, while glycine has been considered as the best candidate for post-synaptic mechanisms (e.g., McGeer & McGeer, 1981; Moore & Caspary, 1983; Davidoff & Hackman, 1985; Caspary et al., 1985; Bormann, 1988; Gage, 1992). Microinjection or iontophoretic application of these compounds or their antagonists have potent inhibitory or disinhibitory actions respectively on IC neurons (Watanabe & Simada, 1971, 1973; Faingold et al., 1989a, 1991). Both GABA and glycine depress spontaneous activity and stimulus-induced responses (Watanabe & Simada, 1973; Faingold et al., 1991). Furthermore, receptor binding studies have demonstrated that the auditory system is characterized by high levels of inhibition. The evidence here is based on the relative abundance of sites marked by various ligands used to probe for inhibitory neurotransmitter receptors
418
Hutson
(e.g., Baker et al., 1986; Glendenning & Baker, 1988; Glendenning et al., 1991, 1992) Considerable evidence supporting GABA as an inhibitory neurotransmitter at the IC has accumulated from pharmacological and neurochemical studies. When GABA is exogenously applied to IC neurons the inhibition produced is rapid in onset and rapid in offset. Benzodiazapine, which acts at GABA receptor sites, and nipecotic acid, a GABA reuptake inhibitor, both inhibit IC neuronal discharge and augment the action of GABA applied simultaneously. Bicuculline, a GABA antagonist, blocks inhibitory effects of exogenously applied GABA and blocks binaural inhibition at the IC. Baclofen, a possible GABA agonist, inhibits IC neuronal firing but its temporal action does not follow the same time course as GABA (Watanabe & Simada, 1973; Johnston, 1976; McGeer & McGeer, 1981; Faingold et al., 1989a, 1991; Klug et al., 1995). Experiments using the GABA antagonist picrotoxin as an inhibitory probe show that the phasic onset-type response of IC neurons to an acoustic stimulus can be converted to a tonic sustained-type discharge by application of picrotoxin. The abolition of synaptic inhibition by picrotoxin is most pronounced during the initial period of stimulation, lasting 40-80 msec. In contrast, picrotoxin had very minor effects on tonic units of the IC only slightly increasing their number of discharges (Watanabe & Simada, 1971, 1973). (This effective disinhibition by picrotoxin further supports the belief of GABA as a pre-synaptic inhibitor, as picrotoxin reverses pre-synaptic inhibition of spinal motorneurons; Eccles et al., 1%3). Neurochemical observations show that there are high concentrations of glutamic acid decarboxylase (GAD), the enzyme responsible for GABA synthesis, within the IC. GABAergic terminals can be labeled in the IC by GAD-immunocytochemistry or by antibodies to GABA itself (Adams & Wenthold, 1979; Thompson et al., 1985; Moore & Moore, 1987; Roberts & Ribak, 1987; Hutson, 1988; Glendenning et al., 1992). Evidence of synaptic terminals in IC visualized by GABA antibodies show that these terminals are found predominately along dendrites, rarely contacting the cell soma (e.g., Oliver & Bekius, 1992). Furthermore, it has been observed that GABA terminals are slightly elevated from the neuron perikaryon, suggestive of pre-synaptic inhibition via axo-axonic synapse (e.g., Thompson et al., 1985). Taken together, there is structural evidence for GABA terminals contacting IC
Auditory Pathway
419
neurons directly (possible post-synaptic mechanism), and indirectly (possible pre-synaptic mechanism). Similar methods of investigation have been employed to assess the possibility of glycine as an inhibitory neurotransmitter in the IC, though these experiments have not been as extensive as the GABA studies. Exogenously applied glycine has a powerful inhibitory effect on phasic neurons of the IC, acting with rapid onset and offset (Watanabe & Simada, 1971, 1973; Faingold et al., 1989a). Strychnine sulfate, a glycine antagonist has no disinhibitory effects on the phasic onset-type IC neurons, though in many cases there was a temporal depression of spike discharge (Watanabe & Simada, 1971, 1973). However, other investigations aimed at evaluating the role of glycinergic inhibition of IC neurons found that strychnine can block binaural interactions, that is, disinhibit an IE neuron (Faingold et al., 1989a; Klug et al., 1995). Glycine receptors have been demonstrated autoradiographically in the IC (Baker et al., 1986; Glendenning & Baker, 1988; Glendenning et al., 1992), and electron microscopic studies show that monoclonal antibodies to glycine receptors are found apposed to synaptic terminals containing flattened vesicles (Wenthold et al., 1988), supporting the view that glycine acts as the neurotransmitter for post-synaptic inhibition. In summary, there appears to be reasonably good evidence for GABA and glycine acting as inhibitory neurotransmitters in the IC. Excitation of IC Neurons. Recordings from electrodes placed intracellularly or extracellularly show that neurons can be excited by acoustic stimuli presented to either the ipsilateral or the contralateral ear (e.g., Nelson & Erulkar, 1963; Aitkin et al., 1981). However, the neurotransmitter mechanisms for excitation at the IC are poorly understood. The best candidates for excitatory neurotransmitters are acetylcholine, glutamate, and aspartate. The evidence for any of these putative neurotransmitters acting at the IC is not overwhelming, although glutamate is currently the most likely candidate based on neurochemical and receptor binding studies (e.g., Glendenning et al., 1992). Unfortunately, in many instances the results of pharmacological studies contradict one another (e.g., Curtis & Koizumi, 1961; Watanabe & Simada, 1973; Farley et al., 1983, Faingold et al., 1989b). About the only conclusive statement that can be made is that excitation probably involves glutamate as the neurotransmitter (Glendenning et al., 1992), that glutamate is associated with terminals containing round synaptic vesicles (Helfert et al., 1992), and this type of terminal morphology has
420
Hutson
long been believed as excitatory in the auditory system (e.g., Rockel & Jones, 1973b; Oliver, 1984, 1985, 1987).
Role of the Ipsilateral Pathway At the present time the evidence from electrophysiological studies suggests that inhibition of IC neurons is the result of activation of ipsilateral pathways (e.g., Roth et al., 1978; Semple & Aitkin, 1979; Semple & Kitzes, 1985; Aitkin & Martin, 1987; Carney & Yin, 1989). While the electrophysiology does show that units in IC can be classified by their response to afferent inputs (e.g., IE or EE), these responses are either net excitatory or net inhibitory, yielding no direct information as to the exact source or pathways by which excitation or inhibition reach the IC. To date, most experiments combining electrophysiological recording and tract-tracing methods have focused on tonotopic or topographic projections to IC from sub-collicular levels having little to offer about excitatory or inhibitory routes of innervation (e.g., Roth et a1.,1978; Aitkin & Schuck, 1985). The relative strength of the ipsilateral inhibition is most striking in autoradiographs obtained by the 14C-2-deoxyglucose method. In these experiments, white noise was presented to one ear and a tone to the other ear. As expected the IC contralateral to the tone stimulus displayed a band of increased activity, while the IC ipsilateral to the tone showed a band of reduced activity against a background of white noise induced activity (Webster et al., 1984, 1985). These authors argue that the reduced band of activity present in the IC ipsilateral to the tone corresponds to an inhibitory contour band, though again no information can be extracted as to the inhibitory pathway other than it originates from the tone stimulated ear. These electrophysiological and 14C-2-deoxyglucose findings together support the prevalent view that the acoustic information arriving at the IC from the ipsilateral ear is predominantly inhibitory. Before proceeding to evaluate the potential sources of the ipsilateral inhibitory pathway, it should first be acknowledged that the ipsilateral pathways are also responsible for excitation of the IC. A brief examination of this excitatory component of the ipsilateral pathway is warranted here for two reasons, first it enables one to estimate the relative size of the ipsilateral excitatory vs. inhibitory pathways and second, by analysis of the probable sites of origin of the excitatory
Auditory Pathway
421
pathway, they can be eliminated confidently from consideration in the inhibitory pathway. The common finding among all electrophysiological investigations of IC neurons is that the majority of cells examined could be excited by stimulation of the contralateral ear, while only about 25% are excited by stimulation of the ipsilateral ear (e.g., Roth et al., 1978; Semple & Aitkin, 1979, 1981; Kuwada et al., 1984; Moore et al., 1984; Semple & Kitzes, 1985). Thus on the basis of excitatory inputs, the ipsilateral pathways are about one-fourth the magnitude of contralateral pathways. But, given the evidence that the ipsilateral pathways also contain an inhibitory component raises the relative magnitude of the overall ipsilateral projection. One can obtain a rough estimate of the magnitude of the ipsilateral inhibitory pathway by deduction from the electrophysiological studies. Of the population of inferior collicular neurons studied electrophysiologically, approximately 20% respond only to contralateral stimulation (OE cells; e.g., Semple & Aitkin, 1979), leaving 80% of the original population being influenced to some degree by the ipsilateral pathways. From this 80%, an additional 25% can be subtracted as being excitable by ipsilateral pathways. Thus, at least 55% of the neurons sampled are capable of being influenced by the ipsilateral pathway in a non-excitatory manner. Granted such a deduction leaves many details unconsidered, yet it does convey the impression that the ipsilateral pathway is characterized by an inhibitory component that is by no means insignificant. As the origin of all ascending auditory pathways, the cochlear nucleus would be the first structure to examine as the origin of the excitatory and inhibitory components of the ipsilateral pathway. It will be recalled that the cochlear nucleus (both DCN and V CN) project to the IC of both sides (e.g., Adams, 1979; Brunso-Bechtold et al., 1981; Nordeen et al., 1983; Oliver, 1985, 1987). Of particular interest here are the studies which utilized the anterograde transport of 3H-leucine from the cochlear nucleus to label terminals in the IC (e.g., Oliver, 1984, 1987). From these injections dense terminal labeling appears in both the ipsilateral and contralateral IC, contralateral being somewhat heavier. Nevertheless, examination of both IC's at the electron microscopic level reveals that all terminals labeled by anterograde transport contained small round synaptic vesicles, with morphology of the terminals in the ipsilateral IC being indistinguishable from those of the contralateral IC. Furthermore these terminals formed asymmetric synaptic contacts
422
Hutson
primarily upon dendrites of IC neurons. The sum of these observations lead one to believe that the cochlear nucleus provides excitatory synapses in both the ipsilateral and contralateral IC. Finally, it is probably safe to conclude that although there may be other sources of ipsilateral excitation in the IC, the cochlear nucleus is probably n o t a direct source of the inhibition seen in the ipsilateral IC. Having eliminated the cochlear nucleus as a direct source of inhibition to the ipsilateral IC, the next most logical consideration would be the possibility of inhibitory interneurons in the IC (Rockel & Jones, 1973b). For this to be a plausible mechanism, excitatory ipsilateral inputs would contact the interneurons which in turn would inhibit other IC neurons. However, the electrophysiological evidence does not tend to support this mechanism. As suggested by Semple and Kitzes (1985), ipsilaterally excited neurons are seen too infrequently to account for the amount of inhibition produced, yet they concede the possibility that electrodes could simply be failing to record from these interneurons. This seems unlikely due to the size of neurons found in the IC, on the order of 17-30/am cell body diameter in cats (Oliver & Morest, 1984). Given the observation that the cochlear nucleus probably does not give rise to inhibitory projections to the ipsilateral IC and that evidence in support of the inhibitory interneuron theory is not particularly convincing, the number of alternative sources of ipsilateral inhibitory input to IC would seem to be reduced to the superior olivary complex and the nuclei of the lateral lemniscus. Further, given the high probability that the inhibitory neurotransmitters operating at the IC are GABA and glycine, one might reasonably suspect two categories of ipsilateral inhibitory pathways. Namely, a GABAergic pathway and a glycinergic pathway. The notion that the superior olivary complex or the nuclei of the lateral lemniscus may be the source of ipsilateral inhibition to IC is not novel, each has previously been suspected to play such a role (e.g., Semple & Aitkin, 1980; Adams & Mugnaini, 1984; Semple & Kitzes, 1985). One additional piece of information makes this point more salient. The ipsilateral superior olivary complex should provide (on the basis of their discharge characteristics) massive excitatory inputs to the IC. Both LSO and MSO receive excitatory inputs from the ipsilateral cochlear nucleus and themselves respond with excitation. However, as stated previously the incidence of ipsilaterally excitable cells in the IC is low, but as noted by Semple and Kitzes (1985) a large ipsilaterally evoked field potential is seen. This
Auditory Pathway
423
observation was puzzling to them, since it is suggestive of a vigorous volley of activity which is reduced to "relative feebleness" at the IC. However, such an observation is exactly what one would expect for an inhibitory pathway. These superior olivary complex neurons would respond to ipsilateral stimulation with action potentials producing a wave of activity that would n o t be replicated at the IC, due to the release of inhibitory rather than excitatory neurotransmitter substance. The only flaw in this explanation is that the recordings were made in anesthetized preparations, which had very low rates of spontaneous activity, so they could not directly observe the strength (or mechanism) of inhibition (Semple & Kitzes, 1985; Kitzes & Semple, 1985). Fortunately, there is reasonably good neurochemical evidence to support many of the known ipsilateral projecting nuclei as falling into one or the other (GABA or glycine) neurochemical pathway. Evidence for a GABAergic pathway begins with the localization of GABA positive neurons within these suspect nuclei. Using GABA antibodies, labeled neurons were found in LSO, DLL, and VLL (Thompson et al., 1985; Hutson, 1988; Helfert et al., 1989; Hutson et al., 1991). Other studies using GAD as a probe found labeled cells in DLL, and few in the dorsal division of VLL, and in LSO (Adams & Mugnaini, 1984; Moore & Moore, 1987). All of these are strong candidates for being the neural substrate for an ascending GABAergic pathway, with the exception of LSO. Many GABA positive cells are reported in LSO of rodents (e.g., Helfert et al., 1989), but only a few in cats (e.g., Glen-denning et al., 1992); however none of the cells had the morphology of LSO principal cells that give rise to ascending projections. This species difference may reflect the role LSO plays in descending projections of the superior olivary complex to the cochlea. A portion of this descending pathway originates from cells within LSO of rodents (e.g., White & Warr, 1983), but not in carnivores or primates (e.g., Warr, 1975; Thompson & Thompson, 1986). Furthermore, it appears to be the GABA positive cells within the rodent LSO that give rise to this descending projection (Vetter et al., 1991; Ostapoff et al., 1997). This leaves nuclei of the lateral lemniscus as the major source of ascending GABAergic inputs to the IC. Although GABA positive cells are found in DLL, VLL (and ILL), DLL contains the highest concentration (Adams & Mugnaini, 1984; Hutson, 1988; Glendenning et al., 1992).
424
Hutson
Perhaps the most convincing evidence in support of the nuclei of the lateral lemniscus being the source of GABAergic projections to the IC is the retrograde transport of radioactively labeled GABA (3H-GABA). When 3H-GABA is injected into the IC, it is rapidly incorporated by axon terminals and transported back to their cells of origin. In cat, this method selectively marks very few cells in VLL and ILL and only on the side ipsilateral to the injection, while many cells are marked in DLL both ipsilateral and contralateral to the injection (Hutson, 1988; Glendenning et al., 1992). Within the medullary auditory nuclei, n o cells were labeled in the cochlear nucleus of either side, none in MTB, MSO, or LSO of either side, with only a meager number of peri-olivary cells marked on the same side as the injection (Hutson, 1988; Glendenning et al., 1992). Other evidence for DLL as the origin of a significant bilateral inhibitory projection to the IC has come from electron microscopic studies of DLL axon terminals, demonstrating these terminals to contain pleomorphic shaped vesicles (Shneiderman & Oliver, 1989), and terminals containing pleomorphic vesicles are associated with GABAergic synapses in the auditory system (Oliver and Bekius, 1992; Helfert et al., 1992). Furthermore, ablation of DLL reduces the amount of induced GABA release in the IC (Shneiderman et al., 1993). Thus from the nuclei of the lateral lemniscus, only DLL appears to be a significant source of an ascending GABAergic projection to the IC. Before leaving this discussion of sources of GABA inputs to the IC, one additional source must be considered, and that is the IC itself. Previously in this paper, the notion of inhibitory interneurons within the IC was dismissed for lack of evidence. However, an inhibitory interneuron does not have to conform to the classic Golgi type II classification (i.e., a small neuron whose axon does not leave its parent nucleus). A projection neuron, with recurrent axon collaterals could serve the same purpose. Evidence for this possibly occuring in the IC comes from the observation that the IC contains a large population GABA positive neurons (e.g., Moore & Moore, 1987; Hutson, 1988; Glendenning et al, 1992; Oliver et al., 1994). Furthermore, if 3H-GABA is injected into the MG, a vast number of IC neurons are vividly marked by retrograde transport (Hutson et al., 1993). Thus, the IC gives rise to a significant ascending inhibitory projection to the MG. In addition, these large projection neurons often give off extensive axon collaterals before leaving the IC en route to MG (Oliver et al., 1991). These studies
Auditory Pathway
425
demonstrate that the IC is capable of supplying its own source of inhibition, yet the arguments against the possibility of interneurons being responsible for ipsilateral inhibition still hold true, few IC neurons are excited by stimulation of the ipsilateral ear. As for an ascending glycinergic pathway providing inhibition at the IC, the evidence is somewhat more compelling. The observations in favor of a glycinergic pathway include the previously noted presence of glycine receptors in the IC, and these receptors are most likely apposed to terminals containing flattened vesicles (Wenthold et al., 1988). Although there are no glycine positive cells in IC as measured by antibodies directed against glycine, there are abundant glycine positive fibers and puncta within IC (e.g., Glendenning et al., 1992), and glycine immunoreactive puncta have been demonstrated to be axon terminals that typically contain flattened vesicles (Helfert et al., 1992). A second line of investigation is, however, a most powerful demonstration of a glycinergic pathway and is based upon the retrograde transport of 3Hglycine itself. When 3H-glycine is injected into the IC of a cat, retrogradely labeled neurons are found in the ipsilateral VLL and ipsilateral LSO (Hutson, 1986; Hutson, et al., 1987; Hutson, 1988; Glendenning et al., 1992). Beyond this, it is important to note that retrogradely labeled neurons were n o t seen in other nuclei that would also be marked by transport of a non-specific tract-tracing substance (e.g., horseradish peroxidase) from an IC injection, these include the contralateral IC, ipsilateral and contralateral DLL, the cochlear nucleus of both sides, the contralateral LSO, and in particular the ipsilateral MSO. Similar results have been reported using retrograde transport of 3H-glycine in chinchillas (Saint Marie & Baker, 1990). Thus the possibility of specific glycinergic pathways appears quite likely, and the same electrophysiological argument used above (based on Semple and Kitzes, 1985) for ascending inhibition at the IC would hold here, with glycine most likely participating in a post-synaptic inhibitory process. It is equally important to point out that not all neurons of LSO retrogradely transport 3H-glycine after its injection into IC. Those that do transport glycine appear to represent a subpopulation of LSO principal cells (the remaining principal cells are candidates for an excitatory projection to the contalateral IC, see below; Glendenning et al., 1992). Furthermore, even though many VLL neurons retrogradely transport glycine, and only on the same side as the injection, they can not account for the observed inhibition of IC neurons resulting from
426
Hutson
stimulation of the ipsilateral ear. Recall that VLL receives almost exclusive afferent innervation from the contralateral cochlear nucleus, and since VLL gives rise to an ascending glycinergic projection it would be expected to have an inhibitory influence on IC neurons, but only after stimulation of the contralateral ear. Nonetheless, the evidence is reasonably convincing that both GABA and glycine are important inhibitory neurotransmitters in the central auditory system, and both are major components of the ipsilateral auditory pathway. This is interesting in view of pharmacological and electrophysiological findings. Pharmacological studies show that picrotoxin disinhibits IC neurons for up to 80 msec (Watanabe & Simada, 1971), while electrophysiological records show IC units can be silenced for up to 120 msec following initial excitation (Erulkar, 1959). These two results have led to the suggestion that the phasic response of IC neurons may be shaped by both pre- and post-synaptic inhibitory events (Erulkar, 1975). Therefore, LSO and DLL appear to be equipped with the necessary connections and neurotransmitters to carry out both pre- and post-synaptic inhibition at the level of the IC subsequent to stimulation of the ipsilateral ear. The observation of LSO neurons containing an inhibitory neurotransmitter (glycine), and that glycine has potent effects at the level of the IC can explain some long perplexing electrophysiolgical observations. Since most IC neurons are excited by contralateral ear stimulation (e.g., Moore et al., 1984; Semple & Kitzes, 1985), and the medial portions of LSO project contralaterally (Glendenning & Masterton, 1983), it would not be surprising to find a subpopulation of LSO neurons located in the medial limb of LSO containing an excitatory neurotransmitter, and indeed there is such a population. Glendenning et al., (1992) demonstrated the contralaterally projecting cells of LSO contain glutamate, and thus are a source of excitation at the IC. Similarly, the lateral portions of LSO project ipsilaterally (Glendenning & Masterton, 1983)and neurons retrogradely labeled by 3H-glycine are found in the lateral portions of LSO (Hutson, 1986; Hutson et al., 1987; Glendenning et al., 1992). Many units in the IC display response types that are IE (ipsilateral inhibitory, contralateral excitatory) and appear to be remarkably similar in their response characteristics to the contralateral LSO (e.g., Roth et al., 1978; Semple & Aitkin, 1979). This can be explained simply by the axons of LSO cells, which are EI in character, crossing the midline and terminating in the contralateral IC at which point they would now be characterized as
Auditory Pathway
427
IE (Glendenning & Masterton, 1983). However, there is a paucity of IC neurons displaying El characteristics which would reflect inputs from the ipsilateral (El) LSO (e.g., Semple & Aitkin, 1979). This has been the perplexing observation, why are there no (or very few) IC units that are "ipsi LSO-like"? Previous explanations of this quandary have been to assume the 'EI'-ness of the ipsilateral LSO projections are lost or masked by convergence and integration of other afferent types, obscuring any ipsilateral LSO-like characteristics (e.g., Roth et al., 1978; Semple & Aitkin, 1979; Glendenning & Masterton, 1983). In light of the findings presented above, that the ipsilateral LSO projections are inhibitory, an alternative explanation could be put forth. Consider for the moment one IC, say the right IC, receiving inputs from ipsilateral and contralateral LSO. Each LSO receiving excitatory inputs from its ipsilateral cochlear nucleus and inhibitory inputs from it contralateral cochlear nucleus via MTB. The left LSO projecting (by its medial portions) to the right IC and having excitatory terminals on IC neurons (see Figure 6). In this figure, one can easily see how El properties of the left LSO would be transformed into IE properties at the right IC. Now attend to the situation as it occurs in the right LSO, still El in character, but projecting inhibitory terminals to the right IC. Here, stimulation of the right ear (ipsilateral) excites the right LSO, however the right LSO, due to the action of the neurotransmitter released by its terminals, does not excite the fight IC but rather inhibits it. On the other hand, stimulation of the left ear (contralateral) would inhibit the right LSO and by doing so would promote a "disinhibitory" effect at the fight IC and appear as net excitation. In this manner, the ipsilateral LSO, while E1 itself, would not project El properties to the IC. Rather, the effect would be more IE-like in nature, and there would be no reason to expect EI or ipsilateral LSO-like unit response properties in the IC. In fact, by overlapping with OE and EE inputs (e.g., Semple & Aitkin, 1979) the ipsilateral LSO inputs would be enhanced further to appear as IE. For example, contralateral ear stimulation would excite an area of overlap by OE or EE cells, which would be potentiated by the withdrawal of inhibition by the ipsilateral LSO (i.e., contralateral stimulation would inhibit the inhibitory ipsilateral LSO projection thus promoting "disinhibition"). Likewise, stimulation of the ipsilateral ear would provide inhibition to the area of overlap.
428
Hutson
Response at right IC
Ear stimulated contra
ipsi o o
E O
~V - I
=E
-I
=E
= increased
firing
IC
o 0 O3 ~-, C G)
<
c~
~r
t ~
= decreased
firing
LSO A
contra ear
I
J
.L LSO
~
I
ipsi ear
midline
Figure 6. Right: schematic diagram illustrating how EI (ipsilaterally excited, contralaterally inhibited) neurons of both the contralateral LSO (1) and ipsilateral LSO (2) could produce IE (ipsilaterally inhibited, contralaterally excited) responses at the fight IC (3). Note that for simplicity, the synapse in MTB has been omitted. Left: 2 x 2 matrix summarizing the predicted changes in firing rate of IC neurons due to excitatory afferents from contralateral LSO vs inhibitory afferents from ipsilateral LSO. For example, stimulation of the fight ear (ipsilateral to IC) would excite the ipsilateral LSO, and in turn the ipsilateral LSO would inhibit the IC. E, +, excitatory; I, -, inhibitory; contra, contralateral; ipsi, ipsilateral; IC, inferior colliculus; LSO, lateral superior olive. Adapated from Hutson, 1986.
Auditory Pathway
429
Role of the Contralateral Pathway
Just as ipsilateral pathways are not exclusively inhibitory, neither are contralateral pathways exclusively excitatory. There is a small yet consistent reference to inhibitory events occurring at the IC with stimulation of the contralateral ear (e.g., Semple & Aitkin, 1979, 1981; Webster et al., 1985; Carney & Yin, 1989). These unit responses have not been studied as intensely as other types, and surveys of IC units indicate that contralaterally inhibited units are meager in number (e.g., Semple & Aitkin, 1979), as are their potential neural substrates. The best candidates for mediating the contralaterally induced inhibition in IC are the DLL, MTB, and VLL. DLL gives rise to a crossed GABAergic projection via the commissure of Probst (e.g., Hutson, 1988; Hutson et al., 1991; Glendenning et al., 1992), although it receives afferent information derived from both ears (e.g., Glendenning et al., 1981; Shneiderman et al., 1988). Both MTB and VLL (for all practical purposes) are exclusively innervated by the contralateral cochlear nucleus, and both project to their ipsilateral IC (Brunso-Bechtold et al., 1981; Glendenning et al., 1981; 1985; Spangler et a1.,1985). The projection from MTB to IC is quite small (Brunso-Bechtold et al., 1981), yet it is almost certainly glycinergic in nature. Principal MTB cell innervation of LSO is probably glycinergic (e.g., Moore & Caspary, 1983, Glendenning et al., 1991) and collaterals from some of these same axons continue to ascend to the IC (Spangler et al., 1985). VLL, on the other hand, projects widely to the IC (e.g., Whitley & Henkel, 1984; Glendenning et al., 1990). Neurons of VLL are glycine positive and transport 3H-glycine from IC injections (e.g., Hutson et al., 1987; Hutson, 1988; Glendenning et al., 1992), and a small number of VLL neurons are also GABA positive and transport 3H-GABA (e.g., Thompson et al., 1985; Hutson, 1988; Glendenning et al., 1992). Therefore the DLL and VLL may be the major source of contralaterally induced inhibition of IC units. Binaural Interaction at the IC
The ipsilateral pathway contributes to binaural interactions in the IC which are measurable in terms of a single neurons response to stimuli delivered to both ears. Hind et al. (1963) demonstrated that activity elicited by a stimulus delivered to both ears may be quite different than
430
Hutson
when it is presented to each ear independently. Thus, information ascending in the ipsilateral and contralateral pathways converge on IC neurons. This was an important observation, for it addresses the neurophysiological basis of sound localization. From a single source of sound in space, different physical cues are available at each ear for localization of that source, i.e., those that reflect different levels of stimulation reaching each ear, such as differing times of arrival of the sound wave at each ear (e.g., Rose et al., 1966). These interaural differences, or disparities, include not only time, but intensity, phase, frequency, and other spectral differences. By manipulating the stimuli applied to each ear, or changing the location of a stimulus in space, a neurons response to these changing locus cues can be examined. For example, clicks presented to one ear typically excite neurons in the contralateral IC, whereas clicks presented to the other (ipsilateral) ear will inhibit this activity in a graded manner dependent upon the intensity of the clicks presented to each ear (e.g., Flammino & Clopton, 1975; Bengry et al., 1977). If the click intensity to the contralateral ear started at 40 dB while at the ipsilateral ear 0 dB, the observed response was excitatory in the contralateral IC. When the intensity to the contralateral ear decreased as the intensity to the ipsilateral ear increased (keeping the overall stimulus intensity constant at 40 dB), there was a systematic decrease, or inhibition, of discharge rate. This inhibition was minimal as long as the stimulus was more intense contralateral than ipsilateral. Inhibition grew in magnitude once the equal intensity point (20 decibels to each ear) was reached and continued as the stimulus grew louder in the ipsilateral ear (Bengry et al. 1977). This interaural intensity difference would mimic a sound as it moved in space from directly opposite the contralateral ear, to the midline, and then opposite the ipsilateral ear. From these studies and others (e.g., Goldberg & Brown, 1969; Brugge er al., 1970; Kuwada et al., 1984, 1987; Batra et al., 1989; 1993; Carney & Yin, 1989; Yin & Chan, 1990; Tsuchitani, 1997) it is evident that the ipsilateral pathway plays a significant role in the binaural processing of directional cues at the level of the superior olivary complex, through the nuclei of the lateral lemniscus, to the IC (for a more complete review of binaural interaction, see Irvine, 1992). Recent pharmacological studies have attempted to dissect the role of inhibition in shaping binaural responses at the IC. One approach has been to simultaneously record from one IC and chemically blocking
Auditory Pathway
43 1
GABA or glycine receptors to study changes in binaural response patterns. Observed changes in binaural response patterns ranged from total elimination of ipsilateral inhibition, to subtle changes in firing patterns, to altering a neurons response to interaural intensity differences (e. g., Faingold et al., 1991; Park & Pollak, 1993; Klug et al., 1995). The response modifications imposed by blocking GABA receptors suggest that GABA inputs to the IC are important in defining a neurons recep-tive field, in other words, important for the central representation of location in auditory space (Park & Pollak, 1993). Another approach has been to record binaural responses in one IC while manipulating components of the ascending pathways to evaluate the source of inhibition. For example, infusion of lidocaine (a local anesthetic) into the contralateral DLL blocks the inhibitory component usually present following binaural acoustic stimulation, whereas induced excitation (electrical or c h e m i c a l ) i n contralateral DLL reduces a neurons firing rate in a manner similar to binaural acoustic stimulation (Faingold et al., 1993). Other investigations used interaural intensity differences (e.g., Kelly & Li, 1997), or interaural time differences (e.g., Kidd & Kelly, 1996) to measure binaural interaction, in combination with local blockade of DLL or the superior olivary complex. These studies demonstrate two sources of inhibition acting on IC neurons. Blocking activity in the ipsilateral superior olivary complex reduces binaural inhibition, as does blocking the contralateral DLL, while blocking the ipsilateral DLL has little effect on binaural inhibition. Furthermore, while demonstrating the importance of the superior olivary complex and DLL as sources of ascending inhibition, they also show that a portion of the inhibition seen in one IC, resulting from stimulation of its ipsilateral ear, is mediated by the contralateral DLL. Thus, information from the right ear can be processed along the contralateral pathway to the left DLL, then from the left DLL back to the fight IC, ultimately inhibiting the right IC in response to stimulating the fight ear. These studies demonstrate the lengths to which the central auditory system must go in order to construct an accurate representation of the right vs. left auditory world.
Spontaneous Activity A consideration of how the ipsilateral pathway contributes to physiological mechanisms would be incomplete without mentioning
432
Hutson
spontaneous activity. Certainly if a neuron maintains a high rate of spontaneous activity, then depression of this activity could be as important a neural "cue" as enhancement of this activity. Thus, in a spontaneously active neuron, afferent inhibition as well as excitation may play a major role in determining that neurons response repertoire and in how it "codes" its efferent discharge. High levels of spontaneous activity characterize all levels of the central auditory system (e.g., Bock et al., 1972; Boudreau & Tsuchitani, 1973; Brownell et al., 1979; Clarey et al., 1992). In particular, units of the IC display spontaneous discharge rates on the average of 15-20 spikes/sec (Bock et al., 1972; Ryan & Miller, 1978), while trapezoid body fibers show rates averaging 35-40 spikes/sec (Spirou et al., 1984) though there is wide variation in these estimates (see Spirou et al., 1990). Spontaneous activity has been reported for neurons in the cochlear nucleus (e.g., Rhode & Smith, 1986; Parham & Kim, 1995), and also the auditory nerve with discharge rates ranging from 0-120 spikes/sec (e.g., Kiang, 1963; Liberman, 1978). However, the electrophysiological demonstrations of spontaneous activity are difficult to reconcile with other investigations using 14C-2deoxyglucose to mark stimulus induced activity. During the course of experiments on stimulus coding in the auditory brainstem, Nudo and Masterton (1984) prepared a case with bilateral destruction of the cochleae, and found little evidence for spontaneous activity in their 14C_ 2-deoxyglucose radiographs. From their observations, it would follow that spontaneous activity is not spontaneous, but driven. Given that the ear is sensitive to Brownian or near Brownian movement (e.g., Sivian & White, 1933; Stevens & Davis, 1938; De Vries, 1948; Harris, 1968; Hudspeth, 1989; Yates et al., 1992), the possibility exists that spontaneous activity observed electrophysiologically is driven by thermal fluctuations at the ear. The evidence of Nudo and Masterton (1984) could also be interpreted as a diminished baseline of activity. There is some electrophysiological support for this conclusion, in that section of the auditory nerve reduces, but does not eradicate, spontaneous discharges within central auditory structures (e.g., Starr & Livingston, 1963). Despite these conflicting interpretations, (spontaneous vs. driven), and no matter which is actually correct, it remains that in the absence of direct acoustic stimulation there is a baseline of activity within the auditory system that could allow for greater resolution, or a more sensitive "gauge", of afferent inputs.
Auditory Pathway
433
Anesthetics depress the spontaneous rates dramatically, for example in the trapezoid body spontaneous rates drop to one-fourth of their normal rate, therefore losing up to 75% of the potential information derivable from changes in their rate of spontaneous discharges (Spirou et al., 1984). At the IC, spontaneous activity of units drops to either very low levels or is non-existent (e.g., Aitkin et al., 1981; Kitzes & Scruple, 1985). Kuwada et al. (1989) demonstrated that even a sub-surgical dose of sodium pentobarbital significantly alters a neurons response characteristics to monaural and binaural stimulation. Furthermore, studies that surveyed the IC to categorize unit response properties were done in anesthetized preparations, where inhibition was revealed only by the fact that responses to binaural stimulation were smaller than to monaural stimulation (e.g., Aitkin et al., 1981). Thus our reliance on IC unit classification should be tempered by the knowledge that they are biased towards excitatory classes. In unanesthetized preparations, all IC cells exhibit spontaneous discharges and depression of this spontaneous activity was a prominent feature of the response patterns in many of the units (Book et al., 1972), and in LSO unanesthetized preparations revealed unforeseen unit response characteristics (e.g., Brownell et al., 1979). Recent investigations using unanesthetized preparations (e.g., Stanford et al., 1992; Fitzpatrick et al., 1995; Covey et al., 1996), or slice preparations (e.g., Wu & Kelly, 1992, 1994; Smith, 1995) have provided unique insights into auditory function, especially with regard to binaural interactions. Needless to say anesthetics have obscured the role of inhibition in the auditory system not only for the ipsilateral pathways but also for the contralateral pathways. The Acoustic Chiasm
Returning to the behavioral data on sound localization, unilateral damage to the auditory system anywhere from the lateral lemniscus through auditory cortex results in sound localization deficits within the sensory hemifield of the opposite side, while damage below the level of the lateral lemniscus results in localization deficits in both sensory fields or in the sensory field on the same side as the damage (e.g., Jenkins & Masterton, 1982). Thus, despite the lack of an orderly structural chiasm in the auditory system, there is a proper functional chiasm, one that
434
Hutson
allows for the representation of each auditory hemifield in the opposite side of the brain. All evidence supports the conclusion that the auditory system, like the visual and somatosensory systems is "contralateralized", such that the right versus left sensory fields are represented in the left versus right cerebral cortex, respectively. Unlike the visual or somatosensory systems, the peripheral receptor distribution in the cochlea can not map its associated sensory field. In the visual system for example, the left visual field maps directly to the nasal half of the retina in the left eye and to the temporal half of the retina in the fight eye. It is a simple matter then, for the retinal ganglion cells to sort the visual fields at the optic chiasm, fibers mapping the left visual field (from each retina) simply continue on as the right optic tract, those that represent the right visual field form the left optic tract. The auditory receptors (hair cells) in the cochlea code frequency and intensity, not a position in space. Therefore, no mere sorting of ascending fibers will work to produce an acoustic chiasm. Apparently, the acoustic chiasm is a process not a structure. How then does the auditory system become contralateralized at the level of auditory cortex? How can a system so hopelessly bilateral in terms of fiber connections become functionally contralateral? The process of the acoustic chiasm does depend on some structural elements, and the behavioral data suggests the structures responsible lie between the trapezoid body and the lateral lemniscus. Somehow between these two fiber pathways, the central nervous system is able to construct a fight and left auditory hemifield. The most likely candidates for this chiasmatic process occur among the afferents, efferents, and integrative properties of the superior olivary complex. Based on the anatomical and biochemical findings discussed above, Glendenning and Masterton (1983) proposed that the chiasm is a two step process, the first step founded on the nature of afferents to LSO, the second on how LSO distributes it's efferent fibers. Briefly, the process works as follows. In the first step, LSO receives excitatory inputs from the ipsilateral cochlear nucleus, and inhibitory inputs from the contralateral cochlear nucleus via synapse in the MTB (the MTB serving as a collection of inhibitory interneurons), giving LSO neurons EI binaural characteristics. Thus from a lateral sound source, the ipsilateral LSO would be excited, while the contralateral LSO would be inhibited. This would allow LSO to do a spectral analysis of the differences in physical cues available at each ear originating from a single sound source in
Auditory Pathway
43 5
space. In essence subtracting information arriving at one ear from information provided by the other ear yields information about positions of sounds in space. As such, this would represent a functional division or 'chiasm' of the auditory field, though backwards with respect to behavioral results. It should be noted that if LSO neurons project contralateral in their entirety, then the chiasm would be achieved, the spectral analysis would be contralater-alized and insofar as that would be sufficient to explain the behavioral results, there would be a sufficient chiasm. However, this is not the case. For over a century it has been suspected, and undisputed for the last 30 years, that LSO projects nearly equally to the IC of both sides of the brain (see earlier discussion on anatomy). However, the projections to IC do not arise equally across the limbs of LSO. The lateral limb of LSO (low frequency) projects more to the ipsilateral IC, while the medial limb (high frequency) projects more to the contralateral IC, few cells project to both colliculi. Although this bilateral separation of projections appears something like an optic chiasm, this is only an illusion. Recall that most LSO neurons are El in binaural terms, while many IC neurons are binaurally IE, virtually none El. What happens to LSO's El outputs at the IC? Since LSO projections to the contralateral IC are glutamatergic (excitatory), simply by crossing the midline they become IE with respect to the contralateral IC. LSO projections to the ipsilateral IC are glycinergic (inhibitory), so it effectively supplies IE information to the ipsilateral IC (see Figure 6). Thus, LSO contralateralizes the representation of a sound source not by virtue of its connections alone, but rather by its connections in combination with its biochemistry (see also Glendenning et al., 1985, 1991,1992; Hutson et al., 1991). As an example, imagine a moving sound source which begins directly opposite the left ear, moves toward the midline, then continues along ending directly opposite the right ear. This sound would first excite the left LSO, while inhibiting the right LSO. Through LSO's efferent distribution to the IC, the left (ipsilateral) IC would be inhibited, the right (contralateral) IC excited, and therefore the sound becomes functionally contralateralized to the right side of the brain. As the sound moves towards the midline, each LSO would become more 'neutral', given that its afferent supply from each ear (becoming nearly equal in terms of excitation and inhibition) would tend to cancel out, and LSO efferent projections to IC would not favor excitation of either the left or
436
Hutson
right IC (a midline source then producing no distinct lateralized excitation of one IC). Finally, when the source stops opposite the right ear, it is the fight LSO that now becomes excited, the left LSO inhibited, and excitation at the level of the IC shifts to the left (contralateral) side. In this manner, each lateral lemniscus contains information about the opposite auditory hemifield that ultimately contralateralizes the representation at the IC. DLL appears to play a role in the chiasmatic process by enhancing (sharpening) the representation of auditory space in IC, particularly for a sound source near the midline (e.g., Ito et al., 1995). The question now arises as to what contribution does MSO make to this chiasmatic process? To answer this, we must reconsider both the anatomy and physiology of MSO, bearing in mind the behavioral results which suggest that each lateral lemniscus carries information about the opposite auditory sensory field. MSO receives excitatory inputs from the VCN of both sides, and its efferent projections are almost exclusively to higher order auditory structures on the same side of the brain, thus fibers from MSO constitute a portion of the ascending ipsilateral pathway. Furthermore, MSO neurons, at least in cat, are intensely stained by antibodies to glutamate (Glendenning et al., 1992), and they do not retrogradely transport 3H-GABA or 3H-glycine injected into IC (Hutson et al., 1987; Hutson, 1988; Glendenning et al., 1992). However, in rodents a small number of GABA positive neurons have been reported in MSO (e.g., Thompson et al., 1985; Helfert et al., 1989). The arrangement of MSO afferents and efferents led to the belief that MSO was the anatomical substrate for sound localization based on differences in time of arrival of a sound to each ear (Jeffress, 1948). Based simply on these observations alone, one would expect a laterally located sound source to excite MSO on both sides and thus excite IC on both sides, though perhaps in differing proportions. The crucial evidence for a functional role of MSO in the acoustic chiasm was provided by Yin and Chan (1990), who found that neurons in MSO are indeed sensitive to interaural time differences. More importantly, MSO cells preferred (i.e., had maximum firing rates) when the time difference corresponded to a stimulus located in the contralateral hemifield. Consequently, MSO's contribution to it's ipsilateral lateral lemniscus would include infor-mation about location of a sound source on the opposite side, confor-ming to the rule of each lemniscus representing contralateral space.
Auditory Pathway
43 7
Turning to the question of what becomes of MSO inputs to the IC raised by Kitzes and Semple (1985), a partial answer may be that since MSO neurons respond best to binaural inputs, the paucity of MSO-like responses in IC of their monaural preparation results from each MSO being deprived of half of its normal afferent supply, yielding impoverished action at the IC. Still, their question remains valid on others issues, such as how the IC integrates low frequency inputs from ipsilateral LSO and MSO, or what becomes of MSO's strongly EE inputs at the IC. Although there are as yet no direct answers, both Yin and Chan (1990) and Glendenning et al. (1992) have suggested the answer may lie in the way LSO and MSO efferents overlap within similar regions of IC, and perhaps in combination with DLL as an inhibitory intermediate. Nonetheless, the weight of the evidence clearly suggests that the ipsilateral pathways are equally important as contralateral pathways for the accurate localization of sounds in space. To maintain localization in one hemifield requires an intact contralateral lateral lemniscus, with constituent fibers originating from both sides of the brain, i.e., a contralateral pathway and an ipsilateral pathway. Thus, the auditory system, through a chiasmatic process, achieves a functional contralateralization of the auditory sensory hemifields. In other words, due to the nature of the receptor surface in the cochlea, the central nervous system must construct a right and left auditory world. However, it is important to note that although the auditory system is functionally contralateralized by the level of the lateral lemniscus, each ear is represented in each lateral lemniscus, and beyond. Beyond the Lateral Lemniscus and IC
Although there are ascending projections to the MG that by-pass the IC (e.g., Hutson et al., 1991; Hutson & Glendenning, 1995), the IC for all practical purposes is an obligatory relay within the ascending auditory system (e.g, Aitkin & Phillips, 1984a). The anatomical pathway from the IC through MG to auditory cortex is essentially ipsilateral, though remaining functionally contralateral (e.g., Jenkins & Masterton, 1982). Still, given that each ear is represented in each lateral lemniscus, it follows that each ear is also represented in each structure above the IC, i.e., in the MG (e.g., Galambos et al., 1952), and auditory cortex (e.g., Tunturi, 1946; Rosenzweig, 1951). These forebrain structures are tonotopically organized, and neurons in the MG and auditory cortex
438
Hutson
can be classified in terms of their binaural response properties (see Clarey et al., 1992 for a review), or their directional sensitivity (e.g., Middlebrooks & Pettigrew, 1981). The IC projection to MG is a faithful transference of auditory information (e.g., Stanford et al., 1992), as is the MG projection to auditory cortex (e.g., Middlebrooks & Zook, 1983). However, the role of inhibition in the ascending auditory system does not stop at the IC. A significant proportion of IC projections to MG are GABAergic (Hutson et al., 1993), and this GABAergic projection probably accounts for the dramatic inhibitory effects seen in unit recordings from MG (e.g., Stanford et al., 1992). At the level of auditory cortex, not only is each ear represented on each side of auditory cortex, stimulation of either ear can produce excitation or inhibition in the auditory cortex of either side (e.g., Calford & Semple, 1995). Thus, we can extend the conclusions of Hind et al. (1963) to say that "excitatory-inhibitory interplay" is commonplace at virtually all levels of the ascending auditory system. Again, it is important to remember that although each ear is represented in auditory cortex, each sensory hemifield is not. To accurately derive a cortical representation of the two auditory sensory hemifields requires both ipsilateral and contralateral pathways. Even though the contralateral pathways to the IC are probably larger than the ipsilateral pathways in terms of absolute number of fibers (e.g., Masterton et al., 1992), the functional difference between these pathways within one lateral lemniscus (and therefore within higher order struc-tures) is more a difference in degree of sensitivity, or relative magnitude. For example, neither the ipsilateral or contralateral pathway is incapable of detecting high or low frequencies, nor is one pathway entirely excitatory or inhibitory, though the ipsilateral pathway may fairly be summarized as more sensitive to low frequencies and plays a larger role in inhibitory mechanisms than does the contralateral pathway. The ultimate fate of these pathways lies in auditory cortex, and many details of their contribution to cortical function remain to be discovered. However, unilateral ablation of auditory cortex has consistently been found to result in the loss of ability to accurately localize sounds in the opposite (contralateral) sensory hemifield of carnivores (e.g., Jenkins & Masterton, 1982; Jenkins & Merzenich, 1984; Kavanagh & Kelly, 1987) and primates, including man (e.g., Sanchez-Longo & Forster, 1958; Cortez et al., 1983; Thompson &
Auditory Pathway
43 9
Cortez, 1983; Heffner et al., 1992; Poirier et al., 1994). Furthermore a discrete ablation of primary auditory cortex, limited to a circumscribed frequency region, results in contralateral localization deficits that are frequency specific (Jenkins & Merzenich, 1984). Thus even a small, focal lesion in auditory cortex disrupts normal ability to localize contralateral sounds. Therefore, the sum of the evidence supports a functional contralateralization of auditory hemifields in primary auditory cortex, that ipsilateral and contralateral pathways are active constituents in this process, and the anatomical substrates for the process are those described in preceding sections of this paper. Given that man is included in the list of species where sound localization is fragile to cortical insult, the question arises as to whether the ipsilateral pathways contribute to higher order cortical functions, i.e., cognition.
V. Consequences to Cognition Ferrier (1876b) observed that unilateral lesions of auditory cortex in monkeys resulted in an immediate hearing loss in the contralateral ear. However, the animals show substantial recovery of hearing in the contralateral ear, presumably due to the presence of an ipsilateral pathway (arising from the same ear) but projecting to the undamaged hemisphere (e.g., Heffner & Heffner, 1989). Thus, information arriving at one ear can reach the cortex via contralateral and ipsilateral pathways. Nevertheless, anatomical and physiological studies have indicated that the contralateral pathway to cortex is the major pathway in terms of number of fibers and latency of response (e.g., Rosenzweig, 1951; Pantev et al., 1986), a finding that has behavioral consequences. In normal subjects, when pure tones or verbal stimuli are presented to each ear independently, performance is identical, neither ear shows an advantage. However, the relative dominance of the contralateral pathway has been repeatedly demonstrated in dichotic listening studies in which different verbal stimuli are simultaneously presented to both ears and subjects are asked to report what is heard (e.g., Broadbent, 1954). Under these conditions, there is a small, but consistent "right-ear-advantage", where stimuli presented to the right ear are correctly reported more often than the stimuli presented to the left ear. This result is explained by the presence of a speech reception area (Wernicke's area) which, in most people, is located in the left temporal lobe whose dominant inputs originate from the fight ear.
440
Hutson
The dominant nature of the contralateral auditory pathway is also important in explaining the effects of temporal lobe damage and corpus callosum section. Unilateral temporal lobe damage affects the perception of dichotic stimuli in the ear contralateral to the damaged hemisphere while the ipsilateral ear is less affected (Kimura, 1961). This result demonstrates, again, the relative importance of the contralateral pathway for auditory stimuli. The contralateral nature of the auditory pathway in patients with section of the corpus callosum is shown by an exaggerated right-ear-advantage, suggesting that in normal individuals verbal stimuli presented to the left ear reach Wernicke's area in the left hemisphere via the contralateral pathway to the right hemisphere and then cross to the left hemisphere via the corpus callosum (e.g., Milner et al., 1968). The fact that split-brain patients readily report verbal stimuli presented to the left ear when no competing stimuli are presented to the right ear indicates that although the ipsilateral pathway can mediate verbal stimuli, the main pathway is contralateral. Furthermore, this exaggerated right-ear-advantage in "split-brain" individuals can be further enhanced if the task demands are increased (Prechstedt, 1986). Given that the magnitude of the "ear-advantage" can be manipulated (e.g., Schwartz & Tallal, 1980; Hugdahl, 1995), and that it is maximal for brief, complex signals, can the study of brief signals tell us anything about central auditory pathways? It has been known for over a century that brief sounds, such as verbal stimuli, carry an immense amount of acoustic information (e.g., Helmholtz, 1875). Perhaps the best demonstration of the importance of brief changes within a stimulus may be found in the recognition of musical instruments. Stumpf (1926) discovered that if deprived of an instruments "attack" (i.e., initial, rapidly changing transients present when an instrument begins to sound, musicians had difficulty recognizing individual instruments. Thus patterned changes in brief acoustic stimuli are likely to be important cues for the recognition of a sound source, which may be cortically dependent. While damage to auditory cortex has little affect on discrimination of physical characteristics of sound, such as intensity or frequency, it results in long lasting deficits in sound localization, lateralization, and discrimination of complex sounds (see Masterton & Berkley, 1974; Ravizza & Belmore, 1978, Heffner & Heffner, 1990b), including discrimination of vocalizations (e.g., Heffner & Heffner, 1986, 1994). The ability to discriminate these acoustic signals appears to reside in both hemispheres, and thus are reliant, to some extent, upon both
Auditory Pathway
441
ipsilateral and contralateral pathways (e.g., Heffner & Heffner, 1986). Although the underlying mechanisms involved remain unknown, some conclusions can be drawn with respect to the contribution of the ipsilateral pathways. At the level of the brainstem, each lateral lemniscus carries information about the location of sound in the contralateral hemifield. Reduction of the brainstem system to one ear connected only to its ipsilateral, or its contralateral lateral lemniscus suggests that the information carried in a contralateral lateral lemniscus may ultimately prove to be the prefered pathway for localization and recognition, while the ipsilateral lateral lemniscus may be the preferred pathway for detecting the presence of a sound (Masterton et al., 1992). At the level of auditory cortex, the situation is somewhat different. Although unilateral ablation of auditory cortex results in localization deficits in the contralateral hemifield, the acoustic chiasm accomplishes the contralateralization of auditory space within the brainstem, therefore cortex itself need not derive auditory space and probably performs other types of acoustic analysis. Unilateral ablation of auditory cortex leaves auditory cortex of one hemisphere connected to each ear (ipsilateral and contralateral), and when each ear is tested separately, the resulting loss in sensitivity has been described as a contralateral "ear" deficit rather than a contralateral field deficit, such that detection and discrimination of sounds is reduced in the ear contralateral to the damaged cortex (Heffner & Heffner, 1989). Taken together, the human and animal investigations imply that without auditory cortex discrimination of brief, complex stimuli suffers. In a normal animal, or human, it seems possible that the ipsilateral pathways may provide the cortex with information about steady, environmental "noise", essentially informing cortex on the state of sound in the auditory field, upon which the contralateral pathways locate and recognize a "signal" relative to the background "noise". In this manner, one function of auditory cortex may be the detection, and recognition, of sounds arriving at the contralateral ear (Heffner & Heffner, 1990a). VI. Conclusions
This paper has focused on the experimental observations thought to be most fruitful in revealing the nature of the ipsilateral auditory pathways. Few investigations (e.g., Kreidl, 1914) have expressly sought
442
Hutson
to examine the ipsilateral system, therefore pertinent information must be retrieved from the total findings. Anatomically there can no longer be any doubt of the existence, and magnitude, of the ipsilateral pathways. There are both electrophysiological and neurochemical grounds for the segregation of ipsilateral and contralateral pathways coursing through the brainstem. At the level of the inferior colliculus, the last point where the auditory system can clearly be separated on the basis of ipsilateral and contralateral sub-structure, the ipsilateral pathways impose a powerful inhibitory influence. However, it should be emphasized that the ipsilateral pathway though predominantly inhibitory, does carry excitatory inputs to the IC. Similarly the contralateral pathway is predominately, though not strictly excitatory in nature. Functionally, the ipsilateral pathways are biased toward, though not dominated by, low to mid-range frequencies. The ipsilateral projections arise chiefly from nuclei, or portions of nuclei, tonotopically preferential for low frequency acoustic stimuli. From a behavioral standpoint, the ipsilateral pathways are of substantial significance. Further, of all the behavioral demands placed on the ipsilateral system, the evidence suggests that only sound localization can not be maintained by the ipsilateral pathways in and of themselves. Finally, although the existence, let alone the functional significance of the ipsilateral pathways, has been in debate for nearly a century, such debate can finally be put to rest. The ipsilateral pathway is a viable, functioning entity deserving of greater efforts to detail its underlying properties. Indeed an understanding of these properties and their contribution to the central auditory system are essential to our accurate interpretation of the science of audition.
Footnote
1. For students just beginning their study of the auditory system, there is frequently some point of confusion regarding the binaural classification of neurons. The practice of classifying a cell based on its response to stimulation of one ear and then the other, dates to Goldberg and Brown (1968). They used this nomenclature to catagorize cells in the superior olivary complex as excited by either ear (EE) or excited by one ear and inhibited by the other (El), without regard to ear order. E1 could mean ipsilateral excitation : contralateral inhibition, or contralateral excitation: ipsilateral inhibition (see their Table 1). Subsequent authors have used this nomenclature to classify neurons throughout the central
Auditory Pathway
443
auditory system, but have never agreed upon a standard order. Thus LSO cells are often described as El, but also as IE. The difference between authors appears to be the result of their scientific upbringing. Those of us raised in the superior olives see LSO as El, excitation the result of ipsilateral ear stimulation, so our frame of reference for the entire central auditory system is ipsilateral : contralateral. We then classify IC cells as IE. Those raised in the IC or above, see excitation as the result of contralateral ear stimulation, and thus view the auditory system as contralateral : ipsilateral, classifying IC cells as El, and LSO cells as IE. The functional interpretations are the same, ear order simply reversed.
Acknowledgments The author would like to thank Dr. H. Heffner for his comments and discussions regarding cognitive issues; Dr. G. Koay and Mr. Shawn Koons for a careful reading of the entire manuscript. To S.L.M. and B.N.B.
References Adams, J.C. (1979). Ascending projections to the inferior colliculus. J.
Comp. Neurol., 183:519-538. Adams, J.C. (1983). Cytology of periolivary cells and the organation of their projections in the cat. J. Comp. Neurol., 215: 275-289. Adams, J.C., & E. Mugnaini (1984). Dorsal nucleus of the lateral lemniscus: a nucleus of GABAergic projection neurons. Brain Res. Bull., 13: 585-590. Adams, J.C., & W.B, Warr (1976). Origins of axons in the cat's acoustic striae determined by injection of horseradish peroxidase into severed tracts. J. Comp. Neurol., 170:107-122. Adams, J.C., & R.J. Wenthold (1979). Distribution of putative amino a c i d t r a n s m i t t e r s , choline a c e t y l t r a n s f e r a s e and g l u t a m a t e decarboxylase in the inferior colliculus. Neurosci., 4: 1947-1951. Ades, H.W. (1944). Midbrain auditory m e c h a n i s m s in cats. J. Neurophysiol., 7:415-424. Ades, H.W., & J.M. Brookhart (1950). The central auditory pathway. J. Neurophysiol., 13: 189-205. Aitkin, L.M., D.J. Anderson, & J.F. Brugge (1970). Tonotopic organization and discharge characteristics of single neurons in the nuclei of the lateral lemniscus of the cat. J. Neurophysiol., 33:421-440. Aitkin, L.M., D.R.F. Irvine, & W.R. Webster (1981). Central neural mechanisms of hearing. In: Handbook of Physiology. The Nervous System. IlL Washington, D.C." American Physiological Society.
444
Hutson
Aitkin, L.M., & R.L. Martin (1987). The representation of stimulus azimuth by high best- frequency azimuth-selective neurons in the central nucleus of the inferior colliculus of the cat. J. Neurophysiol., 57:1185-1200. Aitkin, L.M., & S.C. Phillips (1984a). Is the inferior colliculus an obligatory relay in the cat auditory system? Neurosci. Lett., 44: 259-264. Aitkin, L.M., & S.C. Phillips (1984b). The interconnections of the inferior colliculi through their commissure. J. Comp. Neurol., 228: 210-216. Aitkin, L.M., & D. Schuck (1985). Low frequency neurons in the lateral central nucleus of the inferior colliculus recieve their input predominantly from the medial superior olive. Hearing Res., 17: 87-93. Anderson, R.A., G.L. Roth, L.M. Aitkin, & M.M. Merzenich (1980). The efferent projections of the central nucleus and the pericentral nucleus of the inferior colliculus in the cat. J. Comp. Neurol., 194: 649-662. Ariens Kappers, C.U., G.C. Huber, & E.C. Crosby (1936). The Comparative Anatomy of the Nervous System of Vertebrates, Including Man. New York: Macmillan Co. Arnesen, A.R., & K.K. Osen (1978). The cochlear nerve in the cat: topography, cochleotopy, and fiber spectrum. J. Comp. Neurol., 178: 661-678. Baginski, B. (1886). Uber den Ursprung und den centralen Verlauf des Nervus acusticus des Kaninchens. Arch. f. path. Anat. u. Physiol., 105: 28-46. Baginski, B. (1890). Uber den Ursprung und den centralen Verlauf des Nervous acousticus des Kaninchens und der Katze. Arch. f. path. Anat. u. Physiol., 109:81. Baker, B.N., K.K. Glendenning, & P. Hodges (1986). Acoustic chiasm: distribution of glycine and GABA receptors in the brainstem auditory nuclei of cat. Assoc. Res. Otolaryngol., 9: 7. Banks, M.I., & P.H. Smith (1992). Intracellular recordings from neurobiotin-labeled cells in brain slices of the rat medial nucleus of the trapezoid body. J. Neurosci., 12:2819-2837. Barnes, W.T., H.W. Magoun, & S.W. Ranson (1943). The ascending auditory pathway in the brain stem of the monkey. J. Comp. Neurol., 79: 129-152. Batra, R., S. Kuwada, & T.R. Stanford (1989). Temporal coding of envelopes and their interaural delays in the inferior colliculus of the unanesthetized rabbit. J. Neurophysiol., 61: 257-268. Batra, R., S. Kuwada, & T.R. Stanford (1993). High-frequency neurons in the inferior colliculus that are sensitive to interaural delays of amplitude-modulated tones: evidence for dual binaural influences. J. Neurophysiol., 70: 64-80.
Auditory Pathway
445
Bechterew, W. (1885). Uber die Verbindungen der oberen Olive und ihre wahrscheinliche physiologishe Bedeutung. Neurol. Zbl., 4: 489490. Bengry, M.F., M.S. Silverman, & B.M. Clopton (1977). Effects of lesioning the dorsal and intermediate acoustic striae on binaural interaction at the inferior colliculus. Exp. Brain Res., 28:211-219. Beyerl, B.D. (1978). Afferent projections to the central nucleus of the inferior colliculus in the rat. Brain Res., 145: 209-223. Bledsoe, S.C., C.R. Snead, R.H. Helfert, V. Prasad, R.J. Wenthold, & R.A. Altschuler (1990). Immunocytochemical and lesion studies support the hypothesis that the projection from the medial nucleus of the trapezoid body to the lateral superior olive is glycinergic. Brain Res., 517: 189-194. Bock, G.R., W.R. Webster, & L.M. Aitkin (1972). Discharge patterns of single units in inferior colliculus of the alert cat. J. Neurophysiol., 35: 265-277. Bormann, J. (1988). Electrophysiology of GABAA and GABAB receptor subtypes. Trends Neurosci., 11:112-116. Boudreau, J.C., & C. Tsuchitani (1968). Binaural interaction in the cat superior olive S-segment. J. Neurophysiol., 3: 442454. Boudreau, J.C., & C. Tsuchitani (1970). Cat superior olive S-segment cell discharge to tonal stimulation. In W.D. Neff (ed.): Contributions to Sensory Physiology. Vol. 4. New York: Academic Press. Boudreau, J.C., & C. Tsuchitani (1973). Sensory Neurophysiology with Special Reference to the Cat. New York: Van Nostrand Reinhold Co. Bourk, T.R., J.P. Mielcarz, & B.E. Norris (1981). Tonotopic organization of the anteroventral cochlear nucleus of the cat. Hearing Res., 4:215-241. Brawer, J.R., D.K. Morest, & E.C. Kane (1974). The neuronal architecture of the cochlear nucleus of the cat. J. Comp. Neurol., 155: 251-300. Broadbent, D.E. (1954). The role of auditory localization in attention and memory span. J. Exp. Psychol., 47: 191-196. Brogden, W.J., E. Girden, F.A. Mettler, & E. Culler (1936). Acoustic value of the several components of the auditory system in cats. Am. J. Physiol., 116:252-261. Brownell, W.E., P.B. Manis, & L.A. Ritz (1979). Ipsilateral inhibitory responses in the cat lateral superior olive. Brain Res., 177: 189-193. Browner, R.H., & D.B. Webster (1975). Projections of the trapezoid body and superior olivary complex of the kangaroo rat (Dipodomys merriami).. Brain, Behav. Evol., 11:322-354. Brugge, J.F., D.J. Anderson, & L.M. Aitkin (1970). Responses of neurons in the dorsal nucleus of the lateral lemniscus of cat to binaural tonal stimulation. J. Neurophysiol., 33:441-458. Brugge, J.F., & C.D. Geisler (1978). Auditory mechanisms of the lower brainstem. Ann. Rev. Neurosci., 1: 363-394.
446
Hutson
Brunso-Bechtold, J.K., & G.C. Thompson (1976). Auditory hindbrain projections to the inferior colliculus as demonstrated by horseradish peroxidase in the cat. Anat. Rec., 184: 365. Brunso-Bechtold, J.K., G.C. Thompson, & R.B. Masterton (1981). HRP study of the organization of auditory afferents ascending to central nucleus of the inferior colliculus in cat. J. Comp. Neurol., 197: 705722. Bumm, A. (1889). Experimenteller Beitrag zur kenntniss des Hornervenursprungs des Kaninchens. Allgem. Zeitschr. f. Psychiat., 45: 568-572. Caird, D., & R. Klinke (1983). Processing of binaural stimuli by cat superior olivary complex neurons. Exp. Brain Res., 52: 385-399. Calford, M.B., & M.N. Semple (1995). Monaural inhibition in cat auditory cortex. J. Neurophysiol., 73:1876-1891. Cant, N.B. (1984). The fine structure of the lateral superior olivary nucleus of the cat. J. Comp. Neurol., 227: 63-77. Cant, N.B, & J.H. Casseday (1986). Projections from the anteroventral cochlear nucleus to the lateral and medial superior olivary nuclei. J. Comp. Neurol., 247: 457-476. Cant, N.B., & R.L. Hyson (1992). Projections from the lateral nucleus of the trapezoid body to the medial superior olivary nucleus in the gerbil. Hearing Res., 58: 26-34. Carney, L.H., & T.C.T. Yin (1989). Response of low-frequency cells in the inferior colliculus to interaural time differences of clicks" Excitatory and inhibitory components. J. Neurophysiol., 62: 144-161. Caspary, D.M., L.P. Rybak, & C.L. Faingold (1985). The effects of inhibitory and excitatory amino-acid neurotransmitters on the response properties of brainstem auditory neurons. In D.G. Drescher (ed.): Auditory Biochemistry. Springfield: Charles C. Thomas. Casseday, J.H., & E. Covey (1983). Laminar organization of projections the inferior colliculus: Evidence from deposits of wheat germ agglutinin-horseradish peroxidase in the superior olive of the cat. Assoc. Res. Otolaryngol., 6: 8. Casseday, J.H., & W.D. Neff (1975). Auditory localization: Role of auditory pathways in brain stem of the cat. J. Neurophysiol., 38: 842858. Casseday, J.H., & H.A. Smoak (1981). Effects of unilateral ablation of anteroventral cochlear nucleus on localization of sounds in space. In J. Syka & L. Aitkin (eds.): Neuronal Mechanisms of Hearing. New York: Plenum Press. Castaldi, L. (1926). Studi sulla struttura e sullo sviluppo del mencecfalo. Ricerche in Cavia cobaya. Parte terza. Arch. Anat. Embriol., 23: 481609. Chan, J.C.K., & T.C.T. Yin (1984). Interaural time sensitivity in the medial superior olive of the cat: comparisons with the inferior colliclus. Neurosci. Abst., 10: 844.
Auditory Pathway
447
Clarey, J.C., P. Barone, & T.J. Imig (1992). Physiology of thalamus and cortex. In A.N. Popper & R.R. Fay (eds.): The Mammalian Auditory Pathway: Neurophysiology. New York: Springer-Verlag. Code, R.A., & J.A. Winer (1986). Columnar organization and reciprocity of commissural connections in cat primary auditory cortex (AI).. Hearing Res., 23: 205-222. Cortez, A.M., G.C. Thompson, & I. Diamond (1983). Sound localization in primates following ablation of auditory cortex. Assoc. Res. Otolaryngol., 6: 32. Covey, E., J.A. Kauer, & J.H. Casseday (1996). Whole-cell patch-clamp recording reveals subthreshold sound-evoked postsynaptic currents in the inferior colliculus of awake bats. J. Neurosci., 16:3009-3018. Curtis, D.R. (1968). Pharmacology and neurochemistry of mammalian central inhibitory processes. In C. von Euler, S. Skoglund & U. Soderberg (eds.): Structure and Function of Inhibitory Neuronal Mechanisms. New York: Pergamon Press. Curtis, D.R., & K. Koisumi (1961). Chemical transmitter substances in brainstem of cat. J. Neurophysiol., 24: 80-90. Davidoff, R.A., & J.C. Hackman (1985). GABA: presynaptic actions. In M.A. Rogawski & J.L. Barker (eds.): Neurotransmitter Actions in the Vertebrate Nervous System. New York: Plenum Press. Davis, H., & L.J. Saul (1931). Action currents in the auditory tracts of the midbrain of the cat. Science, 74: 205. De Vries, H.L. (1948). Brownian movement and heating. Physica 14: 48-60. Diamond, I.T., E.G. Jones, & T.P.S. Powell (1968). Interhemispheric fiber connections of the auditory cortex of the cat. Brain Res., 11: 117-193. Diamond, I.T., E.G. Jones, & T.P.S. Powell (1969). The projection of the auditory cortex upon the diencephalon and brainstem in the cat. Brain Res., 15: 305-340. Dixon, N.T. (1973). Tonotopic organization of the inferior colliculus and the effects of lateral lemniscus lesions on hearing sensitivity in the cat. Doctoral dissertation, Indiana Univ. (See W.D. Neff, et al., 1975). Drake, C.G., & K.G. McKenzie (1953). Mesencephalic tractotomy for pain. J. Neurosurgery, 10: 457-462. Eccles, J.C., R.R. Schmidt, & W.D. Willis (1963). Pharmacological studies on presynaptic inhibition. J. Physiol., 168: 500-530. Elverland, H.H. (1978). Ascending and intrinsic projections of the superior olivary complex in the cat. Exp. Brain Res., 32:117-134. Erulkar, S.D. (1959). The responses of single units of the inferior colliculus of the cat to acoustic stimulation. Proc. Roy. Soc. B., 150: 336-355. Erulkar, S.D. (1975). Physiological studies of the inferior colliculus and medial geniculate complex. In W.D. Keidel & W.D. Neff (eds.):
448
Hutson
Handbook of Sensory Physiology. Vol. 2. Auditory System. New York: Springer-Verlag. Fagg, G.E., & A.C. Foster (1983). Amino acid neurotransmitters and their pathways in the mammalian central nervous system. Neurosci., 9: 701-719. Faingold, C.L., C.A. Boersma Anderson, & D.M. Caspary (1991). Involvement of GABA in acoustically-evoked inhibition in inferior colliculus neurons. Hearing Res., 52: 201-216. Faingold, C.L., C.A. Boersma Anderson, & M.E. Randall (1993). Stimulation or blockade of the dorsal nucleus of the lateral lemniscus alters binaural and tonic inhibition in contralateral inferior colliculus neurons. Hearing Res., 69: 98-106. Faingold, C.L., G. Gehlbach, & D.M. Caspary (1989a). On the role of GABA an an inhibitory neurotransmitter in inferior colliculus neruons: iontophoretic studies. Brain Res., 500:302-312. Faingold, C.L., W.E. Hoffman, & D.M. Caspary (1989b). Effects of excitant amino acids on acoustic responses of inferior colliculus neurons. Hearing Res., 40:127-136. Farley, G.R., B.J. Morley, E. Javel, & M.P. Gorga (1983). Single-unit responses to cholinergic agents in the rat inferior colliculus. Hearing Res., 11: 73-91. Feng, J.J., S. Kuwada, E.-M. Ostapoff, R.Batra, & D.K. Morest (1994). A physiological and structural study of neuron types in the cochlear nucleus. I. Intracellular responses to acoustic stimulation and current injection. J. Comp. Neurol., 346: 1-18. Fernandez, C., & F. Karapas (1967). The course and termination of striae of Monakow and Held in the cat. J. Comp. Neurol., 131: 371386. Ferrier, D. (1876a). The Functions of the Brain. London: Smith, Elder & Co. Ferrier, D. (1876b). The Croonian Lecture: Experiments on the brain of monkeys: Second series. Phil.Trans. Roy. Soc. B., 165: 433-488. Ferrier, D. (1890). The Croonian Lectures on Cerebral Localization. London: Smith, Elder & Co. Ferrier, D., & W.A. Turner (1898). An experimental research upon cerebro-cortical afferent and efferent tracts. Phil.Trans. Roy. Soc. B., 190: 1-44. Fitzpatrick, D.C., S. Kuwada, R. Batra, & C. Trahiotis (1995). Neural responses to simple simulated echoes in the auditory brain stem of the unanesthetized rabbit. J. Neurophysiol., 74: 2469-2486. Flammino, F., & B.M. Clopton (1975). Neural responses in the inferior colliculus of albino rat to binaural stimuli. J. Acoust. Soc. Am., 57: 692-695. Flechsig, P. (1876). Die Leitungsbahnen Im Gehirn Und Ruckenmark. Leipzig: W. Englemann.
Auditory Pathway
449
Fullerton, B.C., & N.Y.s. Kiang (1990). The effect of brainstem lesions on brainstem auditory evoked potentials in the cat. Hearing Res., 49: 363-390. Fulton, J.F. (ed.). (1946). Howell's Textbook of Physiology. Philadelphia: W.B. Saunders Co. Fuse, G. (1911). Striae acusticae von v. Monakow beim Menschen. Eine pathologish-anatomische Beobachtung. Neurol. Zbl., 30:912-917. Fuse, G. (1916). Die Organisation und der V erlauf der akustischen Bahnen. Arb. Hirnanat. Inst. Zurich., 10:61-94. Fuse, G. (1919). Innerer Aufbau der zentralen akustischen Bahnen. Arb. Anat. Inst. Jap. Univ. Sendai, 2: 275-384. Gage, P.W. (1992). Activation and modulation of neuronal K+ channels by GABA. Trends Neurosci., 15:46-51. Galambos, R., J.E. Rose, R.B. Bromiley, & J.R. Hughes (1952). Microelectrode studies on medial geniculate body of cat. II. Response to clicks. J. Neurophysiol., 15: 259-380. Galambos, R., J. Schwartzkopff, & A. Ruppert (1959). Microelectrode study of superior olivary neclei. Am. J. Physiol., 197: 527-536. Glendenning, K.K., & B.N. Baker (1988). Neuroanatomical distribution of receptors for three potential inhibitory neurotransmitters in the brainstem auditory nuclei of the cat. J. Comp. Neurol., 275: 288308. Glendenning, K.K., B.N. Baker, K.A. Hutson, & R.B.Masterton (1992). Acoustic Chiasm V: inhibitory and excitatory neurotrans-mitters in LSO's ipsilateral and contralateral projections. J. Comp. Neurol., 319: 100-122. Glendenning, K.K., J.K. Brunso-Bechtold, G.C. Thompson, & R.B. Masterton (1981). Ascending auditory afferents to the nuclei of the lateral lemniscus. J. Comp. Neurol., 197: 673-703. Glendenning, K.K., K.A. Hutson, R.J. Nudo, & R.B. Masterton (1985). Acoustic chiasm II" anatomical basis of binaurality in lateral superior olive of cat. J. Comp. Neurol., 232:261-285. Glendenning, K.K., & R.B. Masterton (1983). Acoustic chiasm: efferent projections of the lateral superior olive. J. Neurosci., 3:1521-1537. Glendenning, K.K., R.B. Masterton, B.N. Baker, & R.J. Wenthold (1991). Acoustic chiasm III: Nature, distribution, and sources of afferents to the lateral superior olive in the cat. J. Comp. Neurol., 310:377-400. Glendenning, K.K., R.B. Masterton, K.A. Hutson, & R.J. Nudo (1990). Ventral nucleus of the lateral lemniscus: non-tonotopic organization. Anat. Rec., 226: 37A. Goldberg, J.M. (1975). Physiological studies of auditory nuclei of the pons. In W.D. Keidel & W.D. Neff (eds.): Handbook of Sensory Physiology. Vol. 2. Auditory System. New York: Springer-Verlag.
450
Hutson
Goldberg, J.M., & P.B. Brown (1968). Functional organization of the dog superior olivary complex: an anatomical and electrophysiological study. J. Neurophysiol., 31:639-656. Goldberg, J.M., & P.B. Brown (1969). Response of binaural neurons of dog superior olivary complex to dichotic tonal stimuli: some physiological mechanisms of sound localization. J. Neurophysiol., 32:613-636. Goldberg, J.M., & R.Y. Moore (1967). Ascending projection of the lateral lemniscus in the cat and monkey. J. Comp. Neurol., 129: 143156. Goldberg, J.M., F.D. Smith, & H.O. Adrian (1963). Response of single units of the superior olivary complex of the cat to acoustic stimuli: laterality of afferent projections. Anat. Rec., 145: 232. Gordinier, H.C. (1899). The Gross and Minute Anatomy of the Central Nervous System. Philadelphia: P. Blakiston's Son & Co. Guinan, J.J., S.S. Guinan, & B.E. Norris (1972a). Single auditory units in the superior olivary complex. I: Responses to sounds and classifications based on physiological properties. Int. J. Neurosci., 4: 101-120. Guinan, J.J., B.E. Norris, & S.S. Guinan (1972b). Single auditory units in the superior olivary complex. II: Locations of unit catagories and tonotopic organization. Int. J. Neurosci., 4: 147-166. Hammerschlag, V. (1899). Uber die Reflexbewegung des Muskulus tensor tympani und ihre centralen Bahnen. Arch. Ohrenheilk., 47: 251-275. Hammerschlag, V. (1901). Die Lage des Reflexcentrums fur den Muskulus tensor tympany. Arch. Ohrenheilk., 56:157-160. Harris, G.G. (1968). Brownian motion in the cochlear partition. J. Acoust. Soc. Am., 44: 176- 186. Harrison, J.M. (1978). The auditory system of the brain stem. In R.F. Naunton & C. Fernandez (eds.): Evoked Electrical Activity in the Auditory Nernous System. New York: Academic Press. Harrison, J.M., & M.L. Feldman (1970). Anatomical aspects of the cochlear nucleus and superior olivary complex. In W.D. Neff (ed.): Contributions to Sensory Physiology. Vol. 4. New York: Academic Press. Harrison, J.M., & M.E. Howe (1974). Anatomy of the afferent auditory nervous system of mammals. In W.D. Keidel & W.D. Neff (eds.): Handbook of Sensory Physiology. Vol. 1. Auditory System. New York: Springer-Verlag. Harrison, J.M., & R. Irving (1964). Nucleus of the trapezoid body: dual afferent innervation. Science, 143: 473-474. Harrison, J.M., & R. Irving (1966). Ascending connections of the anterior ventral cochlear nucleus in the rat. J. Comp. Neurol., 126: 51-64.
Auditory Pathway
451
Hamson, J.M., & W.B. Warr (1962). A study of the cochlear nuclei and ascending auditory pathways of the medulla. J. Comp. Neurol., 119: 341-379. Heffner, H.E., B.A. Carroll, & D.R. Kovach (1992). Effect of unilateral ablation of auditory cortex on sound localization by macaques. Assoc. Res. Otolarygol., 15: 52. Heffner, H.E., & R.S. Heffner (1986). Effect of unilateral and bilateral auditory cortex lesions on the discrimination of vocalizations by Japanese macaques. J. Neurophysiol., 56: 683- 701. Heffner, H.E., & R.S. Heffner (1989). Unilateral auditory cortex ablation in macaques results in a contralateral hearing loss. J. Neurophysiol., 63:789-801. Heffner, H.E., & R.S. Heffner (1990a). Effect of bilateral auditory cortex lesions on absolute thresholds in Japanese macaques. J. Neurophysiol., 64: 191-205. Heffner, H.E., & R.S. Heffner (1990b). Role of primate auditory cortex in hearing. In W.C. Stebbins & M.A. Berkley (eds.), Comparative Perception, Vol. II. Heffner, H.E., & R.S. Heffner (1994). Lateralization of the perception of communicative vocalizations in Japanese macaques. In J.R. Anderson, J.J. Roeder, B. Thierry, & N. Herrenschmidt (eds.), Current Primatology, Vol. IlL Heffner, R.S., & R.B. Masterton (1990). Sound localization in mammals: brain-stem mechanisms. In M.A. Berkley & W.C. Stebbins (eds.): Comparative Perception Vol. L Basic Mechanisms. Held, H. (1891). Die centralen Bahnen des Nervus acusticus bei der Katze. Arch. f. Anat. u. Physio. Anat. Abtheil., 15:271-290. Held, H. (1893). Die centrale Gehorleitung. Arch. f. Anat. u. Physiol. Anat. Abtheil., 17:201-248. Helfert, R.H., J.M. Bonneau, R.J. Wenthold, & R.A. Altschuler (1989). GABA and glycine immunoreactivity in the guinea pig superior olivary complex. Brain Res., 501: 269-286. Helfert, R.H., J.M. Juiz, S.C. Bledsoe, J.M. Bonneau, R.J. Wenthold, & R.A. Altschuler (1992). Patterns of glutamate, glycine, and GABA immunolabeling in four synaptic terminal classes in the lateral superior olive of the guinea pig. J. Comp. Neurol., 323: 305-325. Helfert, R.H., & I.R. Schwartz (1987). Morphological features of five neuronal classes in the gerbil lateral superior olive. Am. J. Anat., 179: 55-69. Helmholtz, H.L.F. (1875). On the Sensations of Tone. London: Longmans, Green, and Co. (Translation by A.J. Ellis from: Helmholtz, H.L.F. (1870). Die Lehre von dem Tonempfindungen. 3rd ed. Braunschweig: F. Wieweg). Henkel, C.K, & J.K. Brunso-Bechtold (1990). Dendritic morphology and development in the ferret medial superior olivary nucleus. J. Comp. Neurol., 294: 377-388.
452
Hutson
Henkel, C.K., & K.M. Spangler (1983). Organization of the efferent projections of the medial superior olivary nucleus in the cat as revealed by HRP and autoradiographic tracing methods. J. Comp. Neurol., 221: 416-428. Hind, J.E., J.M. Goldberg, D.D. Greenwood, & J.E. Rose (1963). Some discharge characteristics of single neurons in the inferior colliculus of the cat. II. Timing of the discharges and observations on the binaural stimulation. J. Neurophysiol., 26: 337-341. Hofmann, F. (1908). Die obere Olive der Saugetiere nebst Bemerkungen uber die Lage der Cochlearisenkerne. Arb. a. d. neur. Inst. d. Wien., 14: 76-328. Hudspeth, A.J. (1989). How the ear's works work. Nature, 341:397-404. Hugdahl, K. (1995). Dichotic listening: Probing temporal love functional integrity. In R.J. Davidson, & K. Hugdahl (eds.), Brain Asymmetry. Cambridge, MA: MIT Press. Hutson, K.A. (1986). On the nature of the ipsilateral auditory pathway: A critical review. Submitted to the Department of Psychology, Florida State University in partial fulfillment of the requirement for the degree of Doctor of Philosophy. Hutson, K.A. (1988). Connections of the auditory midbrain: efferent projections of the dorsal nucleus of the lateral lemniscus, the nucleus sagulum, and the origins of the GABAergic commissure of Probst. Doctoral Dissertation, Florida State University. Hutson, K.A., & K.K.Glendenning (1995). Projection of nucleus sagulum to the dorsal division of the medial geniculate body in cat. Assoc. Res. Otolaryngol., 18: 41. Hutson, K.A., K.K.Glendenning, B.N. Baker, & R.B. Masterton (1993). GABAergic and glycinergic projections of the auditory midbrain. Neurosci. Abst., 19: 1203. Hutson, K.A., K.K. Glendenning, & R.B. Masterton (1987). Biochemical basis for the acoustic chiasm? Neurosci. Absr, 13: 548. Hutson, K.A., K.K. Glendenning, & R.B.Masterton (1991). Acoustic Chiasm IV: Eight midbrain decussations of the auditory system in cat. J. Comp. Neuro 1., 312:105-131. Hutson, K.A., & D.K. Morest (1996). Fine structure of cell clusters in the cochlear nerve root: Stellate, granule, and mitt cells offer insight into the synaptic organiztion of local circuit neurons. J. Comp. Neurol., 371:397-414 Imig, T.J., & J.F. Brugge (1978). Sources and terminations of callosal axons related to binaural and frequency maps in primary auditory cortex of the cat. J. Comp. Neurol., 182: 637-660. Irvine, D.R.F. (1992). Physiology of the auditory brainstem. In A.N. Popper & R.R. Fay (eds.): The Mammalian Auditory Pathway: Neurophysiology. NewYork: Springer-Verlag.
Auditory Pathway
453
Ito, M., B. Van Adel, & J.B. Kelly (1995). Effect of cutting the commissure of Probst on sound localization and auditory orientation response in the albino rat. Assoc. Res. Otolaryngol., 18: 63. Jane, J.A., R.B. Masterton, & I.T. Diamond (1965). The function of the tectum for attention to acoustic stimuli in the cat. J. Comp. Neurol., 125: 165-192. Jean-Baptiste, M., & D.K. Morest (1975). Transneuronal changes of synaptic endings and nuclear chromatin in the trapezoid body following cochlear ablations in cats. J. Comp. Neurol., 162:111-134. Jeffress, L.A. (1948). A place theory of sound localization. J. Comp. Physiol. Psychol., 41: 35-39. Jenkins, W.M., & R.B. Masterton (1982). Sound localization: effects of unilateral lesions in central auditory system. J. Neurophysiol., 47: 987-1016. Jenkins, W.M., & M.M. Merzenich (1984). Role of cat primary auditory cortex for sound-localization behavior. J. Neurophysiol., 52: 819847. Johnston, G.A.R. (1976). Amino acid inhibitory transmitters in the central nervous system. In C.H. Hockman & D. Bieger (eds.): Chemical Transmission in the Mammalian Central Nervous System. Baltimore: University Park Press. Jones, D.R., K.A. Hutson, & D.K. Morest (1992). Growth cones and structural variation of synaptic end-bulbs in the cochlear nucleus of the adult cat brain. SYNAPSE, 10:291-309. Jungert, S. (1958). Auditory pathways in the brain stem. Acta OtoLaryngol. Suppl., 138: 1-67. Kane, E.S., & L.M. Barone (1980). The dorsal nucleus of the lateral lemniscus in the cat: neuronal types and their distributions. J. Comp. Neurol., 192: 797-826. Kavanagh, G.L., & J.B. Kelly (1987). Contribution of auditory cortex to sound localization by the ferret (Mustela putorius). J. Neurophysiol., 57: 1746-1766. Kavanagh, G.L., & J.B. Kelly (1992). Midline and lateral field sound localization in the ferret (Mustela putorius): Contribution of the superior olivary complex. J. Neurophysiol., 67: 1643-1658. Kelly, J.B., & L. Li (1997). Two sources of inhibition affecting binaural evoked responses in the rat's inferior colliculus: the dorsal nucleus of the lateral lemniscus and the superior olivary complex. Hearing Res., 104:112-126. Kiang, N. Y-s. (1963). Spontaneous activity of single auditory nerve fibers in cats. J. Acoust. Soc. Am., 35: 793. Kidd, S.A., & J.B. Kelly (1996). Contribution of the dorsal nucleus of the lateral lemniscus to binaural responses in the inferior colliculus of the rat: interaural time delays. J. Neurosci., 16: 7390-7397. Kimura, D. (1961). Some effects of temporal-lobe damage on auditory perception. Can. J. Psychol., 15:156-165.
454
Hutson
Kiss, A., & K. Majorossy (1983). Neuron morphology and synaptic architecture in the medial superior olivary nucleus. Exp. Brain Res., 52:315-327. Kitzes, L.M., & M.N. Semple (1985). Single-unit responses in the inferior colliculus: effects of neonatal unilateral cochlear ablation. J. Neurophysiol., 53: 1483-1500. Klug, A., T.J. Park, & G.D. Pollak (1995). Glycine and GABA influence binaural processing in the inferior colliculus of the mustache bat. J. Neurophysiol., 74:1701-1713. Kolliker, A. (1896). Handbuch der Gewebelehre des Menschen. Vol. 2. Leipzig: W. Engelmann. Kreidl, A. (1914). Zur Frage der sekundaren Horbahnen. Monats-schr. f. Ohrenheilk. u. Laryn. Rhino., 48: 1-14. Kryter, K.D., & H.W. Ades (1943). Studies on the function of the higher acoustic nervous centers in the cat. Am. J. Psychol., 56: 501-536. Kudo, M. (1981). Projections of the nuclei of the lateral lemniscus in the cat: an autoradiographic study. Brain Res., 221: 57-69. Kudo, M., Y. Nakamura, T. Moriisumi, H. Tokuno, & Y. Kitao (1988). Bilateral projections from the medial superior olivary nucleus to the inferior colliculus in the mole (Mogera robusta).. Brain Res., 463: 352-356. Kudo, M., & K. Niimi (1978). Ascending projections of the inferior colliculus onto the medial geniculate body in the cat studied by anterograde and retrograde tracing techniques. Brain Res., 155:113117. Kudo, M., & K. Niimi (1980). Ascending projections of the inferior colliculus in the cat: an autoradiographic study. J. Comp. Neurol., 191: 545-556. Kuwabara, N., R.A. DiCaprio, & J.M. Zook (1991). Afferents to the medial nucleus of the trapezoid body and their collateral projections. J. Comp. Neurol., 314: 684-706. Kuwabara, N., & J.M. Zook (1991). Classification of the principal cells of the medial nucleus of the trapezoid body. J. Comp. Neurol., 314: 707-720. Kuwada, S., R. Batra, & T.R. Stanford (1989). Monaural and binaural response properties of neurons in the inferior colliculus of the rabbit: effects of sodium pentobarbital. J. Neurophysiol., 61: 269-282. Kuwada, S., T.R. Stanford, & R. Batra (1987). Interaural phase-sensitive units in the inferior colliculus of the unanesthetized rabbit: effects of changing frequency. J. Neurophysiol., 57: 1338-1360. Kuwada, S., T.C.T. Yin, J. Syka, T.J.F. Buunen, & R.E. Wickesberg (1984). Binaural interactions in low-frequency neurons in inferior colliculus of the cat. IV. Comparison of monaural and binaural response properties. J. Neurophysiol.,51: 1306-1325. La Villa, I. (1898). Algunos detalles concernientes a la oliva superior y focos acusticos. Rev. trimest, microgr., 3: 75-83.
Auditory Pathway
455
Lewy, F.H. (1909). Degenerationsversuche am akustischen System des Kaninchens und der Katze. Folia Neurobiol., 2:471-518. Liberman, M.C. (1978). Auditory-nerve responses from cats raised in a low-noise chamber. J. Acoust. Soc. Am., 63: 442-455. Lorente De No, R. (1933). Anatomy of the eighth nerve. III. General plan of structure of the primary cochlear nuclei. Laryngoscope, 43: 327-350. Luciani, L. (1884). On the sensorial localizations in the cortex cerebri. Brain, 7: 145-160. Markovitz, N.S., & G.D. Pollak (1994). Binaural processing in the dorsal nucleus of the lateral lemniscus. Hearing Res., 73: 121-140. Masterton, R.B., & M.A. Berkley (1974). Brain function: Changing ideas on the role of sensory, motor, and association cortex in behavior. Ann. Rev. Psychol., 25:277-312. Masterton, R.B., K.K. Glendenning, & R.J. Nudo (1981). Anatomicalbehavioral analyses of hindbrain sound localization. In J. Syka & L. Aitkin (eds.): Neuronal Mechanisms of Hearing. New York: Plenum Press. Masterton, R.B., E.M. Granger, & K.K. Glendenning (1992). Psychoacoustical contributions of each lateral lemniscus. Hearing Res., 63: 57-70. Masterton, R.B., J.A. Jane, & I.T. Diamond (1967). Role of brainstem auditory structures in sound localization. I. Trapezoid body, superior olive, and lateral lemniscus. J. Neurophysiol., 30: 341-359. McGeer, P.L., & E.G. McGeer (1981). Amino acid neurotrans-mitters. In G.J. Siegel, R.W. Albers, B.W. Agranoff & R. Katzman (eds.): Basic Neurochemistry. 3rd. Ed. Boston: Little, Brown & Co. Merzenich, M.M., & M.D. Reid (1974). Representation of the cochlea within the inferior colliculus of the cat. Brain Res., 77: 397-415. Mettler, F.A., G. Finch, E. Girden, & E. Culler (1934). Acoustic value of the several components of the auditory pathway. Brain, 57: 475-483. Meyer, S. (1896). Uber eine Verbindungsweise der Neuronen. Arch. mikr. Anat., 47: 734-748. Middlebrooks, J.C., & J.D. Pettigrew (1981). Functional classes of neurons in primary auditory cortex of the cat distinguished by sensitivity to sound location. J. Neurosci., 1: 107-120. Middlebrooks, J.C., & J.M. Zook (1983). Intrinsic organization of the cat's medial geniculate body identified by projections to binaural response specific bands in the primary auditory cortex. J. Neurosci., 3: 203-224. Milner, B., L. Taylor, & R.W. Sperry (1968). Lateralized suppression of dichotically presented digits after commissural section in man. Science, 161: 184-186. Monakow, C. von (1887). Uber den Ursprung und den centralen Verlauf des Acusticus. Correspondenzbl. f. Schweizer Aertze, 5: 145-146.
456
Hutson
Monakow, C. von (1890). Striae acusticae und untere Schleife. Arch. f. Psychiat., 22: 1-26. Monakow, C. von (1895). Experimentelle und pathologisch-anatomisch Untersuchungen uber die Haubenregion, den Sehhugel und die Regio subthalamica, nebst Beitragen zur Kenntniss fruh erworbener Gross- und Kleinhirn-defekte. Arch. Psychiat. Nervenkr., 27: 386478. Moore, C.N., J.H. Casseday, & W.D. Neff (1974). Sound localization: the role of the commissural pathways of the auditory system of the cat. Brain Res., 82: 13-26. Moore, D.R., & L.M. Kitzes (1985). Projections from the cochlear nucleus to the inferior colliculus in normal and neonatally cochleaablated gerbils. J. Comp. Neurol., 240: 180-195. Moore, D.R., M.N. Semple, P.D. Addison, & L.M. Aitkin (1984). Properties of spatial receptive fields in the central nucleus of the cat inferior colliculus. I. Response to tones of low intensity. Hearing Res., 13: 159-174. Moore, J.K. (1987). The human auditory brain stem: a comparative review. Hearing Res., 29: 1-32. Moore, J.K., & R.Y. Moore (1971). A comparative study of the superior olivary complex in the primate brain. Folia primat., 16: 35-51. Moore, J.K., & R.Y. Moore (1987). Glutamic acid decarboxylase-like immunoreactivity in brainstem auditory nuclei of the rat. J. Comp.
Neurol., 260:157-174. Moore, M.J., & D.M. Caspary (1983). Strychnine blocks binaural inhibition in lateral superior olivary neurons. J. Neurosci., 3: 237-242. Moore, R.Y., & J.M. Goldberg (1963). Ascending projections of the inferior colliculus in the cat. J. Comp. Neurol., 121: 109-135. Morel, A., & T.J. Imig (1987). Thalamic projections to fields A, AI, P, and VP in the cat auditory cortex. J. Comp. Neurol., 265:119-144. Morest, D.K. (1968). The collateral system of the medial nucleus of the trapezoid body of the cat, its neuronal architecture and relation to the olivo-cochlear bundle. Brain Res., 9:288-311. Morest, D.K., K.A. Hutson, & S. Kwok (1990). Cytoarchitectonic atlas of the cochlear nucleus of the chinchilla, Chinchilla laniger. J. Comp. Neurol., 300, 230-248. Morest, D.K., & D.L. Oliver (1984). The neuronal architecture of the inferior colliculus in the cat: defining the functional anatomy of the auditory midbrain. J. Comp. Neurol., 222: 209-236. Mugnaini, E., W.B. Warr, & K.K. Osen (1980). Distribution and light microscopic features of granule cells in the cochlear nuclei of cat, rat, and mouse. J. Comp. Neurol., 191:581- 606. Neff, W.D., & J.H. Casseday (1977). Effects of unilateral ablation of auditory cortex on monaural cat's ability to localize sound. J. Neurophysiol., 40: 44-52.
Auditory Pathway
457
Neff, W.D., & I.T. Diamond (1958). The neural basis of auditory discrimination. In H.F. Harlow & C.N. Woolsey (eds.): Biological and Biochemical Bases of Behavior. Madison: Univ. of Wisconsin Press. Neff, W.D., I.T. Diamond, & J.H. Casseday (1975). Behavioral studies of auditory discrimination: central nervous system. In W.D. Keidel & W.D. Neff (eds.): Handbook of Sensory Physiology. Vol. 2. Auditory System. New York: Springer-Verlag. Nelson, P.G., & S.D. Erulkar (1963). Synaptic mechanisms of excitation and inhibition in the central auditory pathway. J. Neurophysiol., 26: 908-923. Niemer, W.T., & S.K. Cheng (1949). The ascending auditory system, a study of retrograde degeneration. Anat. Rec., 103:116. Nordeen, K.W., H.P. Killackey, & L.M. Kitzes (1983). Ascending auditory projections to the inferior colliculus in the adult gerbil, Meriones unguiculatus. J. Comp. Neurol., 214:131-143. Nudo, R.J., & R.B Masterton (1984). 2-deoxyglucose studies of stimulus coding in the brainstem auditory system of the cat. Contributions to Sensory Physiology, 8: 79-97. Ohnisi, G. (1932). Uber die Fasern, die aus der Lateralschleife zentralwarts emporsteigen oder absteigend in derselben endigen. Zbl. Neurol. Psychiat., 63: 580. Oliver, D.L. (1984). Dorsal cochlear nucleus projections to the inferior colliculus in the cat: a light and electron microscopic study. J. Comp. Neurol., 224:155-172. Oliver, D.L. (1985). Quantitative analysis of axonal endings in the central nucleus of the inferior colliculus and distribution of 3H_ labeling after injections in the dorsal cochlear nucleus. J. Comp. Neurol., 237: 343-359. Oliver, D.L. (1987). Projections to the inferior colliculus from the anteroventral cochlear nucleus in the cat: possible substrates for binaural interactions. J. Comp. Neurol., 264: 24-46. Oliver, D.L., & G.E. Beckius (1992). Fine structure of GABA-labeled axonal endings in the inferior colliculus of the cat: immunocytochemistry on deplasticized ultathin sections. Neuroscience, 46: 455-463. Oliver, D.L., S. Kuwada, T.C.T. Yin, L.B. Haberly, & C.K. Henkel (1991). Dendritic and axonal morphology of HRP-injected neurons in the inferior colliculus of the cat. J. Comp. Neurol., 303: 75-100. Oliver, D.L., & D.K. Morest (1984). The central nucleus of the inferior colliculus in the cat. J. Comp. Neurol., 222: 237-264. Oliver, D.L., J.A. Winer, G.E. Beckius, & R.L. Saint Marie (1994). Morphology of GABAergic neurons in the inferior colliculus of the cat. J. Comp. Neurol., 340: 27-42. Osen, K.K. (1969a). Cytoarchitecture of the cochlear nuclei in the cat. J. Comp. Neurol., 136: 453-484.
458
Hutson
Osen, K.K. (1969b). The intrinsic organization of the cochlear nuclei in the cat. Acta Oto-laryngol., 67: 352-359. Osen, K.K. (1972). Projections of the cochlear nuclei on the inferior colliculus in the cat. J. Comp. Neurol., 144: 3:55-372. Ostapoff, E.-M., C.G. Benson, & R.L. Saint Marie (1997). GABA- and glycine- immunoreactive projections from the superior olivary complex to the cochlear nucleus in guinea pig. J. Comp. Neurol., 381:500-512. Pantev, C., B. Lutkenhoner, M. Hoke, & K. Lehnertz (1986). Comparison between simultaneously recorded auditory-evoked magnetic fields and potentials elicited by ipsilateral, contralateral and binaural tone burst stimulation. Audiology, 25: 54-61. Papez, J.W. (1929a). Comparative Neurology. New York: T.Y. Cromwell Co. (Reprinted 1967; New York: Hafner Publ. Co.). Papez, J.W. (1929b). Central acoustic tract in cat and man. Anat. Rec., 41: 60. Papez, J.W. (1930). Superior olivary nucleus; its fiber connections. Arch. Neurol. Psychiat., 24: 1-20. Parham, K., & D.O. Kim (1995). Spontaneous and sound-evoked discharge characteristics of complex-spiking neurons in the dorsal cochlear nucleus of the unanesthetized cat. J. Neurophysiol., 73: 550-561. Park, T.J., & G.D. Pollak (1993). GABA shapes sensitivity to interaural intensity disparities in the mustache bat's inferior colliculus: Implications for encoding sound location. J. Neurosci., 13: 2050-2067. Penfield, W., & J. Evans (1934). Functional defects produced by cerebral lobectomies. Proc. Assoc. Res. Nerv. Ment. Dis., 13: 352-377. Phillips, D.P., & J.F. Brugge (1985). Prgress in neurophysiology of sound localization. Ann. Rev. Psychol., 36: 245-274. Poirier, P., M. Lassonde, J.-G. V illemure, G. Geoffroy, & F. Lepore (1994). Sound localization in hemispherectomized patients. Neuropsychologia, 32: 541-553. Poljak, S. (1925). The connections of the acoustic nerve. J. Anat., 60: 465-469. Powell, T.P.S., & S.D. Erulkar (1962). Transneuronal cell degeneration in the auditory relay nuclei of the cat. J. Anat., 96: 249-268. Prechstedt, P.H. (1986). Suppression of ipsilateral auditory pathways increases with increasing task load in commussurotomy subjects. Bull. Clin. Neurosci., 51: 73-76. Probst, M. (1902). Experimentelle Untersuchungen uber die Anatomie und Physiology der Leitungsbahnen des Gehirnstammes. Arch. f. Anat. Physiol. Suppl., 47: 254. Ramon y Cajal, S. (1896). Beitrage zum Studium der Medulla Oblongata, des Kleinhirns und Ursprungs der Gehirnnerven. Leipzig: Barth.
Auditory Pathway
459
Ramon y Cajal, S. (1909). Histologie du System Nerveux de L'Homme et des Vertebres. Vol. 1. Paris: A. Moloine. [Reprinted 1952; Madrid: Instituto Ramon y Cajal. Translation by L. Azoulay from: Ramon y Cajal, S (1904). Textura de Sistema Nervioso del Hombre y de los Vertebrados. Madrid]. Ranson, S.W., & S.L. Clark (1947). The Anatomy of the Nervous System. Philadelphia: W.B. Saunders Co. Rasmussen, G.L. (1946). The olivary peduncle and other fiber projections of the superior olivary complex. J. Comp. Neurol., 84: 141-219. Rasmussen, G.L. (1967). Efferent connections of the cochlear nucleus. In A.B. Graham (ed.): Sensorineural Hearing Processes and Disorders. Boston: Little, Brown & Co. Ravizza, R.J., & S.M. Belmore (1978). Auditory forebrain: evidence from anatomical and behavioral experiments involving human and animal subjects. In R.B. Masterton (ed.): Handbook of Behavioral Neurobiology. Vol. 1. Sensory Integration. New York: Plenum Press. Rhode,W.S., & P.H. Smith (1986). Encoding timing and intesity in the ventral cochlear nucleus of the cat. J. Neurophysiol., 56:261-286. Roberts, R.C., & C.E. Ribak (1987). GABAergic neurons and axonal terminals in the brainstem auditory nuclei of the gerbil. J. Comp. Neurol., 258: 267-280. Rockel, A.J., & E.G. Jones (1973a). The neuronal organization of the inferior colliculus of the adult cat. I. The central nucleus. J. Comp. Neurol., 147:11-60. Rockel, A.J., & E.G. Jones (1973b). Observations on the fine structure of the central nucleus of the inferior colliculus of the cat. J. Comp. Neurol., 147: 61-92. Rose, J.E. (1960). Organization of frequency sensitive neurons in the cochlear nuclear complex of the cat. In G.L.Rasmussen & W.F. Windle (eds.): Neuronal Mechanisms of the Auditory and Vestibular Systems. Springfield: Thomas. Rose, J.E., R. Galambos, & J.R. Hughes (1959). Microelectrode studies of the cochlear nuclei of the cat. Bull. Johns Hopkins Hosp., 104: 211-251. Rose, J.E.,D.D. Greenwood, J.M. Goldberg, & J.E. Hind (1963). Some discharge characteristics of single neurons in the inferior colliculus of the cat. I. Tonotopical organization, relation of spike-counts to tone intensity, and firing patterns of single elements. J. Neurophysio., 26:314-319. Rose, J.E., N.B. Gross, C.D. Geisler, & J.E. Hind (1966). Some neural mechanisms in the inferior colliculus of the cat which may be relevant to localization of a sound source. J. Neurophysiol., 29: 288314. Rosenzweig, M.R. (1951). Representations of the two ears at the auditory cortex. Am. J. Physiol., 167: 147-158.
460
Hutson
Rosenzweig, M.R. (1961). Development of research on the physiological mechanisms of auditory localization. Psych. Bull., 58: 376-389. Rosenzweig, M.R., & A.H. Amon (1955). Binaural interaction in the medulla of the cat. Experientia, 11: 498-500. Rosenzweig, M.R., & D. Sutton (1958). Binaural interaction in the lateral lemniscus of the cat. J. Neurophyiol., 21: 17-23. Rosenzweig, M.R., & E.J. Wyers (1955). Binaural interaction at the inferior colliculus. J. Comp. Physiol. Psych., 48:426-431. Roth, G.L., L.M. Aitkin, R.A. Anderson, & M.M. Merzenich (1978). Some features of the spatial organization of the central nucleus of the inferior colliculus of the cat. J. Comp. Neurol., 182: 661-680. Rouiller, E.M., & F. de Ribaupierre (1985). Origin of afferents to physiologically defined regions of the medial geniculate body of the cat: ventral and dorsal divisions. Hearing Res., 19: 97-114. Rouiller, E.M., & D.K. Ryugo (1984). Intracellular marking of physiologically characterized cells of the ventral cochlear nucleus of the cat. J. Comp. Neurol., 225:167-186. Ryan, A., & J. Miller (1978). Single unit responses in the inferior colliculus of the awake and performing rhesus monkey. Exp. Brain Res., 32: 389-407. Ryugo, D.K., & E.M. Rouiller (1988). Central projections of intracellulary labeled auditory nerve fibers in cats: morphometric correlations with physiological properties. J. Comp. Neurol., 271: 130-142. Ryugo, D.K., F.H. Willard, & D.M. Fekete (1981). Differential projections to the inferior colliculus from ~the cochlear nucleus in the albino mouse. Brain Res., 210: 342-349. Sabin, F.R. (1897). On the anatomical relationships of the nuclei of reception of the cochlear and vestibular nerves. Bull. Johns Hopkins Hosp., 8: 253-259. Saint Marie, R.L., & R.A. Baker (1990). Neurotransmitter-specific uptake and retrograde transport of [3H]glycine from the inferior colliculus by ipsilateral projections of the superior olivary complex and nuclei of the lateral lemniscus. Brain Res., 524: 244-253. Sanchez-Longo, L.P., & F.M. Forster (1958). Clinical significance of impairment of sound localization. Neurology, 8:119-125. Sanes, D.H., W.A. Geary, & G.E. Wooten (1985). The quantitative distribution of 3H-strychnine binding in the lateral superior olivary nucleus of the gerbil. Neurosci. Abst., 11:1051. Saul, L.J., & H. Davis (1932). Action currents in the central nervous system: I. Action currents of the auditory tracts. Arch. Neurol. Psychiat., 28:1104-1116. Scheibel, M.E., & A.B. Scheibel (1974). Neuropil organization of the superior olive of the cat. Exp. Neurol., 43: 339-348.
Auditory Pathway
461
Schofield, B.R., & N.B. Cant (1992). Organization of the superior olivary complex in the guinea pig: II. Patterns of projections from the periolivary nuclei to the inferior colliculus. J. Comp. Neurol., 317: 438-455. Schwartz, J., & P. Tallal (1980). Rate of acoustic change may underlie hemispheric specialization for speech perception. Science, 207: 13801381 Semple, M.N., & L.M. Aitkin (1979). Representation of sound frequency and laterality by units in central nucleus of cat inferior colliculus. J. Neurophysiol., 42: 1626-1639. Semple, M.N., & L.M. Aitkin (1980). Physiology of pathway from dorsal cochlear nucleus to inferior colliculus revealed by electrical and auditory stimulation. Exp. Brain Res., 41: 19-28. Semple, M.N., & L.M. Aitkin (1981). Integration and segregation of input to the cat inferior colliculus. In J. Syka & L. Aitkin (eds.): Neuronal Mechanisms of Hearing. New York: Plenum Press. Semple, M.N., & L.M. Kitzes (1985). Single unit responses in the inferior colliculus: differential consequences of contralateral and ipsilateral stimulation. J. Neurophysiol., 53: 1467-1482. Shaner, R.F. (1934). The development of the nuclei and tracts related to the acoustic nerve in the pig. J. Comp. Neurol., 60: 5-19. Shneiderman, A., & C.K. Henkel (1987). Banding of lateral superior olivary nucleus afferents in the inferior colliculus: a possible substrate for sensory integration. J. Comp. Neurol., 266:519-534. Shneiderman, A., D.L. Oliver, & C.K. Henkel (1988). Connections of the dorsal nucleus of the lateral lemniscus: an inhibitory parallel pathway in the ascending auditory system? J. Comp. Neurol., 276: 188-208. Shneiderman, A., & D.L. Oliver (1989). EM autoradiographic study of the projections from the dorsal nucleus of the lateral lemniscus: a possible source of inhibitory inputs to the inferior colliculus. J. Comp. Neurol., 286: 28-47. Shneiderman, A., M.B. Chase, J.M. Rockwood, C.G. Benson, & S.J. Potashner (1993). Evidence for a GABAergic projection from the dorsal nucleus of the lateral lemniscus to the inferior colliculus. J. Neurochem., 60: 72-82. Sivian, L.J., & S.D. White (1933). On minimum audible sound fields. J. Acoust. Soc. Am., 4:288-321. Smith, P.H. (1995). Structural and functional differences distinguish principle from nonprinciple cells in the guinea pig MSO slice. J. Neurophysiol., 73:1653-1667. Smith, P.H., P.X. Joris, L.H. Carney, & T.C.T. Yin (1991). Projections of physiologically characterized globular bushy cell axons from the cochlear nucleus of the cat. J. Comp. Neurol., 304: 387-407. Smith, P.H., P.X. Joris, & T.C.T. Yin (1993). Projections of physiologically characterized spherical bushy cell axons from the
462
Hutson
cochlear nucleus of the cat: evidence for delay lines to the medial superior olive. J. Comp. Neurol., 331: 245-260. Spangler, K.M., W.B. Warr, & C.K. Henkel (1985). The projections of principal cells of the medial nucleus of the tapezoid body in the cat. J. Comp. Neurol., 238: 249-262. Spirou, G.A., W.E. Brownell, & M. Zidanic (1984). Response properties of trapezoid body fibers. Neurosci. Abst., 10:844. Spirou, G.A., W.E. Brownell, & M. Zidanic (1990). Recordings from cat trapezoid body and HRP labeling of globular bushy cell axons. J. Neurophysiol., 63:1169-1190. Spitzer, M.W., & M.N. Semple (1995). Neurons sensitive to interaural phase disparity in gerbil superior olive: diverse minaural and temporal response properties. J. Neurophysiol., 73: 1668-1690. Stanford, T. R., S. Kuwada, & R. Batra (1992). A comparison of the interaural time sensitivity of neurons in the inferior colliculus and thalamus of unanesthetized rabbits. J. Neurosci., 12: 3200-3216. Starr, A., & R.B. Livingston (1963).. Long-lasting nervous system responses to prolonged sound stimulation in waking cats. J. Neurophysiol., 26:416-431. Stevens, S.S., & H. Davis (1938). Hearing: Its Psychology and Physiology. New York: J. Wiley & Sons. Stokes, J.H. (1912). The acoustic complex and its relation in the brain of the opossum (Didelphis virginiana). Am. J. Anat., 12: 401-445. Stotler, W.A. (1949). The projection of the cochlear nerve on the acoustic relay nuclei of the medulla. Anat. Rec., 103: 561. Stotler, W.A. (1953). An experimental study of the cells and connections of the superior olivary complex of the cat. J. Comp. Neurol., 98:401-431. Strominger, N.L. (1969). Localization of sound in space after unilateral and bilateral ablation of auditory cortex. Exp. Neurol., 25: 521-533. Strominger, N.L. (1973). The origins, course and distribution of the dorsal and intermediate acoustic striae in the rhesus monkey. J. Comp. Neurol., 147: 209-234. Strominger, N.L., L.R. Nelson, & W.J. Dougherty (1977). Second order auditory pathways in the chimpanzee. J. Comp. Neurol., 172: 349366. Strominger, N.L., & R.E. Oesterreich (1970). Localization of sound after section of the brachium of the inferior colliculus. J. Comp. Neurol., 138: 1-18. Strominger, N.L., & A.I. Strominger (1971). Ascending brainstem projections of the anteroventral cochlear nucleus in the rhesus monkey. J.Comp. Neurol., 143: 217-242. Strong, O.S., & A. Elwyn (1953). Human Neuroanatomy. Baltimore: The Williams & Wilkins Co. Stumpf, C. (1926). Die Sprachlaute Experimentell-Phonetische Untersuchungen. Berlin: Springer.
Auditory Pathway
463
Thompson, S.P. (1878). Phenomena of binaural audition. Phil. Mag., 6: 383-391. Thompson, G.C., & A.M. Cortez (1983). The inability of squirrel monkeys to lacalize sound after unilateral ablation of auditory cortex. Behav. Brain Res., 8:211-216. Thompson, G.C., A.M. Cortez, & D.M.K. Lam (1985). Localization of GABA immunoreactivity in the auditory brainstem of guinea pigs. Brain Res., 339:119-122. Thompson, G.C., & A.M. Thompson (1986). Olivocochlear neurons in the squirrel monkey brainstem. J. Comp. Neurol., 254: 246-258. Thompson, A.M., & G.C. Thompson (1991). Projections from the posteroventral cochlear nucleus to the superior olivary complex in guinea pig: Light and EM observations with PHA-L method. J. Comp. Neurol., 311" 495-508. Tolbert, L.P., D.K. Morest, & D.A. Yurgelun-Todd (1982). The neuronal architecture of the anteroventral cochlear nucleus of the cat in the region of the cochlear nerve root: Horseradish peroxidase labeling of identified cell types. Neurosci., 7:3031-3052. Tschermak, A. (1899). Uber die Folgen der Durchschneidung des Trapezkorpers bet der Katze. Neurol. Zbl., 18:674-731. Tsuchitani, C. (1977). Functional organization of lateral cell groups of cat superior olivary complex. J. Neurophysiol., 40:296-318. Tsuchitani, C. (1982). Discharge patterns of cat lateral superior olivary units to ipsilateral tone-burst stimuli. J. Neurophysiol., 47: 479-500. Tsuchitani, C. (1997). Input from the medial nucleus of the trapezoid body to an interaural level detector. Hearing Res., 105:211-224. Tsuchitani, C., & J.C. Boudreau (1969). Stimulus level of dichotically presented tones and cat superior olive S-segment cell discharge. J. Acoust. Soc. Am., 46: 979-988. Tunturi, A.R. (1946). A study on the pathway from the medial geniculate body to the acoustic cortex in the dog. Am. J. Physiol., 147:311-319. Turner, W.A., & W. Hunter (1899). On a form of nerve termination in the central nervous system, demonstrated by methylene blue. Brain, 22: 123-135. Van Gehuchten, A. (1902). Recherches sur la voie acoustique centrale. Le Nevraxe, 4: 253-300. Van Gehuchten, A. (1903). Anatomie Du Systeme Nerveux De L~ Louvain" Trois Rois. Van Noort, J. (1969). The Structure and Connections of the Inferior Colliculus. An Investigation of the Lower Auditory System. Assen: Van Gorcum Co. Vater, M., J.H. Casseday, & E. Covey (1995). Convergence and divergence of ascending binaural and monaural pathways from the superior olives of the mustache bat. J. Comp. Neurol., 351: 632-646.
464
Hutson
Vater, M., & A.S. Feng (1990). Functional organization of ascending and descending connections of the cochlear nucleus of horseshoe bats. J. Comp. Neurol., 292: 373-395. Veratti, E. (1900). Su Particolarita di Struttura dei Centri Acustici ne Mammiferi. Pavia: Tip. e Legatoria Cooperativa. Verhaart, W.J.C. (1970). Comparative Anatomical Aspects of the Mammalian Brain Stem and the Cord. Assen: Van Gorcum. Vetter, D.E., J.C. Adams, & E. Mugnaini (1991). Chemically distinct rat olivocochlear neurons. SYNAPSE, 7: 21-43. V incenzi, L. (1900). Nuove ricerche sui calici di Held nel nucleo del corpo trapezoide. Anat. Anz., 18: 344-348. Walker, A.E. (1942). Relief of pain by mesencephalic tractotomy. Arch. Neurol. Psychiat., 48: 865-883. Warr, W.B. (1966). Fiber degeneration following lesions in the anterior ventral cochlear nucleus of the cat. Exp. Neurol., 14: 453-474. Warr, W.B. (1969). Fiber degeneration following lesions in the posteroventral cochlear nucleus of the cat. Exp. Neurol., 23: 140-155. Warr, W.B. (1972). Fiber degeneration following lesions in the multipolar and globular cell areas in the ventral cochlear nucleus of the cat. Brain Res., 40: 247-270. Warr, W.B. (1975). Olivocochlear and vestibular efferent neurons of the feline brain stem: their location, morphology and number determined by retrograde axonal transport and acetylcholinesterase histochemistry. J. Comp. Neurol., 161: 159-182. Warr, W.B. (1982). Parallel ascending pathways from the cochlear nucleus: neuroanatomical evidence for functional specialization. In W.D. Neff (ed.): Contributions to Sensory Physiology. Vol. 7. New York: Academic Press, Inc. Wart, W.B., & J.E. Beck (1996). Multiple projections from the ventral nucleus of the trapezoid body in the rat. Hearing Res., 93: 83-101. Watanabe, T., & Z. Simada (1971). Picrotoxin: effects on collicular auditory neurons. Brain Res., 28: 582-585. Watanabe, T., & Z. Simada (1973). Pharmacological properties of cat's collicular auditory neurons. Jap. J. Physiol., 23: 291-308. Webster, W.R., J. Serviere, & M. Brown (1984). Inhibitory contour in the inferior colliculus as revealed by the 2-deoxyglucose method. Exp. Brain Res., 56:577-581. Webster, W.R., J. Serviere, R. Martin, & M. Brown (1985). Uncrossed and crossed inhibition in the inferior colliculus of the cat: a combined 2-deoxyglucose and electrophysiological study. J. Neurosci., 5:1820-1832. Wenthold, R.J., D. Huie, R.A. Altschuler, & K.A. Reeks (1987). Glycine immunoreactivity localized in the cochlear nuleus and superior olivary complex. Neuroscience, 22: 897-912.
Auditory Pathway
465
Wenthold, R.J., M.H. Parakkal, M.D. Oberdorfer, & R.A. Altschuler (1988). Glycine receptor immunoreactivity in the ventral cochlear nucleus of the guinea pig. J. Comp. Neurol., 276: 423-435. West, C.D. (1970). The convolutional organization of the lateral superior olivary nucleus. J. Aud. Res., 10: 205-209. White, J.S., & W.B. Warr (1983). The dual origins of the olivocochlear bundle in the albino rat. J. Comp. Neurol., 219: 203-214. Whitley, J.M., & C.K. Henkel (1984). Topographical organization of the inferior collicular projection and other connections of the ventral nucleus of the lateral lemniscus in the cat. J. Comp. Neurol., 229: 257-270. Willard, F.H., & G.F. Martin (1983). The auditory brainstem nuclei and some of their projections to the inferior colliculus in the North American opossum. Neurosci., 10:1203-1232. Willard, F.H., & D.K. Ryugo (1983). Anatomy of the central auditory system. In J.F. Willott (ed.).: The Auditory Psychobiology of the Mouse. Springfield: Thomas Pub. Winer, J.A., I.T. Diamond, & D. Raczkowski (1977). Subdivisions of the auditory cortex of the cat: the retrograde transport of horseradish peroxidase to the medial geniculate body and posterior thalamic nuclei. J. Comp. Neurol., 176: 387-418. Winkler, C. (1911). Experimenteller Beitrag zur Kenntniss der secundaren Horbahnen der Katze. Folia neurobiol., 5: 869-889. Winkler, C. (1921). Anatomie du Systeme Nerveux. Vol. 2. Haarlem: de Erven F. Bohn. Woollard, H.H., & J.A. Harpman (1940). The connexions of the inferior colliculus and of the dorsal nucleus of the lateral lemniscus. J. Anat., 74:441-458. Wortis, S.B., & A.Z. Pfeffer (1948). Unilateral auditory-spatial agnosia. J. Nerv. Ment. Dis., 108:181 - 186. Wu, S.H., & J.B. Kelly (1992). Synaptic pharmacology of the superior olivary complex studied in mouse brain slice. J. Neurosci., 12: 30843097. Wu, S.H., & J.B. Kelly (1994). Physiological evidence for ipsilateral inhibition in the lateral superior olive: synaptic responses in mouse brain slice. Hearing Res., 73: 57-64. Yates, G.K., B.M. Johnstone, R.B. Patuzzi, & D. Robertson (1992). Mechanical preprocessing in the mammalian cochlea. Trends Neurosci., 15: 57-61. Yin, T.C.T., & J.C.K. Chan (1990). Interaural time sensitivity in medial superior olive of cat. J. Neurophysiol., 64: 465-488. Yoshida, I. (1925). Uber die funktionelle Bedeutung der oberen Olive nebst ihren Faserbahnen. Folia anat. Japon., 3:111-136. Zook, J.M., & J.H. Casseday (1979). Connections of the nuclei of the lateral lemniscus in the mustache bat, Pteronotus parnellii parnellii. Neurosci. Abst., 5: 34.
466
Hutson
Zook, J.M., & J.H. Casseday (1982a). Cytoarchitecture of auditory system in lower brainstem of the Mustache bat, Pteronotus parnellii. J. Comp. Neuro l., 207: 1-13. Zook, J.M., & J.H. Casseday (1982b). Origin of ascending projections to inferior colliculus in the mustache bat, Pteronotus parnellii. J. Comp. Neurol., 207: 14-28. Zook, J.M., & J.H. Casseday (1985). Projections from the cochlear nuclei in the mustache bat, Pteronotus parnellii. J. Comp. Neurol., 237: 307-324.
00
SECTION VI:
TACTUAL PROCESSING
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
469
Chapter 13
Role of Sensory and Post-Sensory Factors on Hemispheric Asymmetries in Tactual Perception Joi~l Fagot 1, Agn~s Lacreuse 1,2 & Jacques Vauclair I 1 Center for Research in Cognitive Neuroscience Marseille, France 2 Department of Psychology, University of Georgia, Athens, USA It is well known that lateralization characterizes a broad range of activities in the motor (e.g., handedness), visual and auditory domains. It is much less acknowledged, however, that lateralization is also a property of tactual perception. The main goal of this chapter is thus to present and discuss this issue in detail. As several authors have questioned the robustness of tactual asymmetries in humans (e.g., Duda & O'Adams, 1987; Summers & Lederman, 1990), the main objective of this chapter is two-fold. First, we present a large but selective overview of the main findings in favor (and against) lateralization in tactual processing. Second, this chapter attempts to shed light on the respective role of sensory and post-sensory factors on the expression of tactual cerebral lateralization.
470
Fagot, Lacreuse, and Vauclair
It can be reminded that the sense of touch involves very different types of information (e.g., regarding pressure, temperature, shape) and uses both the cutaneous and the kinesthetic receptors (Gibson, 1962; Loomis & Lederman, 1986). Cutaneous perception refers to superficial deformations of the skin, while kinesthesis refers to information coming from muscles, tendons and joints. Under natural conditions, these senses usually work together and we then talk about "haptic" or "tactilokinesthetic" perception (Loomis & Lederman, 1986). Given the synthetic nature of tactual perception, this chapter will present evidence for lateralization in a broad range of tactual activities involving either the cutaneous or kinesthetic senses only, or both. After a short description of the anatomy of somato-sensory pathways, we propose to present evidence of lateralization for tasks varying in complexity. Thus, in a second part of this chapter, we will look for functional asymmetries in elementary tactile discriminations, for instance in pressure discrimination tasks. Then, we will review laterality findings in tasks having a strong cognitive load, such as the discrimination of orientations, tactual maze learning, or the discrimination of meaningful and meaningless forms. This latter section on form discrimination will emphasize our own researches in this area and their significance for understanding some of the inconsistencies of the literature. Note that the published material on tactual lateralization mainly concerns fight-handed normal subjects. It is thus primarily these data that will be reported here, although studies on patients, children and even nonhuman primates will occasionally be cited. Anatomical bases of tactual perception Two ascending systems carry somesthetic information to the cortex: the lemniscal system and the anterolateral system. The anterolateral system, which conveys information about pain, temperature and crudetouch is out of the scope of this paper, and will thus not be presented here (see Kandel & Schwartz, 1981 for a description of the anterolateral system). The lemniscal system, which is of interest here, carries discriminative touch sensation, vibration sense, and information about joint and limb position. Of the most importance for laterality studies, the lemniscal system has the peculiarity of conveying sensory input from one body side to the opposite hemisphere (Cholewiak & Collins, 1991). This system
Tactual Perception
471
originates in the large diameter axons from dorsal root ganglion cells which receive their input from the periphery. These fibers ascend ipsilaterally to the medulla where they cross the midline and then ascend through the contralateral lower brain stem and terminate in the ventral posterior lateral and the ventral posterior medial nuclei of the thalamus. From there, fibers synapse and project to the cortical areas by thalomocortical pathways. Several cytoarchitectonic regions of the parietal cortex have been identified as receiving somatic inputs: they are primary somatic sensory cortex (SI: Brodmann's areas l, 2 and 3), secondary somatic sensory cortex (SII), corresponding to the preinsular portion of area 2, and the somatic sensory association areas (areas 5 and 7). In practice, laterality studies take advantage of the natural decussation of the lemniscal system by applying tactile stimulations on one hand in order to stimulate the cerebral hemisphere on the opposite side. Even with normal subjects for whom information may transfer from one hemisphere to the other via the interhemispheric commissures (e.g., Iwamura, Iriki & Tanaka, 1994), this technique proved to be an efficient method to reveal laterality effects in tactual processing (e.g., Fagot, Lacreuse & Vauclair, 1993). Moreover, distal hand movements are under the control of the hemisphere on the opposite side (e.g., Brinkman & Kuypers, 1973), thus limiting the implication of the ipsilateral hemisphere. The reader is referred to Corkin (1978), Kandel and Schwartz (1981), and Cholewiak and Collins (1991) for further information about the neural bases of tactual perception.
Functional asymmetries for elementary tactile discriminations Under the heading of "elementary tactile discriminations", we have grouped the experimental paradigms requiring pressure and vibration sensitivity, the localization of cutaneous stimuli, the ability to perceive two points as distant, roughness and weight discrimination, as well as the discrimination of kinesthetic positions. Studies on pressure and vibration sensitivity in normal subjects provided some evidence for a greater sensitivity of the left hand compared to the fight for both static (Ghent, 1961; Semmes, Weinstein, Ghent & Teuber 1960; Weinstein, 1962, 1968, 1978; Weinstein & Sersen, 196 I) and vibrotactile stimulation (Rhodes & Schwartz, 1981, Wiles, Pearce, Rice & Mitchell, 1990). Although emerging from several studies, the left hand bias for pressure and vibration sensitivity is not a
472
Fagot, Lacreuse, and Vauclair
robust phenomenon. First, when it appears this effect is usually small in amplitude (e.g., Wiles et al., 1990). Second, several authors failed to replicate it (e.g., Carmon, Bilstrom & Benton, 1969, Fennell, Satz & Wise, 1967; Greenspan & McGillis, 1994; Thibault, Forget & Lambert, 1994). Bradshaw and collaborators investigated vibrotactile reaction times by the left and fight hand in right-handed subjects (e.g., Bradshaw, Nathan, Nettleton, Pierson & Wilson, 1983; Bradshaw, Bradshaw, Pierson-Savage & Nettleton, 1988; Pierson-Savage & Bradshaw, 1987). These authors systematically reported equal response latencies to detect vibrations by the left and right hand. Moreover, regardless of the stimulated hand, response speed interacted in a complex way with the hemispace in which stimulation was received, the orientation of the head, and trials order, suggesting that lateralization in vibrotactile studies might be attentional in nature (e.g., Bradshaw et al., 1988). Different approaches of lateralization in cutaneous perception involved (1)point localization, in which the subject has to report the spatial location of the tactile stimulus, (2) the two-point discrimination technique, in which the experimenter measures the minimal separation required to perceive the two points as being discriminable, and (3) the discrimination of roughness. The point localization task provided no evidence for a differential ability of the left and fight hemisphere to localize tactual stimuli (Semmes et al., 1960; Weinstein, 1968), although a bilateral sensory loss in this task was more often related to left than to right hemisphere lesions by wound penetration (Semmes et al., 1960). Works with commissurotomized patients mainly showed that intermanual transfer of tactual localization tasks was impossible after complete bisection of callosal commissures, but they did not reveal any significant difference between the two hands (e.g., Gazzaniga, Bogen & Sperry, 1963). The two-point discrimination task showed again no advantage for either the left or right hand and thus for the left or right cerebral hemisphere (Charron, Collin & Braun, 1996; Schiavetto, Lepore & Lassonde, 1993) or showed conflicting results (Weinstein, 1968), suggesting that the spatial resolution of the skin is identical for the two hands. The same pattern of results emerged from roughness discrimination studies (Heller, 1982; Lederman, Jones & Segalowitz, 1984), with the exception of Ardila, Uribe and Angel (1987) who reported lower differential thresholds for the nonpreferred hand in both right- and lefthanded subjects.
Tactual Perception
473
We stated in introduction that tactual perception encompasses the cutaneous sensibility and kinesthesis. Roy and McKenzie (1978) investigated kinesthetic lateralization in a thumb positioning task. They found that the left hand made more sensitive kinesthetic discriminations than the fight. This effect was later confirmed by Nishizawa and Saslow (1987) and Nishizawa (1991; but see Colley, 1984). Interestingly, the left side advantage for sense position was observed regardless of hand preference (Riolo-Quinn, 1991), and for movements of the feet (Carnah a n & Elliot, 1987) or of the whole arm (Kurian, Sharma & Santhakumari, 1989), even though such movements are not under the unique control of the contralateral hemisphere (e.g., Brinkman & Kuypers, 1973). For ankle, shoulder and hip movements, however, Thibault et al. (1994) found no significant difference between the left and fight sides. The overall picture of the literature thus speaks in favor of a fight hemisphere advantage for sense position in a broad range of movements. Indeed, Leonard and Milner (1991) tested brain damaged subjects in a kinesthetic task (i.e., duplicate the movement of a lever) implying the mobility of the left or fight arms. They reported that patients with left frontal-lobe or small right-lobe excisions performed as well as control subjects in this task, but those patients with large frontal-lobe removals were impaired compared to controls, whatever the arm to be used. As sense position appears to be lateralized, it might be expected that weight discrimination tasks would give rise to a left hand advantage, because this task involves both cutaneous and kinesthetic information. In accordance with this hypothesis, Brodie (1985) found a slight lefthand advantage for right-handed male subjects, but unfortunately this author did not carry out statistical analyses on this effect. Asymmetries for weight discrimination were replicated neither by Streitfeld (1985), Nishizawa (1991) nor Brodie (1988). Moreover, differences between left- and right-handed subjects are inconsistent, since Brodie (1988) found better performance of left- over right-handed subjects, while the reverse trend was observed by Ardila et al. (1987). In summary, the tactual studies on elementary discrimination gave mixed indications for lateralization. The pressure and vibration sensitivity tasks indicated some tendency for a left hand advantage, but this effect was not systematically found (Greenspan & McGillis, 1994), or may be attributed to attentional instead or sensory factors (see Bradshaw and collaborators' works). For the other simple discrimination tasks involving cutaneous perception, such as the point localization task
474
Fagot, Lacreuse, and Vauclair
or the two-point discrimination task, the data-base suggests equal perceptual abilities between the left and right hands. In fact, the most reliable form of lateralization concerns sense position, for which there is good evidence for a fight hemisphere advantage. Tactual discrimination of orientations
The task of orientation discrimination is one step further in difficulty than the previous tasks, mostly because it requires an oriented frame of reference. It is thus interesting to verify if studies on orientation discrimination showed laterality effects, and if these effects were (or not) consistent. Benton, Levin and Varney (1973) used a passive mode of stimulus presentation to investigate this issue in normal subjects. The stimuli comprised four orientations (horizontal, vertical, 45 ~ left or fight) which were applied during one second on the palm of each hand: subjects showed a left hand advantage. This effect was later reproduced by Varney and Benton (1975) with normal right-handed subjects, but not with left-handed subjects for whom scores were better for the fight hand. In a different study (Benton, Varney & Hamsher, 1978), subjects had to explore one oriented rod with one hand before recognizing it among 11 possible orientations. Again, a left hand advantage, however restricted to male subjects, was observed. Zoccolotti, Passafiume and Pizzamiglio (1979) confirmed this left hand advantage, but they did not find sex differences. To our knowledge, one study only examined lateralization of tactual recognition of orientations in children (Brizzolara, De Nobili & Ferretti, 1982), revealing a left hand advantage in 6-7-year olds. Interestingly, not only normal subjects, but also patients showed a left hand (right hemisphere) advantage for line orientation discrimination. For instance, Carmon and Benton (1969), and Fontenot and Benton (1971) applied linear stimulations of different directions on the palm of patients with unilateral brain lesions. Whatever the hand, fighthemispheric damaged patients showed a defect in scores for line orientation recognition. Left hemispheric damaged patients, by contrast, performed badly only with their right hand, suggesting a right hemisphere advantage to process orientations. In a complementary study, DeRenzi, Faglioni and Scotti (1971) presented two pairs of rods to brain damaged patients. The task was to set one pair in the same position as the standard pair, using the hand ipsilateral to the side of the
Tactual Perception
475
lesion. Results indicated that fight hemisphere damaged patients performed poorly compared to other subjects, including the control and the left damaged groups. From this overview of the literature, it is striking that, at least in righthanders, line orientation tasks consistently demonstrate a left hand (right hemisphere) advantage in normal subjects, in children and patients, whatever the mode of stimulus presentation (i.e., passive or active). Retention of sequences of touches
When humans have to remember series of touches, lateralization depends on the processing mode. Thus, Nachson and Carmon (1975) bimanually applied non sense tactile stimuli on the left and right fingertips of right-handed subjects. The task was to identify the sequence of dot stimulations, or their spatial distribution on the stimulated fingers. A fight hand advantage occurred for the sequential task, but the spatial task gave rise to a left hand advantage, which demonstrates that the spatial component of the task is critical for the emergence of the left hand advantage. In addition to the sequential/ spatial factor, the length of the to be remembered sequence of touches may also affect lateralization. This effect emerged in a study (Lechelt, 1980) investigating how accurately right- and left-handers reported the number of tactile pulses in a temporal sequence. Whatever the hand preference group, the preferred hand was better than the other one when the trains comprised only a small number of pulses (<8). There was, by contrast, an advantage for the non-preferred hand, and thus of the left hand in right-handers, for longer series of pulses (i.e., >7). Tactual discrimination of dot patterns
Dot patterns can be randomly presented in a display, or can be arranged to provide significant information (e.g., to form numbers or Braille characters). Several studies investigated lateralization for counting the number of dots in a pattern. Young and Ellis (1979) found that the left middle finger of right-handed subjects was better for counting than the right middle finger. This effect occurred when the dots were randomly located on the array, but disappeared when aligned, thus showing the critical role of the spatial component of the task for lateralization. Tactile dots patterns arranged to form digits were also
476
Fagot, Lacreuse, and Vauclair
applied with vibrotactile systems (i.e., Optacon: Heller, Rogers & Perry, 1990); recognition was greater with the left than with the right fingertips. Similar results had previously been found with stationary stimuli forming numbers (Harley & Grafman, 1983). In some studies, dot patterns were arranged to form Braille characters. Again, when asymmetries emerge in normal blindfolded subjects, they are in favor of the left hand (Smith, Chu & Edmonston, 1977), but this effect is not robust, as some studies report no hand differences (Myers, 1976). Moreover, Rudel, Denckla and Spalten (1974) asked 7-14- year old children to learn Braille characters. In 1314-year olds, recognition performance was better if the stimuli were learnt with the left hand, but this effect did not occur before age 11. The same rather inconsistent pattern emerged for accustomed Braille readers. As Braille characters convey linguistic information, a right hand (left hemisphere) advantage would be expected for reading among Braille experts. In spite of the linguistic character of the Braille material, it is striking that studies often showed a left hand instead of a right hand advantage (e.g., Mommers, 1980; Hermelin & O'Connor, 1971). It is possible that the Braille characters are more spatially complex than ordinary letters. This hypothesis would explain the right hemisphere superiority (Iaccino, 1993) in Braille reading. It should be noted, however, that the left advantage was not systematically found (Foulke, 1964; Millar, 1984; Mousty & Bertelson, 1985), which might be explained by the linguistic nature of Braille characters, or by the fact that Braille readers alternatively use their left and right hand to focus on the verbal and spatial aspects of the task (e.g., locate the beginning of the line; Millar, 1987). Wilkinson and Carr (1987) observed in congenitally blind people that the preferred hand (left or right) was the most accurate to identify Braille characters, but noted that the advantage of the preferred hand was greater for left- than for right-handers. More recently, Semenza, Zopello, Giduili and Borgo (1996) used a dichhaptic procedure with skilled blind Braille readers. They found a right hand advantage when the letters could be matched on the basis of their names, but no hand difference when the matching could be performed on the basis of perceptual identity. In short, it appears that several factors intervene in lateralization for Braille reading. One of them is the tactual component of the task which seems to favor the left hand. Another one is the linguistic signification of Braille characters which may advantage the left hemisphere, and thus the right hand.
Tactual Perception
477
Tactual Maze Learning If the activity of the hand reflects the processes at work in the opposite hemisphere, then the tactual maze learning task should reveal a left hand advantage, given the well known right hemisphere advantage for spatial processing (Bradshaw & Rogers, 1992; Hellige, 1993). This hypothesis is indeed supported by most of the literature reported below. An early version of this task required subjects to learn a tactual maze and then reproduce it in a locomotor context (Semmes, Weinstein, Ghent & Teuber, 1955). There was no left/right difference in this task, but the results are difficult to interpret, as solving the loco-motor maze involved different processes than the tactual mazes. Corkin (1965) devised a tactual maze task for patients with various cerebral damages and recorded the number of trials and errors required to learn the correct sequence of turns in the maze. Specific types of brain damaged patients were impaired in this task, in particular patients with right frontal lobe excisions, and those patients with large right posterior excisions. Left brain damaged patients as well as patients with small right parietal lobe excisions were not impaired. These results showed that areas of the fight hemisphere were mainly involved in the learning of tactual mazes. The implication of the right hemisphere for tactual maze learning is additionally supported by a study (DeRenzi, Faglioni & Scotti, 1970) in which brain damaged patients had to find a marble in a maze with the forefinger ipsilateral to the side of the lesion. All patients (left and fight brain damaged) were impaired, but right hemisphere damaged patients had more difficulties to find the marble than left hemisphere damaged patients. Sullivan (1989) also tested the ability of normal subjects to recall a tactual task (i.e., reproduce a pattern of touches) after being given an irrelevant tactual maze activity. Performance was better if the left hand had to trace the maze, whatever the hand which was stimulated in the first task. Moreover, a recent study by Ward, Alvis, Sanford, Dodson and Pusakulich (1989) showed, for both left- and right-handers, faster maze learning with the left than with the fight hand, though each group solved the task more rapidly with their dominant hand.
Haptic discrimination of spatial forms The literature on form discrimination is much larger than the previous one, and only a subset of the relevant papers will be presented
478
Fagot, Lacreuse, and Vauclair
here (for a review, see Summers & Lederman, 1990; Verjat, 1989). The majority of the studies on haptic lateralization focused on the discrimination of complex non-sense shapes, but some researchers used linguistic stimuli (e.g., Witelson, 1974), or other meaningful stimuli such as animal shaped forms (e.g., Minagawa & Kashu, 1989), or familiar objects (Milner & Taylor, 1972). As a general trend, this literature indicates that the stimulus type is a very important factor affecting hand biases. Importance of this factor might be illustrated by the work of Franco and Sperry (1977). These authors tested the ability of commissurotomized patients to recognize four different types of geometrical shapes (i.e., topological, projective, affine and Euclidian shapes). On each trial, the patient was presented with five visual geometric shapes. He had then to choose the tactual shape sharing some properties with the standard set of objects. The left hand outperformed the right hand in this task. However, the type of objects had a significant influence, as right hand scores progressively declined from affine to projective geometry and were at chance level for the topological shapes. Several other studies revealed a fight hand advantage for meaningful stimuli, while meaningless stimuli gave rise to a left hand advantage. For example, Cioffi and Kandel ( 1 9 7 9 ) o b s e r v e d a right hand (left hemisphere) bias in children for words identification, and an opposite left hand bias for recognizing nonsense shapes. The same shift in laterality was observed with letters and shapes in 8-14 years old children (Gibson & Bryden, 1983), and Oscar-Berman, Rehbein, Porfert and Goodglass (1978) found no lateral effect to recognize digits, but a right hand advantage to recognize letters. In addition, lateralization with meaningless shapes may also depend on the stimulus mode of processing, as the degree of left hand advantage is influenced by the instruction to approach the stimuli in a global or in a sequential manner (Webster & Thurber, 1978).
Exploratory strategies for nonsense shape discrimination The experiments conducted in our research group revealed the implication of several other factors in the determination of hand tactual asymmetries for shape discrimination. The impetus for these experiments was the observation that the literature on nonsense form recognition usually used the accuracy score as the main (if not the sole)
Tactual Perception
479
dependent variable (see for a review, Summers & Lederman, 1990). We have argued that this type of approach is of limited heuristic power, because accuracy scores provide no insight into the hemispheric processing mode per se, and the hemispheres can solve the task by different ways but still reach identical performance (Fagot, Lacreuse & Vauclair, 1993). For these two reasons, we used in our laboratory a novel apparatus, which allowed to record the hand exploratory strategies, in complement to the classical measure of hand recognition scores. Figure 1 illustrates this apparatus. It consists of an opaque box equipped with two side-by-side sliding doors, each operated independently by a motor. Raising the left or fight door provides access to the stimuli which are concealed from view inside the box. Stimuli employed with this system are three-dimensional nonsense shapes made of metallic cubes (1 x 1 x 1 cm) fixed side-by-side on a lexan baseboard. Cubes are precisely adjusted so that their junctions are not haptically discernible. Stimuli are mounted on vertical panels inside the box. During the experiment, each cube forming the stimuli was positively polarized and electrically insulated from the other cubes. Thus, providing that the subject was grounded, hand contact with a cube shifted its voltage, which was recorded on-line by computer, and later used to infer the location and duration of hand contacts on the stimulus. One important feature of this apparatus is that, given the small size of the opening, stimulus exploration had to be performed by distal movements rather than by proximal elbow or shoulder movements. This feature is critical, because only distal movements are under the exclusive control of the opposite cerebral hemisphere (Brinkman & Kuypers, 1973). A more detailed technical description of the apparatus and its components can be found in Fagot, Arnaud, Chiambretto and Fayolle (1992). A common experimental design was adopted in all our experiments. During a trial, the subject was seated at a table, facing the apparatus into which either single (monohaptic testing) or two (dichhaptic testing) sideby-side stimuli were placed (see Fig. 1). When the test involved monohaptic exploration, the stimulus was explored inside the box by the hand ipsilateral to the opened door. By contrast, when dichhaptic exploration was required, the two hands were introduced inside the box in order to explore the two stimuli simultaneously, one stimulus by each hand. Altogether, we have conducted four different experiments with this system. In some experiments, a tactual-visual procedure was adopted, that is, subjects had to initially touch a stimulus before recognizing
480
Fagot, Lacreuse, and Vauclair
III
!
Stimulus i
Stimulus
B
I
Door
J
Figure 1: Shape discrimination apparatus. it on a visual array. In other experiments, the two stimuli were haptically inspected, and then, the subject had to compare these two forms in order to make a same-different judgment or to recognize one of them in the visual modality. With the aim to infer the mode of stimulus processing in a direct way, exploratory strategies were analyzed from the mean number of cubes that were simultaneously touched during the exploration. The main results concerning this variable are provided in Table 1. The first important result is that asymmetries may exist in terms of hand exploratory strategy, even though there is no hand difference for scores (e.g., Fagot, Lacreuse & Vauclair, 1993, see also Table 1). This finding is important if we consider that the previous conclusions on the absence of lateralization (e.g., Duda & O'Adams, 1987) derived from studies in which the score was the sole dependent variable. Another conclusion from Table 1 is that, when the lateralization appeared in scores, it always favored the left hand. As our stimuli were meaningless and enhanced a spatial coding, this result might be accounted for by a superiority of the fight hemisphere to process spatial information (Lacreuse, Fagot & Vauclair, 1996). This left hand bias is moreover consistent with the main core of the literature concerning nonsense form discrimination (see for reviews: Summers & Lederman,
Tactual Perception
481
T a b l e 1: Summary of main findings from a series of experiments conducted in our research group. For each task, this table indicates the exploratory mode (monoor dichhaptic), the phase during which exploratory strategies were investigated (learning and recognition phase), and the main results in terms of scores and number of cubes touched on average during stimulus palpation. Number in parentheses refer to the actual reference of these studies: 1= Fagot, Lacreuse & Vauclair, 1993; 2=Fagot, Hopkins & Vauclair, 1993; 3=Lacreuse, Fagot & Vauclair, 19%; 4=Fagot, Lacreuse & Vauclair, 1994; LH=left hand; RH=right hand, ns=non significant effect.
Exploratory mode Monohaptic T-V Dichhaptic T-V Dichhaptic T-T Monohaptic T-T
(1) (2) (3) (4)
Phase
Accuracy
Number of cubes
I_x.aming Learning Recognition Learning Recognition
ns. LH>RH LH>RH ns. ns.
LH>RH LH>RH ns. LH>RH ns.
1990; Verjat, 1988) on both normal subjects (Hatta, 1978; Cohen & Levy, 1986, 1988; Dodds, 1978; Flanery & Bailing, 1979; Riege, Metter & Williams, 1980; Streitfeld, 1985), and brain damaged patients (e.g., Milner & Taylor, 1972; Nebes, 1971). It should be noted, however, that this effect has not always been obtained (e. g., Y amamoto & Hatta, 1980; Webster & Thurber, 1978), while a very limited number of studies (e.g., Cranney & Ashton, 1982; Yandell & Elias, 1983) showed a rightinstead of left-hand superiority for meaningless shape recognition. Our studies also revealed that lateralization may emerge during the initial learning phase, but may disappear during the recognition phase. For instance, whatever the experiment, the left hand was found to touch a greater surface of the stimulus than the right, but this effect was present during the learning phase only (see Table 1). Because the same stimulus set was used for learning and recognition, this phase difference cannot be accounted for by physical characteristics of the input. It could rather be attributed to cognitive factors, such as the knowledge of the shape which provided a frame in the search for form features. In a different perspective, some authors have suggested that monohaptic situations are inadequate to reveal manual perceptive asymmetries
482
Fagot, Lacreuse, and Vauclair
and proposed, instead, to use dichhaptic tasks (e.g., Witelson, 1974). Results reported in Table 1 are in agreement with this position, because score asymmetries were observed in dichhaptic testing conditions only. The question remains, however, to understand why the dichhaptic task is more appropriate for unveiling asymmetries. In an attempt to address this question, Witelson (1974) argued that the dichhaptic task produces a competition in the neural system for the required cognitive processing in the two hemispheres. According to Witelson (1974), this competition gives a leading role to the specialized hemisphere which will thus control haptic information processing. We conducted an experiment which addressed this hypothesis in an indirect way. This study (Fagot, Hopkins & Vauclair, 1993) used the apparatus depicted in Figure 1, and subjects were requested to touch two objects simultaneously, and then to recognize one of them on a visual array. In the analysis, we wanted to verify if the two hands would work in synchrony during the dichhaptic exploration. For that purpose, we distinguished two types of exploration strategies. The first one involved a simultaneous displacement of both hands on the shape. The second one involved contacts of both hands with the shape, but one hand at least was not moving for a minimum of 500 ms. The percentage of time spent with both hands active is represented in Figure 2. On average for the group, 20 percent of the exploration time only was devoted to a simultaneous exploration of the two objects. During the other 80 percent, one hand only was active at a time, and the other one remained in a fixed position, with an intermittent alternation in the function of each hand. A similar division of labor between the two hands was observed in blind people reading Braille (Millar, 1987). If the dichhaptic procedure involved so few simultaneous explorations, why did it reveal asymmetries that the monohaptic procedure was unable to elicit? Rather than referring to a process of inter-hemispheric competition (i.e., Witelson, 1974), we propose that dichhaptic exploration introduces more cognitive constraints than the monohaptic exploration, for instance in terms of attention sharing and memory load, which is favorable for lateralization effects to emerge. It is also likely that the instruction to simultaneously explore the two objects made the linguistic encoding difficult, thus enhancing the spatial component of the task. In brief, it occurs that lateralization exists for tactual form discrimination, but that the left hand bias emerging from somato-spatial
Tactual Perception
483
100
80
60
40
20
1
2
3
4
5
6
7
8
9
1 0 11
12
13
14
Figure 2: Percentage of exploration time spent using bimanual versus unimanual strategy (solid: bimanual search; stippled: unimanual search).
tasks may be masked, or may even be reversed, under the influence of several possible factors. Among these factors are the complexity of the form (e.g., Franco & Sperry, 1977), the subjects' previous knowledge or expectancies (Fagot et al., 1994), the instruction to process haptic information in a given way (Webster & Thurber, 1978) and the need to process several forms in parallel (see Table 1). Of particular importance, the reference to verbal codes for form processing may cancel asymmetries in favor of the left hand, and may even induce a bias in the opposite direction (e.g., Cioffi & Kandel, 1979). In this context, it is relevant to verify if the closest relatives of humans (monkeys and apes) share the same lateralization patterns as humans, as nonhuman primates have no human-like language p e r s e (Vauclair, 1996).
Haptic Perception in nonhuman primates Very few studies investigated lateralization in somatosensory discrimination tasks with nonhuman primates, and most of them focused
484
Fagot, Lacreuse, and Vauclair
on the existence of a preference for either hand rather than on asymmetries in performance. Consequently, this approach limits the possibility to strictly compare human and nonhuman primates, because the available human literature bears on performance measures. Nevertheless, experimental evidence suggest that nonhuman primates exhibit a left hand preference for haptic tasks, even though the neural mechanisms underlying this asymmetry remain unknown (Garcha, Ettlinger, MacCabe, 1982). As a first support for this hypothesis, a series of experiments showed in macaques a trend (though not significant) for preferring the left-hand in haptic form discrimination (Ettlinger, 1961; Ettlinger & Moffet, 1964; Milner, 1969; Brown & Ettlinger, 1983). In these studies, however, the sample size was small and left/right differences were not significant. In a larger group, Hoerster and Ettlinger (1985) showed that 77 rhesus who predominantly used their left hand learned haptic discrimination faster than did 78 monkeys who preferred their right hand. Other evidence for group asymmetry in haptic tasks derives from Fagot, Dr6a and Wallen (1991) whose findings on rhesus macaques were later replicated in cebus monkeys by Parr, Hopkins & de Waal (in press) and Lacreuse & Fragaszy (1996). In this study, rhesus monkeys had to climb a wire netting and to maintain a vertical three point posture while they introduced one hand in an opaque box in order to discriminate peanuts mixed with sand and stones of different sizes. There was for the group (n=29) a significant left hand preference for haptic search. A similar left bias occurred for haptic reach in a sitting position, showing that the posture was not critical for the emergence of the left hand preference. When a visual version of this task was proposed, the left hand bias again emerged, but the frequency of left hand usage was significantly lower than in the tactile version (Fagot et al., 1991). To our knowledge, one study only (Lacreuse & Fragaszy, in press) addressed the relation between performance and preference for haptic task in monkeys. This study showed a left hand preference for obtaining out of view food items fixed on the side surface of three dimensional objects, but detailed analyses of hand movements during haptic exploration failed to reveal hand asymmetries in search strategies. In short, findings resulting from nonhuman primate studies are consistent with the human picture, in that asymmetries for haptic perception are generally in favor of the left hand. The problem, however, is that this left bias in monkeys was inferred from measures
Tactual Perception
485
different from those employed in human studies. The existence of a left hand bias in monkeys suggests that the previously reported human left hand right-hemisphere advantage has a long phylogenetic history, and certainly predates cerebral lateralization for linguistic processing. General discussion
The general picture that emerges from this overview of the literature is presented in a condensed way in Table 2. This table reports, for each task, the presence or absence of hand biases as inferred from the literature and provides additional general comments on these effects. Elementary discrimination tasks are presented in the upper part of Table 2 and complex discrimination tasks are listed in the lower part. When the effects are very consistent in the literature, Table 2 refers to a "hand bias". The wording "hand tendency" is used in Table 2 when effects were repeatedly reported, but several studies failed to replicate them. A quick look at Table 2 indicates a general tendency for a left hand (right hemisphere) advantage for somato-sensory discriminations. With the possible exception of the pressure and vibration sensitivity tasks, it must be remarked that tasks used in the assessment of haptic lateralization have in general a high spatial load. Hence, it is not unreasonable to propose that the left-hand biases reported in Table 2 reflect a fight hemisphere advantage for spatial processing. As the fight hemisphere advantage was also observed for visuo-spatial tasks (e.g., Hellige, 1993), this advantage might not be tied to the tactual modality, but might reflect a supramodal ability to process spatial information. The tasks listed in Table 2 have sensory as well as post-sensory components, since subjects had to feel the stimuli and to report what they felt. Also, given the varying complexity of the tactile input, these tasks differ in terms of cognitive load. Thus, it is likely that comparing the findings on the basis of the demands of the task will shed light on the respective role of sensory and post-sensory factors on brain lateralization. Regarding the possible role of early factors, Table 2 suggests that a distinction must be made between kinesthetic and cutaneous inputs. For kinesthesis, asymmetries emerged and were consistently in favor of the left side. They were described for thumb, arm and feet movements, but it must be acknowledged that the literature in this domain remains limited, which prevents any definitive conclusion but rather calls for an in depth investigation of this possibility. For
486
Fagot, Lacreuse, and Vauclair
T a b l e 2: Summary of tactual studies on hand/hemisphere specialization.
TASK
EFFECT
COMMENTS
Pressure/vibration sensitivity
left tendency
possible attentional effects
Point localization
no asymmetry
too few studies
Two-point discrimination
no asymmetry
too few studies
Roughness discrimination
no asymmetry
too few studies
Weight discrimination
no clear trend
too few studies
Sense position
left bias
reported for thumb, arm, and feet
Orientation discrimination
left hand bias
found with different populations
Retention of sequences
left/fight hand advantage
depends on processing mode
left tendency
too few studies/depends on spatial distribution
left tendency
interference between spatial and linguistic codings apparently robust
Dot pattems -numbers -Braille Tactual maze
left hand bias
Nonsense form discrimination
left tendency
depends on presentation mode (dichhaptic or not) depends on encoding mode (linguistic or not)
Haptic search in monkeys
left hand preference
too few studies on hand performance
cutaneous inputs, only the pressure/vibration sensitivity tasks provided some arguments for lateralization. Lateralization was in this case in favor of the left hand, but the obtained effects, possibly affected by attentional factors, were not systematic, although no studies reported a right hand advantage. Other tasks involving the simple cutaneous stimuli showed either no clear trend for a lateral bias (e.g., roughness discrimination) or no asymmetry at all (e.g., two-point discrimination). The necessary
Tactual Perception
487
conclusion is thus that there is no definitive evidence for a sensory lateralization related to the cutaneous sense p e r se. A t best, the literature indicates some tendencies for a left hand advantage (i.e., for pressure/ vibration sensibility tasks), but the left hand advantage hypothesis awaits further investigations to be validated. Turning now to the discrimination of more complex tactual stimuli, the lower part of Table 2 strongly suggests that tactual lateralization exists, and is most of the time in favor of the left side. Note, however, that laterality biases depend on a series of subject (e.g., handedness) and stimulus related factors (e.g., verbal codability and presentation mode of the stimulus). It happens that the most convincing left hand advantage concerns line orientation discrimination, as this effect was obtained in normal subjects, patients and children. At least three aspects of the literature suggests that the left hand advantage for line orientation was determined by post-sensory factors. Firstly, Honda (1977) demonstrated that the left hand advantage occurred even though the task required the discrimination of imaginary lines, because only the two extreme points of the line, instead of the whole line, were physically applied on the skin. Secondly, right hemispheric advantages for line orientation were repeatedly found in visual studies (e.g., Atkinson & Egeth, 1973; Benton, Hannay & Varney, 1975), suggesting that this asymmetry is not linked to a specific sensorial modality. Thirdly, Oscar-Berman et al. (1978) showed a left hand advantage only when the left hand responded after the right, suggesting that tactual laterality effects for line orientation may emerge after some delay in short-term memory. The largest literature on haptic lateralization concerns nonsense form discrimination. Again, when an asymmetry was found in this domain, it was mostly in favor of the left hand, but the left hand bias was not always robust (see Table 2). However, it is noticeable that this literature used the score and occasionally response times as the unique dependent variables. As the score variable appears to be less sensitive to laterality effects than hand exploratory strategies (e.g., Fagot, Lacreuse & Vauclair, 1993), it is suggested that the inconsistencies in hand biases mainly rest on a restricted selection of the measures, and it is thus presumed that a clear-cut picture would emerge if strategies are more systematically considered. As already stated above, the analysis of hand exploratory strategies has the additional advantage to provide information on the processing mode adopted by each hemisphere (e.g., global vs analytic treatments).
488
Fagot, Lacreuse, and Vauclair
In 1990, Summers and Lederman concluded from their review of the literature that "somatosensory perceptual asymmetries are not robust, although hand superiorities are in the predicted direction when they do occur" (page 221). Our own review of the literature is in agreement with this position, but expands it by allowing conclusions regarding the respective role of early and late factors. In this perspective, if we can suspect that early somato-sensory factors affect lateralization, with the possible exception of kinesthetic tasks, this involvement has not yet been firmly established. Haptic lateralization, however, appears to be affected in a major way by a series of cognitive factors, and in particular, by the spatial and verbal dimension of the task. In this respect, the nonhuman primate model, for which there are already evidence of left hand haptic lateralization, might be informative and is worth developing. References
Ardila, A., Uribe, B.E., & Angel, M.E. (1987). Handedness and psychophysics: Weight and roughness. International Journal of Neurosciences, 36, 17-21. Atkinson, J., & Egeth, H. (1973). Right hemisphere superiority in visual orientation matching. Canadian Journal of Psychology, 27, 152-158. Benton, A.L., Hannay, H.J., & Varney, N.R. (1975). Visual perception of line orientation in patients with unilateral brain disease. Neurology, 25, 907-910. Benton, A.L., Levin, S., & Varney, N. R. (1973). Tactile perception of direction in normal subjects. Neurology, 23, 1248-1250. Benton, A.L., Varney, N.R., & Hamsher, K. de S. (1978). Lateral differences in tactile directional perception. Neuropsychologia, 16, 109-114. Bradshaw, J.L., & Rogers, L.J. (1992). The Evolution of Lateral Asymmetries, Language, Tool Use, and Intellect. San Diego" Academic Press. Bradshaw, J.L., Bradshaw, J.A., Pierson-Savage, J.M., & Nettleton, N.C. (1988). Overt and covert attention and vibrotactile reaction times Gaze direction, spatial compatibility and hemispatial asymmetry.
Canadian Journal of Psychology, 4 2, 44-56. Bradshaw, J.L., Nathan, G., Nettleton, N.C., Pierson, J.M., & Wilson, L.E. (1983). Head and body space to left and right: III. Vibrotactile stimulation and sensory andmotor components. Perception, 12, 651661. Brinkman, C., & Kuypers, H.G.J.M. (1973). Cerebral control of contralateral and ipsilateral arm, hand and finger movements in the split-brain rhesus monkey. Brain, 96, 653-674. Brizzolara, D., De Nobili, G.L., & Ferretti, G. (1982). Tactile discrimination of lines in relation to hemispheric specialization. Perceptual and Motor Skills, 54, 665-660.
Tactual Perception 489 Brodie, E.E. (1985). The measurement of differential thresholds for lifted weights by microcomputer controlled apparatus. Perception, 14, 371-376. Brodie, E.E. (1988). Sex and hand-preference effects in simultaneous and consecutive discrimination of lifted weight. Perception & Psychophysics, 43, 326-330. Brown, J.V, & Ettlinger, G. (1983). Intermanual transfer of mirrorimage discrimination by monkeys. Quarterly Journal of Experimental Psychology, 35, 119-124. Carmon, A., & Benton, A.L. (1969). Tactile perception of direction and number in patients with unilateral cerebral disease. Neurology, 19, 525-532. Carmon, A., Bilstrom, D.E., & Benton, A.L. (1969). Thresholds for pressure and sharpness in the fight and left hands. Cortex, 5, 27-35. Carnahan, H., & Elliot, D. (1987). Pedal asymmetry in the reproduction of spatial locations. Cortex, 2 3, 157-159. Charron, J.F., Collin, I., & Braun, C.M. (1996). Intermanual transfer of somaesthetic information: a two-point discrimination experiment. Neuropsychologia, 34, 873-877. Cholewiak, R.W., & Collins, A.A. (1991). Sensory and physiological bases of touch. In Heller, M.A. & Schiff, W. (Eds.), The psychology of touch (pp. 23-60). Hillsdale: Lawrence Erlbaum Associates. Cioffi, J., & Kandel, G.L. (1979). Laterality of stereognosic accuracy of hildren for words, shapes and bigrams: A sex difference for igrams. Science, 204, 1432-1434. Cohen, L., & Levy, J. (1986). Cerebral and sex differences in the categorization ofhaptic information. Cortex, 22, 253-259. Cohen, H,. & Levy, J. (1988). Hemispheric specialization for tactile perception opposed by controlateral noise. Cortex, 24, 425-431. Colley, A. (1984). Spatial location judgements by right and lefthanders. Cortex, 20, 47-53. Corkin, S. (1965). Tactually-guided maze learning in man: Effects of unilateral cortical excisions and bilateral hippocampal lesions. Neuropsychologia, 3, 339-351 Corkin, S. (1978). The role of different cerebral structures in somesthesic perception. In Carterette, E.C. & Friedman, M.P. (Eds.), Handbook of perception, vol. VIB: Feeling and Hurting (pp. 105155). New York: Academic Press. Cranney, J., & Ashton, R. (1982). Tactile spatial ability : Lateralized performance of deaf and hearing age groups. Journal of Experimental Child Psychology, 34, 123-134 DeRenzi, E., Faglioni, P., & Scotti, G. (1970). Hemispheric contribution to exploration of space through the visual and tactile modality. Cortex, 6, 191-203. DeRenzi, E., Faglioni, P., & Scotti, G. (1971). Judgment of spatial orientation in patients with focal brain damage. Journal of Neurology, Neurosurgery and Psychiatry, 34, 489-495. Dodds, A.G. (1978). Hemispheric differences in tactuo-spatial processing. Neuropsychologia, 16, 247-254. Duda, P.D., & O' Adams, J. 1987. Dichaptic-visual form matching in adults. Cortex, 2 3, 3 99-414.
~
490
Fagot, Lacreuse, and Vauclair
Ettlinger, G. (1961). Lateral preferences in monkeys. Behaviour, 17, 275-287. Ettlinger, G., & Moffet, A. (1964). Lateral preferences in the monkey.
Nature, 204, 606.
Fagot, J., Arnaud, B., Chiambretto, M., & Fayolle, R. (1992). Lateralization in haptic processing: An apparatus to analyze manual strategies. Behavior Research Methods, Instruments and Computers, 24, 54-59. Fagot, J., Dr6a, C., & Wallen, K. (1991). Asymmetrical hand use in rhesus monkeys (Macaca mulatta) in tactually and visually regulated tasks. Journal of Comparative Psychology, 105, 260-268. Fagot, J., Hopkins, W.D., & Vauclair, J. (1993). Hand movements and hemispheric specialization in dichhaptic explorations. Perception, 22, 847-853. Fagot, J., Lacreuse, A., & Vauclair, J. (1993). Haptic discrimination of nonsense shapes: Exploratory strategies but not accuracy reveal laterality effects. Brain and Cognition, 21, 212-225. Fagot, J., Lacreuse, A., & Vauclalr, J. (1994). Hand movement profiles -in a tactual-tactual matching task: Effects of spatial factors and laterality. Perception & Psychophysics, 56, 347-355. Fennell, E., Satz, P., & Wise, R. (1967). Laterality differences in the perception of pressure. Journal of Neurology, Neurosurgery and Psychiatry, 30, 337-340. Flanery, R.C., & Bailing, J.D. (1979). Developmental changes in hemispheric specialization for tactile spatial alsllity. Developmental Psychology, 15, 364-372. Fontenot, D.J., & Benton, A.L. (1971). Tactile perception of direction in relation to hemispheric locus of lesion. Neuropsychologia, 9, 83-88. Foulke, G. (1964). Transfer of a complex perceptual skill. Perceptual and Motor Skills, 18, 733-740. Franco, L., & Sperry, R.W. (1977). Hemisphere lateralization for cognitive processing of geometry. Neuropsychologia, 15, 107-114. Garclaa, H.S., Ettlinger, G., & MacCabe, J.J. (1982). Unilateral removal of the second somatosensory projection cortex in the monkey: Evidence for cerebral predominance.9 Brain, 105, 787-810. Gazzaniga, M.S., Bogen, J.E., & Sperry, R.W. (1963). Laterality effects in somesthesis following cerebral commissurotomy in man.
Neuropsychologia, 1,209-215.
Ghent, L. (1961). Developmental changes in tactual threshold on dominant and nondomlnant sides. Journal of Comparative and Physiological Psychology, 54, 670-673. Gibson, J.J. (1962). Observations on active touch. Psychological Review, 69, 477-491. Gibson, C., & Bryden, M.P. (1983). Dichhaptic recognition of shapes and letters in children. Journal of_Psychology, 37, 132-143. Greenspan, J.D., & McGillis, S.L. (1994). Thresholds for the perception of pressure, sharpness and mechanically evoked cutaneous pain: Effects of laterality_ and repeated testing. Somatosensory and Motor Research, 11, 311-317.
Tactual Perception 491 Harley, J.P., & Grafman, J. (1983). Fingertip number writing errors in hospitalized non-neurologic patients. Perceptual and Motor Skills, 56, 551-554. Hatta, T. (1978). The functional asymmetry of tactile pattern learning in normal subjects. Neuropsychologia, 21, 83-89. Heller, M.A. (1982). Visual and tactual texture perception: Intersensory cooperation. Perception & Psychophysics, 31,339-344. Heller, M.A., Rogers, G.J., & Perry, C.L. (1990). Tactile pattern recognition with the Optacon: Superior performance with active toucfi and the left hand. Neuropsychologia, 28, 1003-1006. Hellige, J.B. (1993). Hemispheric Asymmetry. What's Right and What's Left. London: Harvard University Press. Hermelin, B., & O'Connor, N. (1971). Spatial coding in normal, autistic and blind children. Perceptual and Motor Skills, 33, 127-132. Hoerster, W., & Ettlinger, G. (1985). An association between hand preference and tactile discrimination performance in the rhesus monkey. Neuropsychologia, 23, 411-413. Honda, H. (1977). Hemispheric laterality effects on tactile perception of direction in normal subjects. Tohoku Psychologica Folia, 36, 68-74. Iaccino, J.F. (1993). Left brain-right brain differences. Hillsdale: Lawrence Erlbaum Associates. Iwamura Y., Iriki A., & Tanaka M. (1994). Bilateral hand representation in the postcentral somatosensory cortex. Nature, 369, 554-556. Kandel, E.R., & Schwartz, J.H. (1981). Principles of Neural Science. London: Edward Arnold. Kurian, G., Sharma, N.K., & Santhakumari, K. (1989). Left-arm dominance in active positioning. Perceptual and Motor Skills, 68, 1312-1314. Lacreuse, A., & Fragaszy (in press). Manual exploratory procedures and asymmetries for a haptic search tasks: A comparison between capuchins (Cebus apella) and humans. Laterality. Lacreuse, A., & Fragaszy, D.M. (1996). Hand preferences for a haptic searching task by tufted capuchins (Cebus apella). International Journal of Primatology, 17, 613-632. Lacreuse, A., Fagot, J., & Vauclair, J. (1996). Lateralisation hEmisphErique et strategies d'exploration dans des taches de perception tactilo-kinesthEsique. L'Annde Psychologique, 96, 131-145. Lechelt, E.C. (1980). Laterality differences and shifts as a function of tactile temporal pattern complexity: Implication for differential processing strategies of the cerebral hemispheres. Psychological Research, 41, 319-333. Lederman, S.J., Jones, B., & Segalowitz, S.J. (1984). Lateral symmetry in tactual perception of roughness. Canadian Journal of Psychology, 38, 599-609. Leonard, G. & Milner, B. (1991). Contribution of right frontal lobe to the encoding and recall of kinesthetic distance information. Neuropsychologia, 29, 47-58. Loomis, J.M., & Lederman, S.J. (1986). Tactual Perception. In Boff, K.R., Kaufman, L. & Thomas, J.P. (Eds.), Handbook of Perception
and Human Performance: vol. 2: Cognitive Processes and Performance (pp 1-41). New York: Wiley.
492
Fagot, Lacreuse, and Vauclair
Millar, S. (1984). Is there a "best hand" for Braille? Cortex, 20, 75-87. Millar, S. (1987). The perceptual "window" in two-handed Braille: Do the left and right hands process text simultaneously ? Cortex, 23, 111-122. Milner, A.D. (1969). Distribution of hand preferences in monkeys. Neuropsychologia, 7, 375-377. Milner, B., & Taylor, L. (1972). Right-hemisphere superiority in tactile pattern-recognition after cerebral commissurotomy: Evidence for nonverbal memory. Neuropsychologia, 10, 1-15. Minagawa, N., & K-ashu, K. (1989). Influence of cognitive style and interstimulus interval on the hemispheric processing of tactile stimuli. Perceptual and Motor Skills, 68, 1031-1039. Mommers, M.J.C. (1980). Braille reading: Effects of different hand and finger usage. Visual Impairment and Blindness, 74, 338-343. Mousty, P., & Bertelson, P. (1985). A study of Braille reading: 1. Reading speed as a function of hand usage and context. The Quarterly Journal of Experimental Psychology, 3 7A, 217-233. Myers, D.H. (1976). Right- and left-handed counting of Braille dots in subjects unaccustomed to Braille. British Journal of Psychology, 67, 407-412. Nachson, I., & Carmon, A. (1975). Hand preferences in sequential and spatial discrimination tasks. Cortex, 11, 123-131. Nebes, R.D. (1971). Superiority of the minor hemisphere in commissurotomized man for the perception of part-whole relations. Cortex, 7, 333-349. Nishizawa, S. (1991). Different pattern of hemisphere specialization between identical kinesthetic spatial and weight discrimination tasks. Neuropsychologia, 29, 305-312. Nishizawa, S., & Saslow, C. (1987). Lateralization of kinesthetically guided spatial perception. Cortex, 23, 485-494. Oscar-Berman, M., Rehbein, L., Porfert, A., & Goodglass, H. (1978). Dichhaptic hand-order effects with verbal and nonverbal tactile stimulation. Brain and Language, 6, 323-333. Parr, L.A., Hopkins, W.D., & De Waal, F.B.M (in press). Haptic discrimination in capuchin monkeys (Cebus apella): Evidence of manual specialization. Neuropsychologia. Pierson-Savage, J.M. & Bradshaw, J.L. (1987). Mapping of extracorporeal space by vibrotactile reaction times: A far left side disadvantage. Perception, 16, 283-290. Rhodes, D.L., & Schwartz, G.E. (1981). Lateralized sensitivity to vibrotactile stimulation: Individual differences revealed by interaction of threshold and signal detection tasks. Neuropsychologia, 19, 831-835. Riege, W.H., Metter, E.J. & Williams, M.V. (1980). Age and hemispheric asymmetry in nonverbal tactual memory. Neuropsychologia, 18, 707-710. Riolo-Quinn, L. (1991). Relationship of hand preference to accuracy on a thumb-positioning task. Perceptual and Motor Skills, 73, 267-273. Roy, E., & McKenzl~e, C. (1978). Handedness effects in kinesthetic spatial location judgements. Cortex, 14, 250-258.
Tactual Perception
493
Rudel, R.G., Denckla, M.B. & Spalten, E. (1974). The functional asymmetry of Braille letter learning in normal, sighted children. Neurology, 24, 733-738. Schiavetto, A., Lepore, F., & Lassonde, M. (1993). Somesthetic discrimination threshold in the absence of the corpus callosum. Neuropsychologia, 7, 695-707. Semenza, C., Zopello, M., Gidiuli, O., & Borgo, F. (1996). Dichaptic scanning of Braille letters by skilled blind readers: Lateralization effects. Perceptual and Motor Skills, 82, 1071-1074. Semmes, J. Weinstein, S., Ghent, L., & Teuber, H.L. (1955). Spatial orientation in man after cerebral injury: I. Analysis by locus of lesion. Journal of Psychology_, 39, 227-244. Semmes, J. Weinstein, S., Ghent, L., & Teuber, H.L. (1960). Somatosensory Changes after Penetrating Brain Wounds in Man. Cambridge: Harvard University Press. Smith, M.O., Chu, J., & Edmonston, W.E. (1977). Cerebral lateralization of hapticperception: Interaction of responses to Braille and music reveals a functional basis. Science, 197, 689-690. Streitfeld, B. (1985). Hemispheric advantages and shifts of laterality differences in visual and tactual modalities. International Journal Of Neuroscience, 25, 245-254. Sullivan, E. (1989). Hemispheric asymmetry in tactile forgetting induced by tactually-guided movement. Cortex, 25, 83-92. Summers, D.C. & Lederman, S. (1990). Perceptual asymmetries in the somatosensory system: A dichhaptic experiment and critical review of the literature from 1929 to 1986. Cortex, 26, 201-226. Thibault, A., Forget,. R. & Lambert, J. (1994). Evaluation of cutaneous and proprloceptlve sensation in children: A reliability study. Developmental Medicine and Child Neurology, 36, 796-812. Varney, N.R., & Benton, A.L. (1975). Tactile perception of direction in relation to handedness and familial handedness. Neuropsychologia, 13, 449-454. Vauclair, J. (1996). Animal cognition. An introduction to modern comparative psychology. CamlSridge: Harvard Univeristy Press. Verjat, I. (1988). La dissym6trie fonctionnelle c6r6brale dans la modalit6 tactilo-kinesth6slque manuelle. L'Annde Psychologique, 88, 83-109. Verjat, I. (1989). A propos de l'exploration dichhaptique. L'Annde Psychologique, 89, 277-289. Ward, J.P., Alvis, G.R., Sanford, C.G., Dodson, D.L., Pusakulich, R.L. (1989). Qualitative differences in tactuo-spatial motor learning by left-handers. Neuropsychologia, 2 7, 1091-1099. Webster, W.G, & Thurber, D. (1978). Problem solving strategies and manifest brain asymmetry. Cortex, 14, 474-484. Weinstein, S. (1962). Differences in effects of brain wounds implicating fight or left hemispheres. In Mountcastle, V.B.(Ed.), Interhemispheric Relations and Cerebral Dominance (pp. 159-176). Baltimore: Johns Hopkins Press. Weinstein, S. (1968). Intensive and extensive aspects of tactile sensitivity as a function of body part, sex, and laterality. In Kenshalo, D.R. (Ed.), The Skin Senses (pp. 195-222). Springfield: Thomas.
494
Fagot, Lacreuse, and Vauclair
Weinstein, S. (1978). Functional cerebral hemispheric asymmetry. In Kinsbourne, M. (Ed.) Asymmetrical Function of the Brain (pp 1748). Cambridge: Cambrige University Press,. Weinstein, S., & Sersen, E.A. (1961). Tactual sensitivity as a function of handedness and laterality. Journal of Comparative and Physiological Psycho logy,_:54, 665-669. Wiles, P.G., rearce, S.M., Rice, P.J.S., & Mitchell, J.M.O (1990). Reduced vibration perception in right hands of normal subjects: An acquired abnormality? British Journal of Industrial Medicine, 47, 715-716. Wilkinson, J.M. & Carr, T.H. (1987). Strategic hand use preferences and hemispheric specialization in tactual reading: Impact of the demands of perceptual encoding. Brain and Language, 3 2, 97-123. Witelson, S.F. (1974). Hemispheric specialization for linguistic and nonlinguistic tactual perception using a dichotomous stimulation technique. Cortex, 1 O, 3-17. Yamamoto, M., & Hatta, T. (1980). Hemispheric asymmetries in a tactile thought task for normal subjects. Perceptual and Motor Skills, 50, 469-471. Yandell, L., & Elias, J. (1983). Left hemispheric advantage for a visuospatial dichaptic matching task. Cortex, 19, 69-67. Young, A.W., & Ellis, A.W. (1979). Perception of numerical stimuli felt by fingers of the left and right hands. Quarterly Journal of Experimental Psychology, 31,263-272. Zoccolotti, P., Passafiume, D., & Pizzamiglio, L. (1979). Hemispheric superiorities on a unilateral tactile test: Relationship to cognitive dimensions. Perceptual and Motor Skills, 49, 735-742.
S E C T I O N VII: OLFACTORY PROCESSING
This Page Intentionally Left Blank
Cerebral Asymmetries in Sensory and Perceptual Processing S. Christman (Editor) 9 1997 Elsevier Science B.V. All rights reserved.
49'/
Chapter 14
Laterality in Human Nasal Chemoreception.
Richard L. Doty, Steven M. Bromley, Paul J. Moberg & Thomas Hummel Smell and Taste Center and Department of Otorhinolaryngology: Head and Neck Surgery University of Pennsylvania Medical Center
As we interact with the environment, airborne chemicals and other volatile agents enter our noses and bombard the olfactory receptors. These finely-tuned microscopic structures determine, to a large degree, the flavors of foods and beverages, and warn us of such environmental hazards as leaking natural gas, spoiled food, and polluted air. Just as in the case of our two eyes and two ears, evolution has provided us with two separate nasal passages, each of which harbors a receptor-bearing olfactory epithelium. What is the physiological significance, if any, of this duality with regard to sensory lateralization? Is there a physiological advantage to this anatomical arrangement, other than providing redundancy? Although these questions have been debated for many years, considerable controversy still continues to exist as to whether asymmetries in olfactory function are, in fact, present at all.
498
Doty et al.
A major reason for investigating lateralization in olfactory function becomes apparent when one considers the atypical anatomy of this system. Among the senses, olfaction is unique in that second order neurons (whose cell bodies lie within the olfactory bulb) send information directly, with primarily ipsilateral projections, to the older portions of the telencephalon before reaching the thalamus and the neocortex (Gordon & Sperry, 1969; Youngentob et al., 1982; Hummel et al., 1995). However, some contralateral afferent and efferent connections exist between the two sides of the olfactory system via the anterior commissure, the corpus callosum, and conceivably the poorly understood hippocampal commissure. Although such connections are believed to be comparatively minor, they clearly exist in humans and other mammals. A classic example of the role of the anterior commissure in mediating cross-hemispheric exchange of information can be found in studies of infant rats conditioned to move toward an odor when only one naris is open. These rats are then tested with opposite (previously occluded) naris open to see if the conditioning transferred to the contralateral side. When six-day-old rats, who lack a mature anterior commissure, are trained in this manner, no transfer of training to the opposite side occurs. In contrast, 12-day-old rats with a mature anterior commissure demonstrate recall of information on the contralateral side. If the anterior commissure is sectioned prior to training of such rats, recall is confined to the side of training, just as occurs in 6-day-old rats (Kucharski & Hall, 1987, 1988). Such a phenomenon is seen using a number of behavioral paradigms, including those of odor aversion and habituation (King & Hall, 1990; Kucharski, Arnold & Hall, 1995). Most information related to laterality of human nasal chemoreception comes from the study of six types of subjects: (i) epileptic patients whose epileptogenic foci are lateralized to one or the other side of the brain; (ii) stroke patients with damage localized to one side of the brain; (iii) epileptic patients whose corpora callosa have been sectioned to prevent seizure activity from spreading contralaterally (so-called "split-brain" patients); (iv) epileptic patients who have received unilateral frontal and/or temporal lobe resection to mitigate their seizure activity; (v) patients with hemiparkinsonism whose dopaminergic deficits are asymmetrical; and (vi) normal subjects. Variation in the quality or type of olfactory testing, size and location of the epileptogenic focus, and, in the case of resection studies, the amount and location of the tissue removed, unfortunately complicate comparisons of results among a
Laterality and Olfaction 499 number of these studies. Nevertheless, the bulk of the research, including "split-brain" experiments, suggests that left and right hemispheres can function quite independently and that there may be some differences in the representation and processing of olfactory information by the two hemispheres. This occurs in despite the fact that odor detection, identification, and discrimination can apparently be performed by each hemisphere independently. In this chapter, we provide an overview of what is presently known about the anatomy of major nasochemosensory systems and the degree to which these systems exhibit laterality, drawing upon anatomical, psychophysical, imaging, and electrophysiological studies. This review attempts to shed light on the degree to which the two sides of the nose provide independent information to higher brain structures and how these structures filter, integrate, interpret, and retrieve such information. ANATOMY OF THE O L F A C T O R Y AND T R I G E M I N A L CHEMOSENSORY SYSTEMS Odorants enter the nose during both active (e.g., sniffing) and passive inhalation; most have the propensity to stimulate both olfactory receptors (CN I) located in the upper recesses of the nasal vault, and free nerve endings of the ophthalmic and maxillary divisions of the trigeminal nerve (CN V), distributed throughout the nasal mucosa and the olfactory neuroepithelium (Figure 1). 1 Sensations derived from CN I stimulation are those of odors (e.g., flowers, lemon, grass, fish, etc.), although at low concentrations minute sensations that lack qualitative character, "something more than nothing," can be perceived. Sensations derived from CN V stimulation are somatosensory, and include tactile sensations, burning, cooling, tickling, warming, and the perception of atmospheric humidity, "thickness," or "fullness." Difficult-to-describe "feelings" can also be perceived from low concentrations of CN V stimulants. It is important to realize that a variety of odorants differentially stimulate these two systems, and that CN I and CN V differ in terms of their central projections and the degree to which their pathways project contralaterally and ipsilaterally. The weight of the evidence suggests that CN I stimulants presented to one nasal chamber cannot be localized to that chamber; however, this is not the case with CN V stimulants (von Skramlik, 1925; Schneider & Schmidt, 1967; Ehrlichman, 1986; Kobal, van Toiler & Hummel, 1989).
500
D o t y et al.
/ t~i~i~ ~.,.,
A: l a t e r a l Ant. ethmoidal ,,,~,.,,,~
~ ~ " ' ' / .,..~
Infraorbital nerve
9
,,~iiiiiiiiiiiiiiiiiiiiii~~~~,,a. t ~ - : ' ~ ~ U ~ ;~"."~,i. i~ /../.~J..[ ') :.': g ~ ( ! ' : ~ , // .. . ... .. !: i(.......,...:.... ,.
".
~,~-'"'*~h,.
-
~:
d
L "
:'....."~-,~....i];"~ . f M a x i l l a r y .....-.."..... ~ : .. . ~ ............. nerve : . . . ~'........................ ~ - : ~ < % . !~~~~:~=. ~- = " ................:.----=.-.:.... r~ --...:~.. .... ,,, ...... .. :""~'~:'$.-., ":",%"*%-. .
'
g i 1a olfactona
. . . . ~ ~ . . .
/6,'~'~'x'~'~'a"~.'i',~-~a,
....
...........................
9
.........
~::"
,~:
~.:;:
""r
9
.::,
~ . . . . ,, ................................................... .. ~~ii~::~{~::~i~::~::~::i::~i~::~::~::~::i::~~i~~..`...*....~<~..~..~~.. ,,,,.....:...... ,,......., . , - 9~
"~..':.:.:.:.:.:.:.....-.................. ......,.,,..o...,,,e,,,,,-.,,,,.
B: se p- t' a- I
,,we . . . . .
6w.
.......
g
b. . . . . . n e r v e ~ . . . . ~. . . . . .
Fila
~ ~.).~ ~ . ~ ..~t i, i ,,i i |
Ethmoidal j .~!iiiiiiii!i:<:~~~~:k..~. nerve ; - ~ . ~ ~ ,~i:Q"i'~ ~ ,
~ 9
9
.e~"
.X. ~~.i) a ~ ,. .
olfactoria .....
Posterior
~.ff-':
ethmoidal
: "" ~..~ _~ .. ' ~ , , , ................. ~,, . . . J
Post. sup.
. ~~g'.~ / 9...................... ~ ~ ................................................ ~................................... -~~
med. nasal nerve
NasoDlantlne
N.
Figure 1. Anatomical diagram of olfactory receptors (CN I) located in the upper recesses of the nasal vault, and free nerve endings of the ophthalmic and maxillary divisions of the trigeminal nerve (CN V), distributed throughout the nasal mucosa and the olfactory neuroepithelium.
Laterality and Olfaction
501
O L F A C T O R Y SYSTEM
Olfactory System Sensory Receptors and Primary Neurons CN I consists of millions of bipolar receptor cells whose cell bodies, dendrites, and initial axon segments are found within the olfactory epithelium, a heterogeneous patch of tissue located on the cribriform plate, superior septum, and portions of the superior and middle turbinates (Moran, Rowley, Jafek & Lovell, 1982) (Figure 1). This pseudostratified columnar epithelium, which contains a variety of cell types, is supported by a highly vascularized lamina propria. Up to 50 cilia project into the mucus from the dendritic knob of each bipolar receptor cell. These cilia, a number of which radiate considerable distances across the epithelial surface (i.e., > 30 /~m), lack dynein arms and, hence, do not beat rhythmically, in contrast to cilia in the respiratory epithelium. They do, however, exhibit the typical 9+2 microtubule arrangement. Proteinaceous receptor sites are found on these cilia. The aggregate surface area of the cilia has been estimated to be nearly nine square inches in humans (Doty, 1997). It is presently believed that only one type of receptor is genetically expressed on each receptor cell; individual receptors, however, may respond to a number of chemicals and more than one receptor cell can share the same type of receptor. The unmyelinated axons of the olfactory receptor cells coalesce into bundles of approximately 200 axons surrounded by ensheathing or Schwann cell mesoaxons within the lamina propria. Axons within these bundles are in direct contact with one another and may interact metabolically and electrically (Gesteland, 1986). These bundles, in turn, combine with other bundles to form the olfactory ilia which traverse the foramina of the cribriform plate, subsequently forming a dense layer of axons on the surface of the olfactory bulb ipsilateral to their origin. From this point the receptor cell axons branch and synapse with second order neurons within the bulb.
Olfactory System Secondary Sensory Neurons: Olfactory Bulb and Projections The first synapse of the axon of the receptor cell occurs within a glomerulus, a globe-like structure present in most olfactory systems, including those of many invertebrates (Figure 2). The human olfactory
502
Doty et al.
ONL
....... ~..___~ f1,~
GL
,~
t
,
;i /
,,
,~
...\
. . . .
/~//
~..
..
~
,
k
d
~ ~ ..~ ~
9
I
II
I
\
I'
~..._
EPt
MBL[
....
---
"C ~ GRL
LOT-
Figure 2: Olfactory System Secondary Sensory Neurons: Olfactory Bulb, Glomeruli, and Projections
bulb typically has thousands of these 50 to 200/~m diameter structures arranged in single or double concentric layers. With age, the number and integrity of the glomeruli decrease greatly, so that they are nearly absent in most persons over the age of 80 (Smith, 1942). The development and maintenance of glomeruli depends on trophic influences exerted by receptor cells. A given receptor cell projects ipsilaterally to only one glomerulus, and a given glomerulus appears to receive most of its input from a restricted region of epithelium. As the concentration of an odorant is increased, additional glomeruli are activated, suggesting that groups of cells with higher thresholds are recruited (Stewart, Kauer & Shepherd, 1979).
Laterality and Olfaction 503 The main afferent second-order neurons, whose primary dendrites extend into the glomeruli, are termed mitral and tufted cells and project via the olfactory tract to secondary olfactory areas of the olfactory cortex (see Figure 2). The activity of these cells is modulated by several types of inhibitory neurons, including periglomerular, short-axon, and granule cells (Shipley & Reyes, 1991). Some mitral and tufted cells, when stimulated, respond to odorants in complex ways. For example, some initially exhibit inhibition and subsequently excitation, whereas others do the opposite. The character of the responses often are concentration dependent (Shipley & Ennis, 1996). It is important to note that the mitral and tufted cells form a necessary component of the reciprocal relations between the olfactory bulb and higher brain structures. The interneurons provide the circuitry through which positive and negative feedback occurs; hence, the formation of complex 'reverberating' circuits. Reciprocal inhibition between neighboring mitral or tufted cells through interneural connections reportedly sharpens the contrast between adjacent neural channels -- not unlike the reciprocal inhibition seen in the visual and tactile systems (MacLeod, 1971).
Olfactory System Central Connections In humans, the olfactory tract is relatively flat posteriorly and becomes the olfactory trigone just rostral to the anterior perforated substance, so named because of the many small holes for blood vessels that punctuate this region. At the edges of the trigone, the tract divides into the medial and lateral olfactory striae. Modem studies suggest that there is no medial olfactory tract in mammals (Price, 1990). Thus, all mitral and tufted cell axons apparently leave the olfactory bulb via the lateral olfactory tract to synapse on structures collectively termed the primary olfactory cortex. These structures include (a) the anterior olfactory nucleus (AON), (b) the olfactory tubercle (poorly developed in humans), (c) the prepiriform cortex, (d) the pirform cortex (a major olfactory discrimination region), (e) the periamygdaloid cortex (a region contiguous with the underlying amygdala), (f) the cortical nucleus of the amygdala, and (g) the entorhinal area (Brodal, 1981; Meyer & Allison, 1949; Powell, Cowan & Raisman, 1965; Price, 1990; Shipley & Ennis, 1996; Tanabe et al., 1975) (Figure 3).
504
Doty et al.
The AON is located in a dorsal segment of the olfactory bulb, as well as in a rostral part of the olfactory peduncle near the anterior perforated substance. The dendrites of the pyramidal cells of the AON receive synapses not only from the olfactory bulb, but from both ipsilateral and contralateral brain structures, including the contralateral AON and olfactory cortex (see next section) via the anterior segment of the anterior commissure. Thus, the anterior commissure represents the first CNS structure through which olfactory information crosses contralaterally. This contralateral projection suggests that impulses from one anterior olfactory nucleus can reach bulbar granule cells, as well as periglomerular cells, on both sides, as well as the contralateral anterior olfactory nucleus. Therefore, it is likely that input from one side can be modulated by that from the other side. The effect of this retrobulbar influence is to alter the pattern of information received by the respective cortices within the right and left hemispheres. Although the proportion of olfactory-related fibers within the anterior commissure is unknown, this commissure is the second largest in the brain, with only the corpus callosum being larger. In the rhesus monkey, the anterior commissure contains over three million axons. While this is approximately 5% of the 56 million axons of the corpus callosum, these two commissures together account for 99% of the axons which connect the two hemispheres (Lamantia & Rakic, 1990). Extensive interactions occur among the cells that comprise the superficial and deeper laminae of each component of the olfactory cortex, as well as among the components themselves (Shipley & Reyes, 1991). In addition to receiving direct projections from the olfactory bulb, the entorhinal cortex also receives input from both the prepiriform and periamydaloid cortices (Powell, Cowan & Raisman, 1965). The entorhinal cortex sends productions to a number of cortical structures, including the hiptx)campus (Powell, Cowan & Raisman, 1965). However, no direct pathways exist between the hippocampus and either the olfactory bulb or the AON, and anatomic, physiologic, and ontogenic evidence suggests that the hippocampus is not a main component of the olfactory system (Brodal, 1947; Carpenter, 1985). Nevertheless, it does play a role in odor memory (Squire & ZolaMorgan, 1991). The hippocampus seems to exert its influence on olfaction via its efferent connections to the olfactory cortex. The main effector system of the hippocampal formation is the fornix, which includes both projection and commissural fibers. From the hippo-
Neocortex
l/o I
-I
: ,.-I
Superior ' ' Y l Temporal Culcus q l -
Primary Olfactory Cortex m
c_
Olfactory Epithelium
@
I I
I I
@
P-
I
ii
i
I
Anterior Olfactory Nucleus
' Orbitofrontal and ...--- Dorsolateral , Cortices I I |
Prepiriform and Piriform Cortices Olfactory Bulb
Insular Cortex Thalamus: Mediodorsal n. Medioventral n. Brain Stem Regions
:::::::::::::::::::::::::::::::::::::::::::::::::::::
,Autonomic Functions
:::::::::::::
q...~
r
9 Pituitary
rculation @
O
506
Doty et al.
campus, the fornix forms an arch-like structure in each hemisphere under the splenium of the corpus callosum. In this region, crossover to the opposite hemisphere occurs via a thin sheet of fibers known as the hippocampal commissure (also termed the commissura fornicis). Although this structure is rather poorly developed in man, it does form, nonetheless, a connection between the two hemispheres. In humans, most projections of the mitral and tufted cells are to the more rostral elements of the primary olfactory cortex. No obvious point-to-point topography exists between the olfactory bulb regions and cortical projection sites (Shipley & Reyes, 1991). Thus, small areas of the bulb can project to large areas of the olfactory cortex and vice versa (Price, 1990). Reciprocal connections are present between the olfactory cortex and orbitofrontal cortex, the mediodorsal and submedial thalamic nuclei, the lateral hypothalamus, the amygdala, and the hippocampus (Price, 1990). A major pathway from the olfactory cortex to the orbitofrontal cortex is via the mediodorsal nucleus of the thalamus. One feature of the olfactory system that distinguishes it from other sensory systems is a tremendously rich supply of centrifugal fibers (Figure 4). Centrifugal fibers enable the central nervous system to modify and control the incoming flow of olfactory sensory signals (Hagbarth, 1960). In the olfactory system, most of these fibers act directly on the periglomerular and granule cells of the olfactory bulb (Davis & Macrides, 1983; Shipley & Ennis, 1996) (Figure 3), although additional control is exerted beyond the bulb at higher levels (Davis & Macrides, 1983; Haberly & Price, 1978; Shipley & Ennis, 1996). The most prominent projections to the olfactory bulb appear to come from the pyramidal cells of the AON (Carson, 1984; Kratskin, 1995), although animal studies suggest that a substantial number arise from the piriform cortex (Haberly & Price, 1978), the lateral entorhinal cortex (Kratskin, 1995), the amygdaloid areas (Kratskin, 1995), the nucleus of the lateral olfactory tract (DeOlmos, et al., 1978; Carson, 1984), the diagonal brand of Broca (DeOlmos, et al., 1978; Price & Powell, 1970), the raphe nuclei (Carson, 1984; McLean & Shipley, 1987), the locus coeruleus (Shipley, et al., 1985), and regions of the hypothalamus (Carson, 1984; Shipley & Adamek, 1984; DeOlmos, et al., 1978). With regard to lateralization, centrifugal neurons mostly project to the ipsilateral olfactory bulb, although bilateral projections in addition to those to the AON occur to the nucleus of the lateral olfactory tract, the raphe nuclei, and the locus coeruleus (Kratskin, 1995). The centrifugal
Laterality and Olfaction 507
Basal Forebrain
Olfactory Bulb
Prepiriform and Piriform Cortices
Yi!iii i!iiiii
Lateral Entorhinal Cortex Amygdala: Posterolateral and Medial Cortical Areas
Hypothalamic Region
i!iiiiii:ii!ii:i!i!i!ii!i!i!iiiiiiiiiiiiiiiii!ii
i
i;i i i i i i ::i~i i i~i i i il;i;i i i i i i :.;
I
Figure 4. Centrifugal fibers in the olfactory system. axons appear to terminate mainly on inhibitory interneurons (Kratskin, 1995). Because of these centrifugal connections, it is possible that the olfactory bulb is involved in non-olfactory neural functions associated with the limbic system, in addition to the bulb's usual sensory functions (Cain, 1974, Shepherd, Nowycky, Greer & Mori, 1981; Kratskin, 1995). While it is not completely clear which elements of olfactory processing are handled by various cortical structures, it is generally believed that cortical structures are involved in the conscious perception of smell (Slotnick & Kaneko, 1981, Eichenbaum, Shedlack & Ekmann, 1980, Potter & Butters, 1980; Staubli, Schottler & Nejat-Bina, 1987). Support for this notion comes from the fact that primates with lesions in orbitofrontal cortex have difficulty discriminating odors (Tanabe et al., 1975). Similarly, amygdalotomies reportedly induce deficits in odor discrimination (Andy, Jurko & Hughes, 1975). In multiple sclerosis, the number of lesions (i.e., plaques) in the inferior medial temporal and frontal lobes, as determined by magnetic resonance imaging (MRI), correlate strongly (r - -0.94) with odor identification test scores (Doty, Li, Mannon & Yousem, 1997). Odor memory, a component of many tests of olfactory function, has been specifically linked to the hippocampus by studies reporting rapid forgetting of olfactory information following hippocampal lesions (Staubli, Ivy & Lynch, 1984).
508
Doty et al.
There has been suggestion that subcortical brain regions may play a role in aspects of lateralized odor perception (Figure 3), although the specific circuitry involved is poorly understood. For example, unilateral lesions within lateral sectors of the cat's mesencephalic brain stem appear to produce significant alterations in odor-related behaviors on the side opposite to the lesions (Sprague, Chambers & Stellar, 1961, p. 166): "When fresh meat was presented on the ipsilateral side, the hungry cat [whose vision was blocked] would sniff and bat at it with its paw. When the meat was passed to the contralateral side, however, sniffing ceased and there was no sign of orientation to it." Whether this is an example of a true sensory alteration, a sensory "neglect" phenomenon, or disruption of circuits involved in a motor program related to the execution of an olfactory-related task, is unknown. TRIGEMINAL SYSTEM The nasal cavity is innervated by the ophthalmic and maxillary branches of the trigeminal nerve (Lang, 1989; see Figure 1). The anterior ethmoidal nerve runs through the cribroethmoidal foramen and then passes forward and downward along the border between the nasal septum and the lateral wall of the nose. Within the nasal cavity numerous branches of the nerve pass to the anterior segment of the nasal septum, from the cribriform plate down to an area lying above the premaxilla. The nasopalatine nerve arises from the maxillary nerve and passes through the sphenopalatine foramen to reach the nasal septum. During its course within the nasal cavity it sends numerous branches to the septal mucosa. Electrophysiological data indicate that an area of increased trigeminal chemosensitivity might be found at the anterior third of the septum (Hummel, Schiessl, Wendler & Kobal, 1996). Fibers of the trigeminal nerve have been shown to innervate the olfactory epithelium (Finger, Getchell, Getchell & Kinnamon, 1990~ Read, 1908; von Brunn, 1892). Several studies indicate that olfactory receptor responses to chemical stimuli can be modified by activation of the trigeminal nerve and that such changes result from both central (e.g., Stone, Williams & Carregal, 1968; Stone, 1969) and peripheral (Bouvet, Delaleu & Holley, 1987, 1988) interactions. The latter interactions may take place via an axon reflex (Finger, Getchell, Getchell & Kinnamon, 1990).
Laterality and Olfaction 509 Trigeminal afferents mediate sensations of touch, temperature and pain (for review, see Kelly & Dodd, 1991). In the case of gaseous CO2, which produces stinging and/or burning sensations when applied to the mucosa of the nose (e.g., Kobal, 1985), chemosensory nociceptive afferents (C-fibers, Adelta-fibers) are activated by the intracellular accumulation of protons (Steen, Wegner, Kreysel & Reeh, 1995). This increases a cation membrane conductance (Bevan & Yeats 1991; Konnerth et al., 1987) which exhibits slow desensitization (Steen, Reeh, Anton & Handwerker, 1992). In all probability, the same channels are also opened by capsaicin, the hot ingredient in chili peppers (Bevan, Forbes & Winter, 1993; Steen & Reeh, 1993). The cell bodies of the trigeminal afferents lie in the Gasserian ganglion (Ggl. Semilunare) where it is possible to obtain electrophysiologic responses after chemical stimulation of the nasal mucosa with CO2 (Thurauf, personal communication). The axons project to the trigeminal sensory nucleus (that extends from the rostral spinal cord to the midbrain), and to the spinal, principal, and mesencephalic trigeminal nuclei; nociceptive afferents descend in the trigeminal tract and terminate in the spinal nucleus. Chemosensory fibers from the nasal cavity have been shown to project to the superficial laminae of the spinal nucleus, namely the subnucleus caudalis, and subnucleus interpolaris (Anton & Peppel, 1991). Trigeminal information is thought to be relayed to the amygdala from the trigeminal sensory nuclei via the lateral parabrachial complex (Bernard, Peschanski & Besson, 1989). Neurons of the spinal nucleus project to the ventral posterior medial and intralaminar nuclei of the thalamus; most ascending fibers cross to the contralateral side and travel with the anterolateral system while some fibers ascend ipsilaterally (Barnett et al., 1995). Olfactory-trigeminal interactions may take place in the mediodorsal nucleus of the thalamus where convergence between olfactory and trigeminal afferents occurs. Inokuchi, Kimmelman and Snow (1993) have shown that single neuronal responses following odorant stimulation in rats are enhanced after trigeminal afferent activity is blocked with a local anesthetic. The projection from the ventral posterior medial nucleus terminates in the primary somatosensory cortex (SI), however, trigeminal chemosensory stimulation produces activation of the secondary somatosensory cortex (Huttunen, Kobal, Kaukoronta & Hari, 1986; Kettenmann, Hummel, Stefan & Kobal, 1996). Interestingly, trigeminal stimulation of either the left or fight nasal chamber produces bilateral activation in the
510
Doty et al.
secondary somatosensory cortex (Kettenmann et al., 1996). Further cognitive processing of trigeminally mediated information may also occur in the ventral orbital cortex (Snow, Lumb & Cervero, 1992). THE SEARCH FOR A N A T O M I C A L A S Y M M E T R I E S IN BRAIN REGIONS R E L A T E D TO O L F A C T I O N
Numerous studies have found anatomic differences between olfactory structures on the left and right sides of the brain. In fact, the volume of the entire olfactory bulb, as well as its outer stratum, is significantly greater on the right than on the left in rats (Heine & Galaburda, 1986). Within some age groups of rats, the hippocampus is larger on the fight than on the left, although this phenomenon is either attenuated or reversed in females (Diamond, Murphy, Akiyama & Johnson, 1982; Sherman & Galaburda, 1985). Such differences extend to the entire hemisphere in many species including mice, cats, rabbits and dogs (Kolb et al., 1982; Rosen et al., 1989; Tan & Caliskan, 1987). For example, in a study of infant and adult rats, it was noted that the right cerebral cortex was thicker than the left by about 5%, and that the fight hemisphere was "heavier, wider, longer, taller, and extended further anteriorly than the left" (Kolb, Sutherland, Nonneman & Whishaw, 1982). In humans, similar structural asymmetries appear to be present that could conceivably affect olfactory processing. The left lateral ventricle is typically much larger than the right (i.e., implying the right hemisphere has greater mass), especially in men (Geschwind & Levitsky, 1968; LeMay, 1984). The anterior right hemisphere, incorporating a number of structures associated with higher-order olfactory processing, is larger than its left-side counterpart (Galaburda et al., 1978; LeMay, 1976; Kertesz, Polk, Black & Howell, 1990), particularly in righthanders. The right frontal and left occipital lobes are larger in terms of overall volume than their opposing counterparts (Weinberger, Luchins, Morihisa & Wyatt, 1982). In lower-order animals, this phenomenon appears to be attenuated or reversed in females (Bear et al., 1986). Whether the larger mass of the right hemisphere reflects asymmetry in olfactory processing is unknown. Although caution is warranted in inferring differential function from volumetric differences in brain mass, relationships among volumetric differences and function are seen in some other neural systems (e.g., those associated with language), adding weight to the possibility that the right hemisphere may play a
Laterality and Olfaction 51 1 role in olfactory processing different in type or degree from that of the left hemisphere. As noted later in this chapter, a number of psychophysical and functional imaging studies also support this notion, although the data are by no means unequivocal and contradictions abound. THE SEARCH FOR FUNCTIONAL ASYMMETRY IN HUMAN O L F A C T O R Y PATHWAYS Even though recent studies suggest electrophysiological responses within the two hemispheres may differ among stimuli that preferentially stimulate CN I and CN V (Hummel et al., 1995), the functional significance of such responses is not clear and no psychophysical studies of laterality have implicitly characterized odorants in terms of these properties. This is in spite of the fact that there is now a reasonable understanding, largely based upon human studies of anosmics (Doty et al., 1978) and rat electrophysiological studies (Silver & Moulton, 1982), of the trigeminal stimulative properties of a larger number of odorants. As discussed in this section, split-brain patients with anterior commissure and corpus callosum severance recall odor quality and identity information only from the side of the brain ipsilateral to the nasal chamber stimulated. CN V is presumably not involved in this phenomenon, since it does not depend upon these commissures for cross-hemispheric communication of information. Psychophysical Studies of Commissurectomized Epilepsy Patients In the late 1960s, Gordon and his associates, most notably Bogen and Sperry, performed olfactory tests on patients whose anterior commissures and corpora callosa were sectioned in an effort to mitigate the spread of intractible epileptic seizures. In one of these "split brain" experiments, approximately a half-dozen common odorants (e.g., coffee, fish, garlic, tobacco) were presented separately to each side of the nose for a small group of such individuals (Gordon & Sperry, 1969). Odors presented to the left naris were correctly and confidently named, whereas those presented to the fight were identified by name at a chance rate, implying that the information was not transferred to the left (verbal) side of the brain. Gordon (1974, p. 144) notes that some patients expressed concern about perceiving nothing in their fight naris,
512
Doty et al.
but usually rationalized the problem, saying, for example, "I have congestion in my right nostril today, so I can't smell anything." To determine whether the right side of the nose of these split-brain patients was able to perceive and respond to odors, Gordon and Sperry provided the patients with a set of objects representing the odorants given (e.g., a plastic lemon for lemon juice, a perfume bottle for perfume, a carved fish for fish oil, etc.). When the patients were allowed to point to the object representing the odor, four of five patients tested on the fight side were able to correctly identify the odor non-verbally. Severe fight hemisphere damage was suspected in the sole patient unable to perform this task, as inferred from a presurgical seizure history and post-operative motor deficits. In another experiment, two subjects were tested in a paradigm where the objects chosen to represent the odorants were kept out of sight behind a blind but could be felt and manipulated by each hand. In this situation, the odors presented to the right naris were consistently correctly identified by the left hand (which is controlled by the right hemisphere), but not by the right hand. Conversely, when the odors were presented to the left naris they were correctly identified by the right hand and not the left. "Further scrutiny of olfactory-tactual data ... showed that the left hemisphere performance was about 25 percent superior to that of the fight. That is, odor-object pairings were more often correct for the left nostril-right hand combination than for the fight nostril-left hand. A superiority, though much less striking, was also seen for the left nostril naming versus right nostril pointing tasks (Gordon, 1974, p. 148)." The basis for the aforementioned superiority of left hemisphere performance is not known. Among the possible explanations are that (i) the left hemisphere is more efficient at making cross-modal comparisons, (ii) receptors in the left olfactory epithelium, or olfactoryrelated pathways in the left side of the brain, are more efficient than those of the right, (iii) functional damage from the underling epilepsy was greater in the fight hemisphere than in the left in these specific patients, and (iv) the left hemisphere, with its centers for verbal function and bilateral motor control, interferes with the responses of the right hemisphere. Apparently, instructions given to the patient must first be dealt with by the left hemisphere, even in cases when functions of the fight hemisphere are being tested.
Laterality and Olfaction
513
In the early years of the following decade, Gordon, Bogen and Sperry (1971) tested olfaction in two patients who had undergone partial commissurotomy for intractible epilepsy. In these patients, only the anterior commissure and the anterior two-thirds of the corpus callosum were sectioned, leaving the splenium intact. One of these patients had very little smell function prior to the surgery, which remained so post-operatively. However, this was not the case with the other patient, whose test results were summarized by Gordon (1974, p. 149) as follows: In contrast to previous cases with complete surgical section, this patient retained the ability to name odors presented to his fight nostril at a level significantly above chance guessing even though left hemisphere lateralization of speech was confirmed by pre-surgical administration of intracarotid amobarbital. Furthermore, he could cross-match olfacto-tactual stimuli by successfully pairing odors with their associated objects for any of the nostril-hand combinations. This remarkable lack of disconnection symptomology for olfaction extended to other modalities as well. However, cross-transfer of visual and tactual information between the hemispheres can be attributed to fiber cross-projections from the occipitoparietal areas coursing through the intact splenium. On the contrary, olfactory fibers are known to cross the midline only in the anterior commissure which, according to the surgeons, is completely divided in this patient. One is forced to conclude that the nature of information that passes through the posterior one third of the corpus callosum either includes olfactory fibers p e r se or carries some abstracted version of additionally processed olfactory stimuli. It is noteworthy that this patient exhibited the same left nostril superiority as seen in those patients whose corpus callosum had been completely sectioned. Thus, over 25% more odors were recognized by the left than by the fight side of the nose. Such superiority was observed for the left nasal chamber in all tests, i.e., those of naming, pointing, and odor-object pairing. In 1975, Gazzaniga et al. confirmed, in another split-brain patient, that transfer of olfactory information occurs via the corpus callosum. In this patient, the anterior commissure had been sectioned, but only the anterior one-third of the corpus callosum. This patient (p. 12) "... showed an unstable performance on this [transfer] test. In one session he b e h a v e d as t h o u g h no o l f a c t o r y i n f o r m a t i o n was b e i n g interhemispherically communicated, while in other tests he behaved normally." One wonders whether the session-to-session variability noted in this study might reflect fluctuations in attentional or motivational
514
Doty et al.
state, as seen in some cases of unilateral neglect (Mesulam, 1984). Analogous transfer did not occur in other patients of this series whose anterior commissures and corpus callosa were completely sectioned. A 1978 study by this same group, which evaluated olfactory function in patients with severed corpora callosa and intact anterior commissures, also found interhemispheric transfer of olfactory information (Risse et al., 1978). As noted by the authors (p. 28), "These data suggest that the anterior commissure plays a prominent role in the interhemispheric integration of modality-specific sensory information. Although the [four] patients had widely differing histories and extracallosal brain damage, evidence of interhemispheric transfer of sensory information was obtained in every case following complete section of the corpus callosum." A decade after these classic studies, Eskenazi, Cain, Lipsitt and Novelly (1988) tested, on each side of the nose, the olfactory function of two callosotomy patients who also had intact anterior commissures. These patients had seizure onsets arising largely in right brain structures. Postoperatively, both patients exhibited "clinically normal thresholds with inconsistently better sensitivity in the left nostril." Furthermore, both patients exhibited a left side superiority for naming, as well as remembering, odors. Odor discrimination was superior on the left in the patient who received this test. Unlike earlier studies by Gordon et al., in which the anterior commissure was also sectioned, neither subject matched more odors to objects representing the odors when the tactile and olfactory information projected to the same hemisphere than when such information projected to different hemispheres. This suggests that an intact anterior commissure may somehow balance out the performance of the two hemispheres on this specific task. In summary, split brain studies uniformly demonstrate that while each side of the brain appears to be able to recognize and discriminate among odors, the left side of the brain seems to be superior with respect to recognition and identification.
Psychophysical Studies of Nonsurgicaily Treated Epilepsy Patients Temporal lobe epilepsy (and less commonly frontal lobe epilepsy) exhibits lateralized neurophysiological changes in brain regions associated with olfactory processing. The question arises as to whether the ability to smell is changed by epilepsy and, if so, whether the
Laterality and Olfaction 515 function is differentially influenced by left- or right-sided foci. Unfortunately, early threshold studies -- most of which reported heightened olfactory sensitivity -- did not specify the side of epileptogenic foci, and only bilateral testing was performed (e.g., Dimov, 1973; Campanella, Filla & DeMichele, 1978; DeMichele, Filla & Campanelli, 1976; Santorelli & Marrota, 1964). Such testing is insensitive to most unilateral deficits, since it largely reflects the best functioning side of the nose (Betchen & Doty, 1997). In the first study to test patients with respect to side of epileptogenesis, Eskenazi, Cain, Novelly and Mattson (1986) reported normal bilaterally-determined n-butanol odor detection thresholds in 18 epileptic patients relative to 17 normal controls, an observation confirmed for this odorant for each side of the nose by Martinez et al. (1993) in 21 epileptic patients. In accord with this observation, West, Doty, O'Connor and Sperling (1993) found, in 16 patients with left and 14 with right foci, no detection threshold sensitivity deficits on either side of the nose, relative to matched controls, to the odorant phenyl ethyl alcohol using a single staircase procedure. Several studies have found suprathreshold deficits in epilepsy patients. For example, Abraham and Mathai (1983) noted decreased performance on an odor-matching task in 28 epileptic patients with fight-sided foci. However, patients with left-sided foci did not exhibit the same difficulty in performing the bilaterally-administered olfactory matching task. Three years later, Eskenazi, Cain, Novelly and Mattson (1986) reported a small bilateral odor identification deficit in 18 epileptic patients relative to 17 normal controls which was apparently independent of the side of seizure focus. More recently, Carroll and Richardson (1993) found, relative to controls, a decrease in immediate odor memory to common odorants (e.g., vinegar, coconut, coffee, nail varnish, and garlic) in ten patients with right-sided temporal lobe epilepsy. Ten patients with left-sided temporal lobe epilepsy, as well as ten epileptics with a non-temporal lobe epileptic focus, did not show this deficit. Paradoxically, this difference was seen only for nameable (i.e. common) odors, and was not seen for nonnameable, uncommon, odors. Two other psychophysical studies have appeared on this general topic. In the first, Martinez et al. (1993) found deficits in odor discrimination, short- and long-term odor memory, and odor naming in 21 epileptics relative to normal controls. No differences between the left and right sides of the nose were present in these patients, who
516
Doty et al.
subsequently underwent temporal lobe resection. In the second study, West, Doty, O'Connor & Sperling (1993) also found, in 20 patients with left foci and 20 with right foci, decrements on each side of the nose in tests of odor identification and memory relative to matched controls. In summary, several psychophysical studies from the 1960's found bilaterally-determined olfactory sensitivity to be increased in patients with epilepsy. More recent studies, however, including ones using wellvalidated test procedures, find no such increases in sensitivity. Several studies suggest that fight foci may have a larger influence on measures of bilaterally-determined olfactory function than left foci, implying that either the left side is somewhat more protected from epilepsy-related impairment than the right or that the impairment, p e r se, is simply greater on the right than on the left. Studies that have examined olfactory function on each side of the nose are unanimous in concluding that detection thresholds are not altered in medicated patients with epilepsy, but other measures, including those of odor identification, memory, and discrimination, are depressed, usually bilaterally, in such patients. Most studies find no association between the side of epileptogenesis and the degree of dysfunction, although there is some suggestion that, on tests of odor discrimination, the right side of the nose may be more influenced than the left.
Psychophysicai Studies of Epilepsy Patients Following Temporal Lobe Resection Unilateral temporal lobe resection for intractable seizure activity invariably infringes upon or ablates anatomic areas associated with olfaction, including the uncus, amygdala, and hippocampus, as well as their connections to other olfaction-related structures. The complexity of the influence of temporal lobe structures on olfactory processing in humans is illustrated by the case of H.M., the only patient on record to have received a bilateral temporal lobe resection (Scoville and Milner, 1957). Despite this loss, H.M. retained partial olfactory functioning with seemingly normal scores on tests of odor detection threshold, suprathreshold intensity discrimination, and adaptation. However, severe deficits on tests of odor identification, discrimination, and matching were apparent (Eichenbaum, Morton, Potter & Corkin, 1983), deficits that have been subsequently observed by others in patients undergoing unilateral lobectomies (e.g., West, Doty, O'Connor & Sperling, 1993;
Laterality and Olfaction 517 Eskenazi, Cain, Novelly & Friend, 1983; Eskenazi, Cain, Novelly & Mattson, 1986; Jones-Gotman & Zatorre, 1988; Martinez et al., 1993; Zatorre & Jones-Gotman, 1992). The effect of temporal lobe resection on measures of olfactory threshold sensitivity is controversial, with some authors reporting decreased sensitivity, others increased sensitivity, and still others no change in sensitivity relative to normals. For example, in a pioneering quantitative study of olfactory function in temporal lobe resection patients, Rausch and Serafetinides (1975) reported bilaterally decreased sensitivity (increased odor detection thresholds) to pyridine, amyl acetate, and phenyl ethyl alcohol following lobectomy; however, odor recognition thresholds were unchanged. Martinez et al. (1993) similarly noted a post-operative increase in odor detection thresholds, but only on the side ipsilateral to the lobe resection and only with right-sided resections. In a study of 16 left- and 14 right-resected patients, West, Doty, O'Connor and Sperling (1993) found decreased detection threshold sensitivity for phenyl ethyl alcohol on the side of the operation in both groups of patients, with more marked deficits on the left than on the fight. However, the degree of dysfunction was greater in the left resection group. In contrast to these findings are those which, like Eichenbaum et al.'s finding in H.M., show no decline in odor detection thresholds following temporal lobe resection (i.e., Huber, Pruszewicz, Szmeja & Bialek, 1%5; Henkin, Comiter, Fedio & O'Doherty, 1977; Eskenazi, et al., 1983, 1986; Jones-Gotman and Zatorre, 1988). An elevation of odor recognition thresholds was reported by Huber et al. (1965) and Henkin et al. (1977) in such patients, although in both cases small numbers of subjects were tested and details as to the manner in which threshold testing was performed were not provided. It should be noted that all of these studies evaluated function bilaterally. Since, as noted earlier, bilateral measurement seems to reflect the best functioning side of the nose (Betchen & Doty, 1997; Hornung et al., 1990), such evaluation is likely insensitive to unilateral deficits such as those observed in the Martinez et al. (1993) and West, Doty, O'Connor and Sperling (1993) studies. There is lack of agreement as to whether the side of temporal lobe resection plays a role in determining the extent of suprathreshold postoperative olfactory deficits in bilaterally-administered tests. No difference between left- and right-sided temporal lobectomies on bilaterallyadministered tests, including tests of odor memory and discrimination,
518
Doty et al.
have been reported (e.g., Eskenazi, Cain; Novelly & Friend, 1983; Eskenazi, Cain, Novelly & Mattson, 1986; Jones-Gotman & Zatorre, 1988; Zatorre & Jones-Gotman, 1991). In contrast, Rausch, Serafetinides and Crandall (1977) found both left- and right-temporal lobe resection groups performed worse than controls on an odor discrimination task, with patients with fight-side resections showing a significantly greater decrement than those with left-side resections. Subsequently, Abraham and Mathai (1983) reported that the ability to match odorants was affected by fight temporal lobe lesions. Left-sided lesions were not associated with any decrement in odor matching ability. Recently, Martinez et al. (1993) and West, Doty, O'Connor and Sperling (1993) examined olfactory functioning in epileptic patients both before and after their operations (the preoperative findings are noted in the previous section). In the Martinez et al. study, postoperative testing occurred about six months after lobectomy, whereas in the West et al. study the post-operative testing was performed immediately after the operation, although a subgroup of patients was also tested six months later. Martinez et al. noted a post-operative elevation in detection threshold on the right side in patients with rightsided operations, and a decrease in threshold (i.e., increase in sensitivity) on the contralateral side. In the West, Doty, O'Connor and Sperling (1993) study, both left and right lobectomized patients evidenced deficits in odor detection threshold and identification in the nostril ipsilateral to the side of operation. Interestingly, in contrast to the Martinez et al. study, the deficit was greater for the left resected patients. No evidence of change in the test measures over the six-month postoperative interval was found in a subgroup of the patients so evaluated. Psychophysical Studies of Patients with Hemiparkinsonism Patients with early-stage hemiparkinsonism tend to have more nigrostriatal dopaminergic damage on the side of the brain opposite to the side of motoric dysfunction, reflecting the crossed pathways of the motor system. To determine whether asymmetry in olfactory function was present in such patients, either in relation to the side of hemiparkinsonism or in relation to left vs. right, Doty, Stern, Pfeiffer, Gollomp and Hurtig (1992) tested the odor identification ability of 20 unmedicated hemiparkinsonian patients, 20 medicated hemiparkinsonian patients, and 20 age- and gender-matched normal controls. No
Laterality and Olfaction 519 meaningful left:right asymmetry or relationship between the side of hemiparkinsonism and scores on the lateralized odor identification test was observed.
Psychophysical Studies of Patients with Cerebral Vascular Accidents Bellas, Novelly and Eskenazi (1989) obtained n-butanol detection threshold measures separately for the left and fight sides of the nose in 15 control subjects and in 15 patients who had suffered a fight cerebral vascular accident. The threshold values did not differ between the two sides of the nose or between the two study groups. To our knowledge, this is the only study of stroke patients in the literature. 2
Psychophysical Studies of Normal (Non-epileptic) Human Subjects Relatively few investigators have studied lateralization of olfactory function in normal human subjects, perhaps because of difficulty in overcoming the intact commissural connections between the hemispheres, as described earlier in this chapter. To our knowledge, the first researchers to quantitatively assess lateralization of olfactory function in humans were Toulouse and Vaschide (1899, 1900). They reported marked asymmetries between the left and right sides of the nose: 56 of 64 largely right-handed subjects (including adult men and women, as well as some children) demonstrated increased sensitivity to camphor odor on the left side of the nose. In 5 left-handed or ambidextrous subjects, the right side was conversely more sensitive. In a subsequent experiment using ammonia with 15 female right-handed subjects, higher sensitivity was found on the right side. Taken together, these observations suggested that the side of greater nasal sensitivity for camphor is contralateral to hand dominance, whereas that for ammonia is ipsilateral to hand dominance. Toulouse and Vaschide speculated that this difference could reflect the fact that camphor is an olfactory stimulant (thereby activating ipsilateral projections), whereas ammonia is a trigeminal stimulant (activating projections that cross the midline). However, it has since been noted that camphor, as well as ammonia, have trigeminal components, at least at higher concentrations, making this explanation less tenable (Doty et al., 1978). Toulouse and Vaschide's remarkable threshold findings have received little support from subsequent studies, although Frye, Doty and
520
Doty et al.
Shaman (1992) found, in 16 right-handed men, 21 right-handed women, 17 left-handed men, and 21 left-handed women, that righthanders were slightly more sensitive than left-handers on the left side of the nose, and left-handers were slightly more sensitive than righthanders on the fight side of the nose. An opposite pattern of results, however, was reported by Youngentob et al. (1982) in a study of 9 lefthanders and 10 right-handers: left-handers consistently evidenced greater n-butanol sensitivity on the left side of the nose, whereas righthanders evidenced a weak tendency towards greater sensitivity on the fight side of the nose. In contrast to the aforementioned studies, a number of experiments, all employing relatively large samples, have found no left:right differences in detection thresholds. For example, Koelega (1979) reported no such differences in n-amyl acetate thresholds for 20 men and 20 women (all right-handed). Zatorre and Jones-Gotman (1990) also found no laterality in a phenyl ethyl alcohol detection threshold task in 49 men and 50 women (approximately half left-handed), a finding that was recently replicated by Betchen and Doty (1997) in a study of 66 men and 72 women (all but 6 fight-handed). Lateralization of function has also been explored for suprathreshold olfactory tasks. In the sole study addressing this issue for odor recognition memory, Bromley and Doty (1995) found no differences in test scores for the left and right sides of the nose. In contrast, a number of studies report a right nostril superiority for other types of tests. Zatorre and Jones-Gotman (1990), for example, reported that performance on a same-different odor discrimination task is better on the right than on the left side of the nose. Pendse (1987), in one of the first studies to report a right-sided performance bias, found that right-handed women show a fight-nostril superiority in estimating the intensity of five suprathreshold concentrations of n-butanol. However, of the 10 right-handed females and 10 right-handed males used in the study, only the females demonstrated such asymmetry, a phenomenon the author was unable to explain. Additional support for the notion that the right hemisphere may play a special role in olfactory information processing comes from a study by Zucco and Tressoldi (1988). Using a reaction time paradigm, these investigators found that when verbal and pictorial stimuli were presented to the left visual field (i.e., the right hemisphere)of 12 fighthanded subjects, they were more quickly recognized as being the same or different from a preceding odor stimulus than when the same stimuli
Laterality and Olfaction
521
were presented to the right visual field. One explanation of this phenomenon is that the initial odor more effectively "primed" the fight hemisphere and facilitated processing of the second stimulus (i.e., a picture or a word). Reasons for right hemispheric effects in odor information processing, if indeed present, remain unclear. It is possible that such effects are associated with the anatomical asymmetries noted earlier in this paper, in that right-sided structures are larger. However, more research is needed on this point, as mass alone need not be a predictor of functionality. Furthermore, a right-side bias may not be unique to olfaction. As reviewed by Christman (1995), several studies have presented identical visual or auditory input to both hemispheres simultaneously and compared the performance to that obtained from the presentation of the same information to each hemisphere alone. Under this condition, one hemisphere typically predominates over the other in bihemispheric trials, a phenomenon termed metacontrol. As noted by Christman (p. 232), Of particular interest is the finding that the hemisphere with the superior processing ability is not necessarily the one that assumes metacontrol; for example, Hellige et al. (1989) employed a consonant-vowel-consonant (CVC) identification task and showed that under conditions of bilateral presentation, the less efficient strategy associated with the fight hemisphere dominated and was applied to both LH
and RH stimuli. Hellige and associates have extended this paradigm to examine the qualitative nature of performance for foveally presented stimuli and have found that the pattern of errors for central presentations also resembled that for RH,but not LH, presentations (Hellige, Cowin & Eng, 1991). This counterintuitive finding may represent a case of interhemispheric cooperation: because the more efficient LH strategy, based on phonetic processing, is not available to the fight hemisphere, both hemispheres may adopt the less efficient RH strategy because it is available to both. An area of significant controversy, which has some bearing on the current discussion, is the possible role of language and verbal labeling on tests of olfactory function. Although several investigators suggest that linguistic processes are involved in the recognition of odors (e.g., Rabin & Cain, 1984; Lyman & McDaniel, 1990), there are numerous studies that show no significant effect of verbal labeling in association with olfaction (Lawless & Cain, 1975; Davis, 1977; Lawless, 1978; Engen, 1987). Understanding that language is typically lateralized to the left hemisphere in right-handed patients, a right side superiority in
522
Doty et al.
olfactory function might be viewed as supporting the hypothesis that specialization of hemispheric function for olfactory processing is not dependent upon language. However, until the nature of both right and left hemisphere "olfactory processing" is better defined, it would seem difficult to test this hypothesis. In addition to studies focusing on possible perceptual asymmetries between the two sides of the nose are those which examine how one side of the nose may influence the other, and how the two sides may be working together to provide a richer bilateral experience of an odor. It was well known in the 19th Century that when one side of the nose was adapted to an odorant, the threshold for the other side was raised, implying central interaction (Zwaardemaker, 1925). Cain (1977)asked subjects to report the bilateral intensity of an odor experience when each side of the nose was presented, at the same time, with different concentrations of a stimulant. He found that with an imbalance in stimulus concentration between the left and the right, the side of the nose receiving the stronger stimulus disproportionately accounts for total perceived intensity of the odor. In order to maintain the same perceived intensity, significantly more stimulation is required of an odorant presented primarily unilaterally than bilaterally. Medina and Cain (1982) subsequently reported a similar phenomenon for trigeminal stimuli, and Bromley and Doty (1995) reported that odor memory scores were higher when stimuli were presented bilaterally than when they were presented unilaterally. In addition, a monotonic decay in performance across delay intervals was seen under unilateral, but not bilateral, test conditions. Whether these instances of apparent interaction between the two sides of the nose simply are manifestations of the tendency for bilateral testing to reflect the best functioning side of the nose merits additional study.
Eiectrophysiological Studies of Normal and Epileptic Patients Several investigators have reported that, in normal subjects, chemosensory evoked potentials are of larger amplitude in the right than in the left hemisphere, regardless of side of nose stimulated (e.g., Becker et al., 1993; Hummel & Kobal, 1992; Kobal, Hummel & Pauli, 1989; Kobal, Hummel & van Toiler, 1992). 3 Despite these indications that more total activation may occur in the right hemisphere, pleasant smelling stimuli (e.g., vanillin)presented to the left nostril induce,
Laterality and Olfaction
523
relative to unpleasant smelling stimuli (e.g., hydrogen sulfide), more hemispheric activation in both hemispheres. Conversely, unpleasant smelling stimuli presented to the right nostril tend to induce greater amplitude responses than pleasant smelling ones. Such observations led Kobal, Hummel and Pauli (1989) to hypothesize that lateralized response differences may relate to specialization of the hemispheric processing of positive (left) and negative (right) emotions (see Davidson, 1984; Dimond, Farrington & Johnson, 1976; Davidson et al., 1990; Jones & Fox, 1992). If this is true, it would be of interest to determine what, in fact, constitutes "pleasant" and "unpleasant" odorants. For example, can an individual be conditioned to perceive an odorant that was initially unpleasant as pleasant and, if so, will the pattern of evoked cortical responses change in accord with the change in perceived pleasantness of the stimulus? Do persons who find a particular odor pleasant (e.g., licorice)provide a different pattern of responses than persons who find the odorant unpleasant? Recently, Hummel et al. (1995) obtained chemosensory eventrelated potentials (CSERPs) for each side of the nose for 12 epilepsy patients with left temporal lobe foci and l0 epilepsy patients with fight temporal lobe foci. Olfactory (vanillin and low concentrations of hydrogen sulfide) and trigeminal (carbon dioxide) stimulants were used. For both patients and controls, the latency of the evoked potential to the trigeminal stimulus was longer when the stimulus was presented to the left naris than to the right naris, regardless of the side of the epileptic focus. However, the latencies for the evoked potentials to the olfactory stimuli were longest on the side of the epileptic focus; i.e., patients with right foci had longer latencies on the right, and patients with left foci had longer latencies on the left. No evidence of overall right- or left-side bias was present in these patients. The situation with evoked potential amplitudes was somewhat different. These investigators found that the topographical distribution of chemosensory evoked potential amplitudes changed as a consequence of the side of epileptogenic focus. Thus, in patients with a right-sided focus, but not those with a left-sided focus, the anteriorposterior amplitude distribution differed from that seen in normal controls. Specifically, in normal controls the amplitudes increase across the Fz, Cz, and Pz axis, but in patients with right-side foci, the Cz amplitude was greater than the Fz or Pz sites, which showed the normal anterior-posterior relationship.
524
Doty et al.
A number of studies examining odor-induced changes in the EEG have been performed in human subjects. While one EEG study reports that the right hemisphere exhibits more pronounced alpha blocking than the left in response to odorants (Schwartz et al., 1992), this phenomenon was not observed in another very thorough and extensive study that involved repetitive testing of subjects on different test days (Brauchli, Regg, Etzweiler & Zeier, 1995). Odorants also have been reported to increase alpha activity at left-hemisphere parietal sites (Sawada et al., 1992), and theta activity at left-hemisphere frontal sites (Klemm, Lutes, Hendrix & Warrenburg, 1992). The apparent discrepancies among studies on the topographical distribution of EEG changes in response to odorous stimulation may be due to a number of factors, including (i) changes in cortical activity across time (Burgess & Gruzelier, 1993), (ii) effects of stimulus exposure time or repetition (Kobal & Hummel, 1991), (iii) specific odorants evaluated (van Toiler et al., 1993), (iv) differences in EEG recording paradigms (e.g., different references), (v) different control conditions (Lorig, 1991), and (vi) the procedures used for odorant presentation (Doty & Kobal, 1995). Functional Imaging Studies There has recently been a rapid expansion of technologies that allow for monitoring, in vivo, human brain activity induced by sensory stimulation, including positron emission tomography (PET), single photon emission tomography (SPECT), magnetoencephalographic source imaging (MSI), and functional magnetic resonance imaging (fMRI). To date, however, there has been comparatively little application of these techniques to olfaction. Such techniques have promise in providing important information regarding differential hemispheric processing of olfactory information, as well as defining regional abnormalities in such processing, although PET and SPECT are limited in terms of temporal resolution, and, thus, cannot distinguish the order or pattern of activation of the brain regions involved. Two studies have used PET to examine cerebral structures activated by olfactory stimulation. In the first, Zatorre, Jones-Gotman, Evans and Meyers (1992) studied 5 male and 6 female healthy college students under control and activation conditions. Under the activation condition, the subjects smelled, bilaterally, one of eight different odors presented
Laterality and Olfaction
525
with a Q-tip, whereas in the control condition, the Q-tip without odor was presented for sniffing. Subtraction images (activation condition minus control condition) were then obtained. Changes in regional cerebral bloodflow (rCBF) were observed bilaterally at the fronto-temporal junction (piriform area), and in the right orbitofrontal cortex. These authors suggested, from these observations, that olfactory asymmetry was present only in higher-order olfactory processing centers. In the second study, Nadel et al. (1992) examined PET responses in four groups of subjects: (a) a non-stimulated control group (n=8); (b) a group receiving left nasal chamber stimulation only (n=3); (c) a group receiving fight nasal chamber stimulation only (n=4); and (d) a group receiving simultaneous stimulation of both nasal chambers (n=5). Following placement of venous and arterial catheters, subjects were injected with 114/aCi/kg of 18F-florodeoxyglucse (FDG). During the 45-minute FDG uptake phase, the patients were blindfolded and their hearing was attenuated using earplugs and/or a foam headholder. Beginning two minutes before FDG injection and throughout the uptake phase, hundreds of presentations of University of Pennsylvania Smell Identification Test (UPSIT) items were made, one after another, which were timed to coincide with each inhalation. In a unilateral stimulation condition, the contralateral naris was occluded using a piece of MicrofoamrM tape (3M Corporation, Minneapolis) cut to conform to the naris perimeter. Odorants presented to the right naris produced more metabolic activity in the right inferior temporal lobe than in the left, whereas odorants presented to the left naris or both nares together did not produce such an asymmetry. More recently, PET was used to assess cortical metabolic rate during an olfactory memory (match to sample) task in six patients with Alzheimer's disease (Buchsbaum, Kesslak, Lynch & Chui, 1991). Metabolic activity was compared with that of both age-matched controls performing the olfactory task and controls resting with their eyes closed. Overall, patients had lower metabolic rates in the anterior portion of the medial temporal cortex than did controls, and the difference was greatest between patients and controls performing the odor memory task. In general, the patients, compared to controls, exhibited larger absolute and relative glucose metabolic rates on the left than on the right side of the brain. Several studies have examined metabolic activity in olfactory-related regions in patients with schizophrenia. Clark, Kopala, Hurwitz and Li
526
Doty et al.
(1991) examined regional cerebral glucose metabolism in 16 male schizophrenia patients and eight healthy controls. Eight of the patients had normal olfactory function and eight were microsmic, as determined by UPSIT scores. The schizophrenic patients had lower levels of overall frontal metabolism than the normal controls. However, the olfactorydeficient schizophrenics had lower right basal ganglia and thalamic metabolism than the non-olfactory deficient schizophrenics, suggesting dysfunction in subcortical brain regions associated with olfaction. While limited in sample size, this study argues for a relative decrement in right hemisphere brain regions and increased activity in contralateral left hemisphere regions in olfactory-deficient schizophrenic patients. Woo et al. (1993) examined odor memory and identification ability in 28 healthy control and 20 neuroleptic-naive schizophrenic subjects, some of whom also underwent concurrent PET scanning. Among the patients with schizophrenia, significant correlations were observed between UPSIT scores and the degree of regional metabolism in the frontal lobes (r = 0.53, p < 0.05), especially in the left middle frontal (r = 0.55~ p < 0.05) and left inferior frontal gyri (r=0.66, p < 0.01). The left frontal/occipital ratio was also positively related (r = 0.66, p < 0.01). Bertollo, Cowen & Levy (1996) examined local cerebral metabolic rate within two olfactory cortical projection regions in eight normal and eight schizophrenic male subjects. A greater degree of hypometabolism was observed in the right lateroposterior quadrant of the orbitofrontal cortex, which receives largely uncrossed projections from the olfactory bulbs via the pyriform and entorhinal cortex, along with a small number of olfactory and other limbic inputs from the medial subdivision of the dorsomedial thalamic nucleus. A smaller but more symmetrical hypometabolism was seen in the medial anterior aspect of the orbitofrontal cortex which receives crossed afferents from the limbic system. While the authors did not obtain psychophysical measures of olfactory function or clinically define olfactory dysfunction in their subjects, these data are consistent with the notion that the olfactory deficit in schizophrenia reflects a rhinencephalic deficit that is more pronounced in the right hemisphere. More recently, Malaspina and colleagues (1996) examined six male schizophrenics and seven age and sex matched controls using SPECT. Resting baseline (matching pictures without delay) and activation (birhinal stimulation with UPSIT items)data were obtained. The authors found a contiguous cluster in the right cortical area with a significantly
Laterality and Olfaction 527 lower rCBF in schizophrenia patients relative to controls. Notably, controls, but not schizophrenics, had significantly increased rCBF in olfactory activation than in resting baseline conditions. Activation was seen specifically in the right hippocampus, right medial temporal/lateral, left occipital and left medial temporal lobes. This study suggests deficient activation of the (tertiary) cortical and medial temporal lobe olfactory areas in patients with schizophrenia. Despite the small number of studies and heterogeneity of methods, the results of neuroimaging studies suggest that there may be a slight fight-hemisphere advantage for the processing of olfactory information. Nevertheless, left hemispheric and subcortical regions also appear to be important in the processing of olfactory information, but seem to be less activated by olfactory stimuli in both patients or controls. It should be noted, however, that most of these studies did not use an activation paradigm, simply correlating psychophysical scores with functional indices. Future studies using olfactory stimuli in activation paradigms will prove important in better defining the degree of relative involvement of left and fight higher-order olfactory structures. CONCLUSIONS It is apparent from this review that there are many contradictory findings and perspectives regarding olfactory system lateralization. In the case of split brain patients, both halves of the brain can discriminate and recognize odors, although only the left side can verbally report their identity. Studies of such patients suggest that higher-order olfactory tasks, such as odor identification and cross-modal matching, are performed better by the left than the right side of the brain, regardless of the nature of the required response or the apparent involvement of linguistic processes. In contrast, a number of studies using non-split brain subjects suggest either no left:fight differences in odor processing or predominantly fight hemisphere superiority. The latter findings fit in conceptually with observations that general regions of the right side of the brain associated with olfactory function (e.g., the temporal lobe) may, in fact, be larger than their left side counterparts. Temporal lobe resection studies that suggest olfactory processing is more dependent upon the right hemisphere should be interpreted cautiously. Traditionally, more cortical tissue is removed in right hemisphere resections than in left hemisphere resections, given the need
528
Doty et al.
to preserve speech areas within the left side of the brain (see, for example, Table 1 in Rausch, Serafetinides & Crandall, 1977). Furthermore, temporal lobe resection studies in which preoperative testing was not performed cannot discern whether the deficits they observe are due to operative procedures or to the underlying epilepsy-related neuropathology. Although pre- and post-operative test data are now available from two studies, the findings conflict in the degree to which right and left lesions influenced the test scores. Ongoing research in our laboratory is seeking to establish more definitively whether the volume of the excised tissue, as measured by MRI, is a probable contributing factor to such disparate left:right findings. Reports of a greater influence of fight than left lesions on olfactory function, along with PET studies suggesting more odor-induced activation of right than left hemisphere structures and psychophysical studies noting better right side than left side performance on suprathreshold tests, makes it tempting to conclude that the right hemisphere is "dominant" for olfaction, just as the left hemisphere is "dominant" for language. However, such a conclusion is premature, given (i) the rather clear-cut data from split brain patients that suggest left hemisphere superiority in a variety of odor-mediated tasks, (ii) the lack of a clear consensus for right side superiority on most psychophysical tasks, and (iii) the possibility that metacontrol or metacontrol-like processes may exist within the olfactory system. Indeed, an argument can be made that if right temporal lobe lesions influence odor perception more than left temporal lobe lesions, but some function is still preserved in both hemispheres, it is the left hemisphere that, in a global sense, is more dominant, given that it is less influenced by lesioning and presumably has more tissue devoted to the maintenance of olfactory function. The interesting observation by Carroll and Richardson (1993) that lesions of the right hemisphere altered performance on an odor memory task for only namable common odors, but not for non-namable uncommon odors, suggests that the hemispheres may work together to coordinate higher order olfactory processing and that left hemisphere activities may be represented within the fight hemisphere. It is clear that much more work is needed to understand olfactory laterality. Unfortunately, numerous methodologic and conceptual pitfalls are evident in existing work in this field, not the least of which is the use of too few subjects, a lack of standardized testing, and ignorance
Laterality and Olfaction
529
concerning the validity and reliability of many the tests employed. Not all studies have carefully counterbalanced the order of testing of the two sides of the nose, resulting in the confounding test order with laterality. In the case of thresholds, it has been well established in both animals and humans that repeated testing results in decreases in threshold values (Doty & Ferguson-Segall, 1989; Pierce, Doty & Amoore, 1996). Thus, if counterbalancing schemes are not adhered to strictly, right-nostril superiority would be found in studies where the left nostril is tested first more often than the right. Furthermore, investigators are cautioned in assuming that the nominal results of olfactory tests truly reflect the functions that are purportedly measured by them. For example, scores on an odor memory test may be depressed when the olfactory epithelium or bulb are damaged, reflecting deficiencies in activation of the system rather than in brain regions associated with odor memory processing, per se. The recent demonstration, using principal components analysis, that, in non-brain damaged or lesioned subjects, tests of odor memory, detection, and identification all tend to load on a common principal component reiterates the need to cautiously interpret test findings in regards to the trait or mechanism being measured (Doty, Smith, McKeown & Raj, 1994). Along this same vein, it is important to note that some portion of the population may have better olfactory function on one side than the other as a result of unilateral damage to the olfactory epithelium and/or bulb, tract, or stria due to local nasal infections, head trauma, substance abuse, neoplasms, vascular dysfunction, or even normal aging (Duncan & Smith, 1995) Although smell losses are typically bilateral, considerable disparity between the two sides of the nose has been found in some cases. For example, Furukawa, Kamide, Miwa and Umeda (1988) reported that seven of 94 patients evaluated had significant unilateral threshold deficits, even though their bilateral thresholds were indicative of normal function. While there is no evidence that such discrepancies favor one side over the other, investigators should be aware of the possibility of the contamination of their study groups with such cases. The degree to which left:right changes in nasal engorgement (e.g., the so-called "nasal cycle") influence measures of unilateral olfactory function has received little attention. It is well documented that left:fight fluctuations occur over time in nasal engorgement and airflow, reflecting alterations in autonomic tone of the nasal vasculature (Frye & Doty, 1992). Such changes correlate with a number of ultradian
530
Doty et al.
rhythms, including asymmetries in left:right EEG activity and differential performance on visual/spatial psychological tasks (Werntz et al., 1983). However, recent data suggest that the classic reciprocity of nasal engorgement across the two sides of the nose is the exception rather than the rule (Gilbert & Rosenwasser, 1987), particularly in older persons. For example, in a study of 60 persons tested at 15-minute intervals over 6-hour periods (Mirza, Kroger & Doty, 1997), only 9% exhibited a classic nasal cycle. One-third of the 60 subjects exhibited no cyclic changes on either side of the nose, whereas 28.3% exhibited parallel cycles (rhythms that were correlated positively, not negatively, across the two sides of the nose) and 23.8% exhibited hemicycles (cyclic changes on only one side of the nose). Such observations suggest that predictable side-to-side changes in olfactory function based upon changes in nasal engorgement are likely the exception, rather than the rule, and that the potential exists for systematic and difficult-to-control biases entering into studies of olfactory system laterality. Interestingly, there is suggestion that olfactory sensitivity may be enhanced bilaterally when relatively more airflow is present in the right nasal chamber, presumably reflecting increased sympathetic nervous system activity (Frye & Doty, 1992). Clearly, a challenge for the future is to determine what precise aspects of olfactory function are lateralized in human beings. To achieve this end, investigators will need to pay careful attention to design details that take into account and correct for limitations in specific tests and test populations. Footnotes
1. Free nerve endings from other nerves within the nasopharynx and oral cavity can also respond to vapors, including the glossopharyngeal (CN IX) and vagus (CN X) nerves. However, most inhaled chemicals do not produce much, if any, stimulation of these nerves and for this reason they are not further considered herein. 2. These authors also reported that the degree to which an odor could be localized to the left or fight nasal chamber was decreased in patients compared to controls, with patients performing at chance level in the left nasal chamber. This finding was interpreted as reflecting left nasalneglect, in accord with the fact that all the patients had left unilateral neglect syndrome. However, the study implies that odors can be localized to one or the other nasal chamber. This differs from the observations of others that odorants cannot be so localized unless they have a trigeminal component (e.g., von Skramlik, 1926; Schneider & Schmidt, 1967;
Laterality and Olfaction 531 Ehrlichman, 1986; Kobal, van Toiler & Hummel, 1989). Several of the stimuli used by Bellas et al. (e.g., ground coffee beans, shaved Ivory soap) would seem to the present authors to have at least some degree of trigeminal activity, making their conclusions questionable. 3. An apparent exception to this rule is nicotine. Hummel et al. (1992) report that this stimulus produces larger amplitude evoked responses in the left hemisphere than in the fight. Only left nostril stimulation was performed, so the generality of this phenomenon to the fight nostril is unknown.
Acknowledgements Supported by Grant PO 1 00161 from the National Institute on Deafness and Other Communication Disorders, National Institutes of Health. We thank Hans Kroger, Igor Kratskin, Ritu Bagla, and David Yousem for their comments on a previous version of the manuscript.
References Abraham, A. & Mathai, K.V. (1983). The effect of fight temporal lobe lesions on matching of smells. Neuropsychologia, 21, 277-281. Andy, O.J., Jurko, M.F, & Hughes, J.R. (1975). The amygdala in relation to olfaction. Confin. Neurol., 3 7, 215-222. Anton, F. & Peppel, P. (1991). Central projections of trigeminal primary afferents innervating the nasal mucosa" a horseradish peroxidase study in the rat. Neuroscience, 41, 617-628. Barnett, E.M., Evans, G.D., Sun, N., Perlman, S. & Cassell, M.D. (1995). Anterograde tracing of trigeminal afferent pathways from the murine tooth pulp to cortex using herpes simplex virus type I. J. Neuroscience, 15, 2972-2984. Bear, D., Schiff, D., Saver, J., Greenberg, M. & Freeman, R. (1986). Quantitative analysis of cerebral asymmetry. Arch. Neurol., 43, 598' 603. Becker, E., Hummel, T., Piel, E., Pauli, E., Kobal, G. & Hautzinger, M. (1993). Olfactory event-related potentials in psychosis-prone subjects. Internat. J. Psychol., 15, 51-58. Bellas, D.N., Novelly, R.A. & Eskenazi, B. (1989). Olfactory lateralization and identifica tion in right hemisphere lesion and control patients, Neuropsychologia, 27, 1187-1191. Bernard, J.F., Peschanski, M. & Besson, J.M. (1989). A possible spino (trigemino)-ponto-amygdaloid pathway for pain. Neurosci. Letters, 100, 83-88 ~ Bertollo, D.N., Cowen, M . A . & Levy, A.V. (1996). Hypometabolism in olfactory cortical projection areas of male patients with
532
Doty et al.
schizophrenia: an initial positron emission tomography study.
Psychiat. Res., 60, 113-116. Betchen, S.A. & Doty, R.L. (1997). Bilateral olfactory detection thresholds reflect the functioning of the better side of the nose. Chem. Senses, in press. Bevan, S., Forbes, C.A. & Winter, J. (1993). Protons and capsaicin activate the same ion channels in rat isolated ganglion neurones. J. Physiol., 459, 401P. Bevan, S. & Yeats, J. (1991). Protons activate a cation conductance in a subpopulation of rat dorsal root ganglion neurones. J. Physiol., 433, 145-161. Buchsbaum, M.S., Kesslak, J.P., Lynch, G. & Chui, H. (1991). Temporal and hippocampal metabolic rate during an olfactory memory task assessed by positron emission tomogra phy in patients with dementia of the Alzheimer type and controls: Preliminary studies. Arch. Gen. Psychiat., 48, 840-847. Bouvet, J.F., Delaleu, J.C. & Holley, A. (1987). Olfactory receptor cell function is affected by trigeminal nerve activity. Neurosci. Letters, 77, 181-186. Bouvet, J.F., Delaleu, J.C. & Holley, A. (1988). The activity of olfactory receptor cells is affected by acetylcholine and substance P. Neurosci. Res., 5, 214-223. Brauchli, P., Regg, P.B., Etzweiler, F. & Zeier, H. (1995). Electrocortical and autonomic alteration by administration of a pleasant and an unpleasant odor. Chem. Senses, 20, 505- 51:5. Brodal, A. (1947). The hippocampus and the sense of smell. A review. Brain, 70, 179-222. Brodal, A. (1947). The amygdaloid nucleus in the rat. J. Comp. Neurol., 87, 1-16. Brodal, A. (1948). The origin of the fibers of the anterior commissure in the rat. Experimental studies. J. Comp. Neurol., 88, 157-205. Brodal A. (1981). Neurological Anatomy in Relation to Clinical Medicine. New York: Oxford University Press. Bromley, S.M. & Doty, R.L. (1995). Odor recognition memory is better under bilateral than unilateral test conditions. Cortex, 31, 25-40. Buchsbaum, M.S., Kesslak, J.P., Lynch, G. & Chui, H. (1991). Temporal and hippocampal metabolic rate during an olfactory memory task assessed by positron emission tomogra phy in patients with dementia of the Alzheimer type and controls: Preliminary studies. Arch. Gen. Psychiat., 48, 840-847. Burgess, A. & Gruzelier, J. (1993). Individual reliability of amplitude distribution in topographical mapping of EEG. Electroenceph. Clin. Neurophysiol., 86, 219-223. Cain, D.P. (1974). The role of the olfactory bulb in limbic mechanisms.
Psychol. Bull., 81,654-671. Cain, W.S. (1977). Bilateral interaction in olfaction. Nature, 268, :50-53.
Laterality and Olfaction 533 Campanella, G., Filla, A. & DeMichele, G. (1978). Smell and taste acuity in epileptic syndromes. Eur. Neurol., 17, 136-141. Carpenter, M.B. (1985). Core Textbook of Neuroanatomy, 3rd ed. Baltimore: Williams and Wilkins. Carroll, B. & Richardson, J.T.E. (1993). Olfactory information processing and temporal lobe epilepsy. Brain Cogn., 22, 230-243. Carson, K.A. (1984). Quantitative localization of neurons projecting to the mouse main olfactory bulb. Brain Res. Bull., 12, 629-634. Christman, S.D. (1995). Independence versus integration of right and left hemisphere processing: Effects of handedness. In F.L. Kitterle (Ed.), Hemispheric Communication: Mechanisms and Models. Hillsdale, N.J.: Lawrence Erlbaum Associates, pp. 231-253. Clark, C., Kopala, L., Hurwitz, T. & Li, D. (1991). Regional metabolism in microsmic patients with schizophrenia. Can. J. Psychiat., 36, 645650. Davidson, J.D. (1984). Affect, cognition, and hemispheric specialization. In Izard, C., Kagan, J. & Zajonc, R. (eds.), Emotions, Cognition, and Behavior. Cambridge" Cambridge University Press, pp. 320-365. Davidson, R.J., Ekman, P., Saron, C.D., Senulis, J. & Friesen, W. (1990). Approach-withdrawal and cerebral asymmetry: Emotional expression and brain physiology I. J. Pers. Soc. Psychol.,58, 330-341. Davis, B.J. & Macrides, F. (1983). Tyrosine hydroxylase immunoreactive neurons and fibers in the olfactory system of the hamster. J. Comp. Neurol., 214, 427-440. Davis, R.G. (1977). Acquisition and retention of verbal associations to olfactory and abstract stimuli of varying similarity. J. Exp. Psychol." Human Learn. Mem., 3, 37-51. DeMichele, G., Filla, A. & Campanelli, G. (1976). Ulteriori dati sull'acuita olfattiva negli epilettici. Acta Neurol. Napoli, 31, 250-256. DeOlmos, J., Hardy, H. & Heimer, L. (1978). The afferent connections of the main and accessory olfactory bulb formations in the rat: an experimental HRP-study. J. Comp. Neurol., 181,213-244. Diamond, M.C., Murphy Jr., G.M., Akiyama, K. & Johnson, R.E. (1982). Morphologic hippocampal asymmetry in male and female rats. Exper. Neurol., 76, 553-565. Dimond, S.J., Farrington, L. & Johnson, P. (1976). Differing emotional response from right and left hemispheres. Nature, 261,690-692. Dimov, D. (1973). The condition of olfaction in epilepsy.VestnOtorhinolaryngol, 35, 22-23. Doty, R.L. (1997). Cranial Nerve 1" Olfaction. I n C.G. Goetz & E.J. Pappert (Eds.), Textbook of Clinical Neurology. Philadelphia: W.B. Saunders, in press. Doty, R.L., Brugger, W.E., Jurs, P.C., Orndorff, M.A., Snyder, P.J. & Lowry, L.D. (1978). Intranasal trigeminal stimulation from odorous volatiles: Psychometric responses from anosmic and normal humans. Physiol. Behav., 2 0, 175-185.
534
Doty et al.
Doty, R.L. & Ferguson-Segall, M. (1989). Influence of castration on the odor detection performance of rats. Behav. Neurosci., 103, 691-693. Doty, R.L. & Kobal, G. (1995). Current trends in the measurement of olfactory function. In Doty, R.L. (ed.), Handbook of Olfaction and Gustation. New York: Marcel Dekker, pp. 191-225. Doty, R.L., Li, C., Mannon, L. & Yousem, D.G. (1997). Olfactory dysfunction in multiple sclerosis: Correlation with plaque load in inferior frontal and temporal lobes. New Eng. J. Med., in press. Doty, R.L., Shaman, P. & Dann, M. (1984). Development of the University of Pennsylvania Smell Identification Test: A standardized microencapsulated test of olfactory function. Physiol. Behav. (Monograph), 3 2, 489-502. Doty, R.L., Shaman, P., Applebaum, S.L., Giberson, R., Sikorsky, L. & Rosenberg, L. (1984). Smell identification ability: Changes with age. Science, 226, 1441-1443. Doty, R.L., Smith, R., McKeown, D. & Raj, J. (1994). Tests of human olfactory function: Principal components analysis suggests that most measure a common source of variance. Percept. Psychophys., 56, 701-707. Doty, R.L., Stern, M.B., Pfeiffer, C., Gollomp, S.M. et al. (1992). Bilateral olfactory dysfunction in early stage treated and untreated idiopathic parkinson's disease. J. Neurol. Neurosurg. Psychiat., 55, 138-142. Duncan, H.J. & Smith, D.V. (1995). Clinical disorders of olfaction. In: R.L. Doty (Ed.), Handbook of Olfaction and Gustation. N.Y.: Marcel Dekker, pp. 345-365. Ehrlichman, H. (1986). Hemispheric Asymmetry and Positive-Negative Affect. Paper presented at the Wenner-Gren Center International Symposium on The Dual Brain" Unified Functioning and Specialization of the Hemispheres, Stockholm, Sweden. Eichenbaum, H., Morton, T.H., Potter, H. & Corkin S. (1983). Selective olfactory deficits in case H.M. Brain, 106, 459-472. Eichenbaum, H., Shedlack, K.J. & Ekmann, K. (1980). Thalamocortical mechanisms in olfaction. I. Effects of lesion of the mediodorsal thalamic nucleus and frontal cortex on olfactory discrimination in the rat. Brain Behav. Evol., 17, 255-275. Engen, T. (1987). Remembering odors and their names. Amer. Sci., 75, 497-503. Eskenazi, B., Cain, W.S., Lipsitt, E.D. & Novelly, R.A. (1988). Olfactory functioning and callosotomy: A report of two cases. Yale J. Biol. Med., 61,447-456. Eskenazi, B., Cain, W.S., Novelly, R.A., & Friend, K.B. (1983). Olfactory functioning in temporal lobectomy patients. Neuropsychologia, 21, 365-374.
Laterality and Olfaction 535 Eskenazi, B., Cain, W.S., Novelly, R.A. & Mattson, R. (1986). Odor perception in temporal lobe epilepsy patients with and without temporal lobectomy. Neuropsychologia, 24, 553-562. Finger, T.E., Getchell, M.L., Getchell, T.V. & Kinnamon, J.C. (1990). Affector and effector functions of peptidergic innervation of the nasal cavity. In" B.G. Green, J.R. Mason & M.R. Kate (Eds.), Chemical Senses: Irritation, N.Y.: Marcel Dekker, pp. 1-20. Frye, R.E. & Doty, R.L. (1992). The influence of ultradian autonomic rhythms, as indexed by the nasal cycle, on unilateral olfactory thresholds. In R.L. Doty & D. MUller-Schwartze, D. (Eds.), Chemical Signals in Vertebrates VI. New York: Plenum Press, pp. 595-598. Frye, R.E., Doty, R.L. & Shaman, P. (1992). Bilateral and unilateral olfactory sensitivity: Relationship to handedness and gender. IIn R.L. Doty & D. Mtiller-Schwartze, D. (Eds.), Chemical Signals in Vertebrates VL New York: Plenum Press, pp. 559-564. Furukawa, M., Kamide, M., Miwa, T. & Umeda, R. (1988). Importance of unilateral examination in olfactometry. Auris Nasus Larynx (Tokyo), 15, 113-116. Galaburda, A.M., LeMay, M., Kemper, T.L. & Geschwind, N. (1978). Right-left asymmetries in the brain. Science, 199, 852-856. Gazzaniga, M.S., Risse, G.L., Springer, S.P., Clark, A.B. & Wilson, D.H. (1975). Psychologic and neurologic consequences of partial and complete commissurotomy. Neurology, 25, 10-15. Geschwind, N. & Levitsky, W. (1968). Human brain: Left-right asymmetry in temporal speech region. Science, 161, 186-187. Gesteland, R.C. (1986). Speculations on receptor cells as analyzers and filters. Experientia, 4 2, 287-291. Gilbert, A.N. & Rosenwasser, A.M. (1987). Biological rhythmicity of nasal airway patency: A re-examination of the 'nasal cycle.' Acta Otolaryngol. (Stockh), 104, 180-186. Gordon, H.W. (1974). Olfaction and cerebral separation. In: M. Kinsbourne & W.L. Smith (Eds), Hemispheric Disconnection and Cerebral Function. Springfield: Charles C. Thomas, pp. 137-154. Gordon, H.W. & Sperry, R.W. (1969). Lateralization of olfactory perception in the surgically separated hemispheres of man. Neuropsychologia, 7, 11 l-120. Gordon, H.W., Bogen, J.E. & Sperry, R.W. (1971). Absence of disconnection syndrome in two patients with partial section of the neocommissures. Brain, 94, 327-336. Hagbarth, K.E. (1960). Centrifugal mechanisms and sensory control. Ergebn. Biol., 22, 47-66. Haberly, L.B. & Price, J.L. (1978). Association and commissural fiber systems of the olfactory cortex of the rat. I. Systems originating in the piriform cortex and adjacent areas. J. Comp. Neurol., 178, 711740.
536
Doty et al.
Heine, O. & Galaburda, A.M. (1986). Olfactory asymmetry in the rat brain. Exper. Neurol.,91, 392-398. Hellige, J., Cowin, E. & Eng, T. (1991). Processing of CVC syllables from LVF, RVF, and central locations. Paper presented at the 19th Annual Meeting of the International Neuropsychological Society, San Antonio, TX. Hellige, J., Taylor, A. & Eng, T. (1989). Interhemispheric interaction when both hemispheres have access to the same stimulus information. J. Exp. Psychol." Human Percept. Perform., 15, 711-722. Henkin, R.I., Comiter, H., Fedio, P. & O'Doherty, D. (1977). Defects in taste and smell recognition following temporal lobectomy. Trans. Amer. Neurol. Assoc., 102, 146-150. Hornung, D.E., Leopold, D.A., Mozell, M.M., Sheehe, P. & Youngentob, S.L. (1990). Impact of left and right nostril olfactory abilities on binasal olfactory performance. Chem. Senses, 15, 233-237. Huber, Z., Pruszewicz, A., Szmeja, Z. & Bialek, E. (1965). Study of smell, taste, heating, balance, sight, and tactile sensation following excision of the anterior temporal lobe. Pol. Neurol. Neurosurg. Psychiat., 15, 475-480. Hummel, T. & Kobal, G. (1992). Differences in human evoked potentials related to olfactory or trigeminal chemosensory activation. Electroenceph. Clin. Neurophysiol., 84, 84-89. Hummel, T. & Kobal, G. (1994). Chemosensory event-related potentials: Effects of dichotomous stimulation with eugenol and dipyridyl. In Kurihara, K., Suzuki, N. & Ogawa, H. (eds.), Olfaction and Taste XI. Berlin" Springer, pp. 659-663. Hummel, T., Livermore, A., Hummel, C. & Kobal, G. (1992). Chemosensory event-related potentials" Relation to olfactory and painful sensations elicited by nicotine. Electroenceph. Clin. Neurophysiol., 84, 192-195. Hummel, T., Pauli, E., Stefan, H., Kettenmann, B., Schler, P. & Kobal, G. (1995). Chemosensory event-related potentials in patients with temporal lobe epilepsy. Epilepsia, 36, 79-85. Hummel, T., Schiessl, C., Wendler, J. & Kobal, G. (1996). Peripheral electrophysiological responses decrease in response to repetitive painful stimulation of the human nasal mucosa. Neurosci. Letters, 212, 37-40. Huttunen, J., Kobal, G., Kaukoronta, E. & Hari, R. (1986). Cortical responses to painful CO2-stimulation of nasal mucosa" a magnetencephalographic study in man. Electroenceph. Clin. Neurophysiol., 64, 347-349. Inokuchi, A., Kimmelmann, C.P. & Snow, J.B. (1993). Convergence of olfactory and nasotrigeminal inputs on possible trigeminal contributions to olfactory responses in rat thalamus. Eur. Arch. Otorhinolaryngol., 249, 473-477.
Laterality and Olfaction 537 Jones, F.N. (1953). A test of the Elsberg technique of olfactometry. Amer. J. Psychol., 66, 81-85. Jones, N.A. & Fox, N.A. (1992). Electroencephalogram asymmetry during emotionally evocative films and its relation to positive and negative affectivity. Brain Cogn., 20, 280- 299. Jones-Gotman, M. & Zatorre, R.J. (1988). Olfactory identification deficits in patients with focal cerebral excision. Neuropsychologia, 26: 387-400. Jones-Gotman M, & Zatorre RJ. (1993). Odor recognition memory in humans: Role of right temporal and orbitofrontal regions. Brain Cogn., 22, 182-198. Kelly, J.P. & Dodd, J. (1991). Trigeminal system. In Kandel, E.R., Schwartz, J.H. & Jessell, T.M. (eds.), Principles of Neural Science. New York: Elsevier, pp. 701-710. Kertesz, A., Polk, M., Black, S.E. & Howell, J.A. (1990). Sex, handedness and the morphometry of cerebral asymmetries on magnetic resonance imaging. Brain Res., 530, 40-48. Kettenmann, B., Hummel, C., Stefan, H. & Kobal, G. (1996). Magnetoencephalographical recordings: separation of cortical responses to different chemical stimulation in man. Funct. Neurosci. (EEG Suppl.), 46, 287-290. King, C. & Hall, W.G. (1990). Developmental change in unilateral olfactory habituation is mediated by anterior commissure maturation. Behav. Neurosci., 104: 796-807. Klemm, W.R., Lutes, S.D., Hendrix, D.V. & Warrenburg, S. (1992). Topographical EEG maps of human responses to odors. Chem. Senses, 17: 347-361. Kobal, G. (1985). Pain-related electrical potentials of the human nasal mucosa elicited by chemical stimulation. Pain, 22, 151-163. Kobal, G. & Hummel, T. (1991). Olfactory evoked potentials in humans. In T.V. Getchell, R.L. Doty, L.M. Bartoshuk, & J.B. Snow, Jr. (Eds.), Smell and Taste in Health and Disease. New York: Raven Press, pp. 255-275. Kobal, G., Hummel, T. & Pauli, E. (1989). Correlates of hedonic estimates in the olfactory evoked potential. Chem. Senses, 14, 718. Kobal, G., Hummel, T. & van Toiler, S. (1992). Differences in chemosensory evoked potentials to olfactory and somatosensory chemical stimuli presented to left and right nostrils. Chem. Senses, 17, 233-244. Kobal, G., van Toiler, S. & Hummel, T. (1989). Is there directional smelling? Experientia, 45:130-131. Koelega, H.S. (1979). Olfaction and sensory asymmetry. Chem. Senses Flay.,4: 89-95. Kolb, B., Sutherland, R.J., Nonneman, A.J. & Whishaw, I.Q. (1982). Asymmetry in the cerebral hemispheres of the rat, mouse, rabbit and cat: The fight hemisphere is larger. Exper. Neurol., 78: 348-359.
538
Doty et al.
Konnerth, A., Lux, H.D. & Morad, M. (1987). Proton-induced transformation of calcium channels in chick dorsal root ganglion cells. J. Physiol., 386, 603-633. Kratskin, I.L. (1995). Functional anatomy, central connections, and neurochemistry of the mammalian olfactory bulb. In: R.L. Doty (Ed.), Handbook of Olfaction and Gustation, N.Y.: Marcel Dekker, pp. 103-126. Kucharski, D. & Hall, W.G. (1987). New routes to early memories. Science, 238: 786-788. Kucharski, D. & Hall, W.G. (1988). Developmental change in the access to olfactory memories. Behav. Neurosci., 102: 340-348. Kucharski, D., Arnold, H.M. & Hall, W.G. (1995). Unilateral conditioning of an odor aversion in 6-day-old rat pups. Behav. Neurosci., 109: 563-566. Lamantia, A.S. & Rakic, P. (1990). Cytological and quantitative characteristics of four cerebral commissures in the rhesus monkey. J. Comp. Neurol., 291:520-537. Lang, J. (1989). Clinical Anatomy of the Nose, Nasal Cavity and Paranasal Sinuses. Stuttgart: Thieme V erlag. Lawless, H.T. (1978). Recognition of common odors, pictures, and simple shapes. Percept. Psychophys., 24: 493-495. Lawless, H.T. & Cain, W.S. (1975). Recognition memory for odors. Chem. Senses Flay., 18: 17-26. LeMay, M. (1976). Morphological cerebral asymmetries of modern man, fossil man and nonhuman primate. Ann. N.Y. Acad. Sci., 280: 349-366. LeMay, M. (1984). Radiological, developmental and fossil asymmetries. In N. Geschwind & A. M. Galaburda (Eds.), Cerebral Dominance: The Biological Foundations. Cambridge, Mass: Harvard University Press, pp. 26-42. Lorig, T.S. (1991). Cognitive and non-cognitive effects on odour exposure: electrophysiological and behavioral evidence. In S. Van Toller & G.H. Dodd (Eds.), Fragrance - The Psychology and Biology of Perfume. London: Elsevier, pp. 16 l- 173. Lyman, B.J. & McDaniel, M.A. (1990). Memory for odors and names: Modalities of elaboration and imagery. J. Exp. Psychol.: Learn. Mem. Cogn., 16: 656-664. MacLeod, P. (1971). Structure and function of higher olfactory centers. In L. Beidler (Ed.), Handbook o.f Sensory Physiology, Vol. 14: Olfaction. N.Y.: Springer-Verlag, pp. 181-204. Malaspina, D., Lignelli, A., Pererra, G., Marshall, R., Gorman, J. & van Heertum, R. (1996). SPECT brain activation by odor discrimination in schizophrenia and controls. Presented at 35th Annual Meeting of the Academy of Neuropsychopharmacology. San Juan, Puerto Rico.
Laterality and Olfaction 539 Martinez, B.A., Cain, W.S., de Wijk, R.A., Spencer, D.D., Novelly, R.A. & Saas, K.J. (1993). Olfactory functioning before and after temporal lobe resection for intractable seizures. Neuropsychology, 7: 351-363. McLean, J.H. & Shipley, M.T. (1987). Serotonergic afferents to the rat olfactory bulb. I. Origins and laminar specificity of serotonergic inputs in the adult rat. J. Neurosci., 7:3016-3028. Medina, M.R.G. & Cain, W.S. (1982). Bilateral integration in the common chemical sense. Physiol. Behav., 29: 349-353. Meyer, M. & Allison, A.C. (1949). An experimental investigation of the connexions of the olfactory tracts in the monkey. J. Neurol. Neurosurg. Psychiat., 12: 274-286. Mesulam, M.-M. (1984). A cortical network for directed attentional and unilateral neglect. In: A. Ardila & F. Ostrosky-Solis (Eds.), The Right Hemisphere: Neurology and Neuropsychology. New York: Gordon and Breach, pp. 61-96. Milner, B. (1968). Visual recognition and recall after right temporallobe excision in man. Neuropsychologia, 6, 191-209. Mirza, N., Kroger, H. & Doty, R.L. (1997). Influence of age on the 'nasal cycle.' Laryngoscope, 107: 62-66. Moran, D.T., Rowley, J.C. III, Jafek, B.W. & Lovell, M.A. (1982). The fine structure of the olfactory mucosa in man. J. Neurocytol., 11" 721-746. Nadal, D.M., Doty, R.L., Souder, J.E., Alavi, A., Crawford, D.Z., et al. (1992). Postitron-emission tomography suggests right temporal lobe dominance in olfactory processing. Unpublished manuscript. Pendse, S.G. (1987). Hemispheric asymmetry in olfaction on a category judgement task. Percept. Mot.Skills, 64: 495-498. Pierce, J.D., Jr., Doty, R.L., & Amoore, J.E. (1996). Analysis of position of trial sequence and type of diluent on the detection threshold for phenyl ethyl alcohol using a single staircase method. Percept. Mot. Skills, 8 2:451-458. Potter, H. & Butters, N. (1980). An assessment of olfactory deficits in patients with damage to prefrontal cortex. Neuropsychologia, 18: 621-628. Powell, T.P.S. & Cowan, W.M. (1963). Centrifugal fibers in the lateral olfactory tract. Nature, 199:1296-1297. Powell, T.P.S., Cowan, W.M. and Raisman, G. (1965). The central olfactory connections. J. Anat., 99: 791-813. Price, J.L. (1990). Olfactory system. In G. Paxinos (Ed.), The Human Nervous System. San Diego: Academic Press, pp. 979-998 Price, J.L. & Powell, T.P.S. (1970). The afferent connexions of the nucleus of the horizontal limb of the diagonal band. J. Anat., 107: 239-256. Rabin, M.D. & Cain, W.S. (1984). Odor recognition: Familiarity, identifiability, and encoding consistency. J. Exp. Psychol." Learn. Mem. Cogn., 10:316-325.
540
Doty et al.
Rausch, R. & Serafetinides, E.A. (1975). Specific alteration of olfactory function in humans with temporal lobe lesions. Nature, 225: 557558. Rausch, R., Serafetinides, E.A. & Crandall, P.H. (1977). Olfactory memory in patients with anterior temporal lobectomy. Cortex, 13: 445-452. Read, E. (1908). A contribution to the knowledge of the olfactory apparatus in dog, cat and man. Am. J. Anat., 8, 17-47. Risse, G.L., De Doux, J., Springer, S.P., Wilson, D.H. & Gazzaniga, M.S. (1978). The anterior commissure in man: Functional variation in a multisensory system. Neuropsychologia, 16:23-31. Rosen, G.D., Sherman, G.F., Mehler, C., Emsbo, K. & Galaburda, A.M. (1989). The effect of developmental neuropathology on neocortical asymmetry in New Zealand black mice. Internat. J. Neurosci., 45: 247-254. Santorelli, G. & Marotta, A. (1964). La sogli olfattometrica dell'epilettico in condizioni di base e dopo crisi. Rivista OspPsichiatr., 3 2:185-190. Sawada, K., Koyama, E., Kubota, M., Hayashi, I., Komaki, R., Inui, M. & Torii, S. (1992). Effects of odors on EEG relaxation and alpha power. Chem. Senses, 17: 88. Schneider, R.A. & Schmidt, C.E. (1967). Dependency of olfactory localization on non-olfactory cues. Physiol. Behav., 2: 305-309. Schwartz, G.E., Wright, K.P., Polak, E.H., Kline, J.P. & Dikman, Z. (1992). Topographic EEG mapping of conscious and unconscious odors. Chem. Senses, 17: 695. Scoville, W.B. & Milner, B. (1957). Loss of recent memory after bilateral hippocampal lesions. J. Neurol. Neurosurg. Psychiat., 20:11-21. Shepherd, G.M. (1972). Synaptic organization of the mammalian olfactory bulb. Physiol. Rev., 52: 864-917. Shepherd, G.M., Nowycky, M.C., Greer, C.A. & Mori, K. (1981). Multiple overlapping circuits within olfactory and basal forebrain systems. Adv. Physiol. Sci., 30: 263-278. Sherman, G.F. & Galaburda, A.M. (1985). Asymmetries in anatomy and pathology in the rodent brain. In S.D. Glick (Ed.), Cerebral Lateralization in Nonhuman Species. N.Y.: Academic Press, pp. 185-231. Shipley, M.T. & Adamek, J.P. (1984). The connections of the mouse olfactory bulb: a study using orthograde and retrograde transport of wheat germ agglutinin conjugated to horseradish peroxidase. Brain Res. Bull., 12: 669-687. Shipley, M.T. & Ennis, M. (1996). Functional organization of olfactory system. J. Neurobiol., 3 0:123-176. Shipley, M.T., Halloran, F.J. & de la Torre, J. (1985). Surprisingly rich projection from locus coeruleus to the olfactory bulb in the rat. Brain Res., 329: 294-299.
Laterality and Olfaction 541 Shipley, M & Reyes, P. (1991). Anatomy of the human olfactory bulb and central pathways. In: D.G. Laing, R.L. Doty & Breipohl, W., Eds., The Human Sense of Smell. Berlin: Springer-Verlag, pp. 29-60. Silver, W.L. & Moulton, D.G. (1982). Chemosensitivity of rat nasal trigeminal receptors. Physiol. Behav., 28:927-931. Slotnick, B. & Kaneko, N. (1981). Role of the mediodorsal thalamic nucleus in olfactory discrimination learning in rats. Science, 214: 9192. Smith, C.G. (1942). Age incidence of atrophy of olfactory nerves in man. J. Comp. Neurol., 77: 589-595. Snow, P.J., Lumb, B.M. & Cervero, F. (1992). The representation of prolonged and intense, noxious somatic and visceral stimuli in the ventrolateral orbital cortex of the cat. Pain, 48: 89-99. Sprague, J.M., Chambers, W.W. & Stellar, E. (1961). Attentive, affective, and adaptive behavior in the cat. Science, 133: 165-173. Squire, L.R. & Zola-Morgan, S. (1991). The medial temporal lobe memory system. Science, 253: 1380-1386. Staubli, U., Ivy, G. & Lynch, G. (1984). Hippocampal denervation causes rapid forgetting of olfactory information in rats. Proc. Natl. Acad. Sci. U.S.A., 81: 5885-5887. Staubli, U., Schottler, F. & Nejat-Bina, D. (1987). Role of dorsomedial thalamic nucleus and piriform cortex in processing olfactory information. Behav. Brain Res., 25:117-129. Steen, K.H. & Reeh, P.W. (1993). Sustained graded pain and hyperalgesia from harmless experimental tissue acidosis in human skin. Neurosci. Letters, 154:113-116. Steen, K.H., Reeh, P.W., Anton, F. & Handwerker, H.O. (1992). Protons selectively induce lasting excitation and sensitization of nociceptors in rat skin. J. Neurosci., 12: 86- 95. Steen, K.H., Wegner, H., Kreysel, H. & Reeh, P. (1995). The pH-release of rat cutaneous nociceptors correlates with extracellular [Na+] and is increased under amiloride, in vitro. Soc. Neurosci., 21: 648. Stewart, W.B., Kauer, J.S. & Shepherd, G.M. (1979). Functional organization of rat olfactory bulb analyzed by the 2-deoxyglucose method. J. Comp. Neurol.,185: 715-734. Stone, H. (1969). Effect of ethmoidal nerve stimulation on olfactory bulbar electrical activity. In C. Pfaffmann (Ed.), Olfaction and Taste. New York: Rockefeller University Press, pp. 216-220. Stone, H., Williams, B. & Carregal, E.J.A. (1968). The role of the trigeminal nerve in olfaction. Exp. Neurol.,21:11-19. Tan, U., & Caliskan, S. (1987). Asymmetries in the cerebral dimensions and fissures of the dog. Internat. J. Neurosci., 32: 943-952. Tanabe, T., Yarita, H., Iino, M., Ooshima, Y, & Takagi, S.F. (1975). An olfactory projection area in orbitofrontal cortex of the monkey. J. Neurophysiol., 3 8:1269-1283.
542
Doty et al.
Toulouse, E. & Vaschide, N. (1899). L'Asymetrie sensorielle olfactive. Comp. Rend. Soc. Biol., 51: 785-787. Toulouse, E. & Vaschide, N. (1900). L'Asymetrie sensorielle olfactive. Revue Philosophique, 4 9:176-186. Van Toiler, S., Behan, J., Howells, P., Kendall-Reed, M. & Richardson, A. (1993). An analysis of spontaneous human cortical EEG activity to odours. Chem. Senses, 18, 1-16. von Bekesy, G. (1964). Olfactory analog to directional hearing. J. Appl. Physiol., 19: 369- 373. von Brunn, A. (1892). Beitrge zur mikroskopischen Anatomie der menschlichen Nasenhole. Arch. Mikr. Anat., 39:632-651. von Skramlik, E. (1925). Uber die Lokalisation der Empfindungen bei den niederen Sinnen. Z. Sinnesphysiol., 56: 69-140. Weinberger, D.R., Luchins, D.J., Morihisa, J. & Wyatt, R.J. (1982). Asymmetrical volumes of the right and left frontal and occipital regions of the human brain. Ann. Neurol., 11: 97-100. Werntz, D.A., Bickford, R.G., Bloom, F.E., & Shannahoff-Khalsa, D.S. (1983). Alternating cerebral hemispheric activity and the lateralization of autonomic nervous function. Hum. Neurobiol., 2: 39-43. West, S.E., Doty, R.L., O'Connor, M.J. and Sperling, M.A. (1993). Preand post-operative studies of olfactory function in patients with anterior temporal lobectomy. Chem. Senses, 18: 649. Wu, J., Buchsbaum, M.S., Moy, K., Denlea, N., Kesslak, P., Tseng, H., Plosnaj, D., Hetu, M., Potkin, S. et al. (1993). Olfactory memory in unmedicated schizophrenics. Schizophrenia Res., 9: 41-47. Youngentob, S.L., Kurtz, D.B., Leopold, D.A., Mozell, M.M. & Hornung, D.E. (1982). Olfactory sensitivity: Is there laterality? Chem. Senses, 7:11-21. Zatorre, R.J. & Jones-Gotman, M. (1990). Right-nostril advantage for discrimination of odors. Percept. Psychophys., 4 7:526-531. Zatorre, R.J. & Jones-Gotman, M. (1991). Human olfactory discrimination after unilateral frontal or temporal lobectomy. Brain, 114: 71-84. Zatorre, R.J., Jones-Gotman, M., Evans, A.C. and Meyers, E. (1992). Functional localization and lateralization of human olfactory cortex. Nature, 360: 339-340. Zucco, G.M. & Tressoldi, P.E. (1989). Hemispheric differences in odour recognition. Cortex, 25:607-615. Zwaademaker, H. (1925). L'Odorat. Paris: Doin.
543
Name Index A Abraham, 515, 518, 531 A bramson, 329, 377 Ackles, 346, 354, 376 Adamek, 540 Adams, 386, 392, 394, 396, 400, 402, 418, 421-424, 443, 464, 489 Ades, 409, 414, 415, 443, 454 Aitkin, 311, 319, 388391, 399-401, 415, 416, 419-422, 426, 427, 429, 433, 437, 443-446, 455, 456, 460, 461 Akiyama, 510, 533 Albrecht, 40, 49 Albright, 143, 157 Ali, 364, 371, 374 Allison, 503, 539 Allman, 127, 155 Alvis, 477, 493 Amadeo, 36, 53 Amon, 415, 460 Amoore, 529, 539 Andersen, 136, 157, 158 Anderson, 107, 122, 127, 141, 155, 274, 290, 401, 443-445, 448, 451, 460 Andy, 507, 531 Angel, 472, 488 Angelelli, 46, 48, 53 Anton, 509, 531, 541 Antonini, 58, 77 A pkarian, 40, 54 Ardila, 472, 473, 488, 539 Arguin, 187, 193 Ariens Kappers, 400, 444 Arnaud, 479, 490 Amesen, 391, 4 AA Arnold, 498, 5 3 8 Artola, 143, 153
Ashton, 481, 489 A tkinson, 31, 50, 487, 488 Ault, 58, 79 Axford, 44, 51
B Babcock, 266, 290 Badcock, 45, 52, 199, 229, 274, 291 Baginski, 384, 394, 402, 4 A 4 Bagot, 37, 41, 44, 53, 268, 295 Baker, 335, 375, 398, 418, 419, 425, 444, 449, 452, 460 Ball, 313, 320 Ballard, 145, 153 Ballesteros, 143, 154 Bailing, 481, 490 Banich, 149, 150, 153, 264, 293 Banks, 396, 397, 4 4 A Barlow, 5, 26 Barnes, 392, 394, 400, 444 Barnett, 509, 531 Barone, 388, 447, 453 Barrett, 84, 122, 293 Barsalou, 26 Bash, 234, 260 Bashore, 234, 258 Bates, 356, 374, 375, 381 Batra, 416, 430, 444, 448, 454, 462 Baylis, 222, 226, 228, 229, 235, 258 Beach, 236, 237, 243, 258 Bear, 510, 531 Beasley, 10, 28 Beaton, 8, 11, 26 Beaumont, 150, 154 Bechterew, 384, 445 Bechtoldt, 272, 290
Beck, 386, 464 Becker, 522, 531 Beeman, 118, 120 Beer, 283, 290 Bekius, 418, 424 Belger, 149, 150, 153 Bellas, 519, 531 Belmore, 401, 440, 459 Bender, 60, 78, 131, 155 Benedict, 356, 375 Bengry, 430, 445 Benigni, 356, 374 Benton, 45, 49, 472, 474, 487-490, 493 Berardi, 8-10, 18, 19, 27, 37, 50, 61, 6371, 73-78, 267, 269, 290, 292 Berkley, 440, 451, 455 Berlucchi, 58, 70, 72, 75, 77, 280, 290 Bernard, 44, 53, 509, 531 Bernstein, 224, 230 Berta, 10, 28 Bertelson, 476, 492 Bertollo, 526, 531 Besson, 509, 531 Best, 356, 375 Betchen, 515, 517, 520, 532 Betz, 325, 354, 378 Bevan, 509, 532 Bever, 135, 153 Beverly, 274, 295 Beyerl, 392, 445 Bialek, 517, 536 Biederman, 128, 143, 154, 155 Bignamini, 226, 231 Bilstrom, 472, 489 Binder, 306, 319, 321 Bindra, 6, 30, 129, 158 Birbaumer, 327, 380 Bisiach, 47, 49 Bisti, 36, 40, 52, 76 Black, 70, 73, 77, 510, 537
544
N a m e Index
Blackwood, 45, 52 Blake, 6, 8, 27, 267, 268, 290 B lakemore, 8, 11, 26, 36, 49 Bledsoe, 398, 445, 451 Bloch, 84, 121 Bloom, 356, 375 Blume, 185, 188, 194, 278, 295 Blumstein, 301, 319, 335, 375 Bobak, 40, 50 Bock, 432, 433, 445 Bodis-Wollner, 40, 46, 50, 77 Bogen, 55, 78, 472, 490, 511, 513, 535 Boles, 10, 27, 91, 120, 257, 258 Bonnet, 37, 44, 53, 268, 295 Borgo, 476, 493 Bormann, 417, 445 Boudreau, 388, 390, 391, 397, 432, 445, 463 Bourk, 391, 445 Bouvet, 508, 532 Bowling, 45, 52 Braddick, 18, 30, 31, 50 Bradshaw, 4, 27, 106, 120, 129, 135, 154, 155, 226, 229, 472, 473, 477, 488, 492 Bragdon, 359, 375 Brauchli, 524, 532 Braun, 472, 489 Brawer, 386, 445 Breitmeyer, 31, 45, 50, 96-98, 120, 124, 267, 272, 273, 290, 293, 295 Bretherton, 356, 374, 375, 381 Brinkman, 471, 473, 479, 488 Brizzolara, 474, 488 Broadbent, 16, 28, 439, 445 Broca, 3, 27 B rodal, 503, 504, 532
Brodie, 473, 489 Brogden, 408, 409, 445 Bromley, 520, 522, 532 Brookhart, 414, 443 Brouwers, 273, 291 Brown, 35, 49, 50, 115, 120, 184, 185, 188, 194, 234, 236, 258260, 358, 375, 388390, 395, 430, 442, 450, 464, 484, 489 Brownell, 398, 432, 433, 445, 462 Browner, 392, 395, 396, 398, 400, 404, 445 B rUck, 278, 296 Brugge, 302, 319, 388, 390, 401, 413-416, 430, 443, 445, 452, 458 Brunso-Bechtold, 387, 392, 394-396, 398400, 402, 421, 429, 446, 449, 451 Bryden, 85, 120, 265, 290, 327, 375, 478, 490 Buchsbaum, 225, 230, 525, 532, 542 Buckles, 163, 193, 285, 290 Buckner, 127, 154 Budhoska, 9, 30 Buhrke, 336, 340, 379 Bullier, 56, 58, 78 Bumm, 384, 446 Bundesen, 163, 190, 193, 23 5-237, 246, 250, 256, 259 Bunt, 58, 68, 77 Bures, 273, 291 Burger-Judisch, 328, 334, 339, 379 Burgess, 524, 532 Burgund, 133, 154 Burr, 13, 25, 29, 37, 46, 48, 50, 53, 294 Burton, 46, 54, 264, 293 Butler, 37, 54, 73, 77 Butters, 507, 539
C Cain, 507, 514, 515, 517, 518, 521, 522, 532, 534, 535, 538, 539 Caird, 389, 390, 446 Calford, 438, 446 Caliskan, 510, 541 Callaway, 327, 375 Caltagirone, 335, 377 Calvanio, 135, 156 Calvert, 275, 290 Camaioni, 356, 374 Campanella, 515, 533 Campanelli, 515, 533 Campbell, 5, 18, 27, 29, 31, 36, 40, 49, 50 Cant, 386, 387, 395397, 402, 446, 461 Carey, 135, 154 Carletti, 279, 293 Carmon, 272, 290, 472, 474, 475, 489, 492 Camahan, 473, 489 Carney, 416, 420, 429, 430, 446, 461 Carpenter, 254, 260, 504, 533 Carr, 476, 494 Carregal, 508, 541 Carroll, 515, 528, 533 Carson, 506, 533 Carter, 234, 258, 260 Caspary, 398, 417, 429, 446, 448, 456 Casseday, 387, 395-400, 402, 411-413, 446, 447, 456, 457, 463, 465, 466 Castaldi, 400, 446 Cauley, 356, 376 Cavanagh, 187-188, 193 Cave, 84, 122, 137, 143, 154, 157, 161, 163, 191, 193-195, 199, 231, 293 Cervero, 510, 541 Chabris, 33, 34, 51, 89, 95, 122, 136, 141, 156, 157, 320
Name Index 545 Chaiken, 276, 278, 290 Chambers, 508, 541 Chan, 279, 291, 391, 416, 430, 436, 437, 446, 465 Chapman, 356, 358, 359, 375, 377, 380 Charron, 472, 4 8 9 Chavis, 308, 319 Cheal, 224, 229 Chedru, 276, 290 Cheng, 78, 396, 398, 457 Chiambretto, 479, 490 Cholewiak, 470, 471, 489 Christen, 19, 29 Christman, 7-10, 12-16, 21-24, 27-29, 33, 35, 50, 51, 64, 65, 78, 95, 113, 114, 116, 117, 120-124, 151, 154, 155, 188, 193, 194, 198, 205, 230, 259, 267, 268, 270272, 275-279, 285, 286, 288-291, 2 9 3 , 294, 521, 533 Chu, 476, 493 Chui, 525, 532 Chung, 254, 260 Churcher, 279, 292 Churchland, 136, 144, 154
Cimino, 279, 295 Cioffi, 478, 483, 489 Clarey, 391, 432, 438, 447 Clark, 356, 375, 384, 406, 459, 525, 533, 535 Clopton, 430, 445, 448 Code, 401, 447 Coffey-Corina, 366, 368, 377 Cohen, 148, 154, 222, 229, 278, 294, 330, 340, 359, 380, 481, 489 Cohn, 37, 50
Coles, 234, 258, 282, 285, 296 Collewijn, 274, 295 Colley, 473, 489 Collin, 472, 489 Collins, 317, 319, 346348, 351, 352, 354, 373, 380, 470, 471, 489 Colombo, 279, 291, 308, 319 Comiter, 517, 536 Conesa, 12, 14, 29, 64, 65, 78, 95, 121, 198, 205, 230, 270, 293 Con te, 46, 54 Cook, 95, 96, 120 Cooper, 143, 154, 157, 328, 377, 464 Corballis, 135, 154 Corbin, 276, 290 Corkin, 471, 477, 489, 516, 534 Cortez, 438, 439, 447, 463 Corwin, 13, 28 Courchesne, 326, 345, 375 Courtney, 279, 291 Covey, 398, 416, 433, 446, 447, 463 Cowan, 326, 343, 375, 503, 504, 539 Cowen, 526, 531 Cowin, 84, 85, 97, 99, 109, 110, 119-121, 286, 287, 291, 521, 536 Cragg, 265, 291 Crandall, 518, 528, 540 Cranney, 481, 489 Crebolder, 85, 120 Crick, 225, 229 Crosland, 254 Crossman, 278, 294 Curcio, 170, 193 Curtis, 417, 419, 447
D Damasio, 129, 154, 322
Daniloff, 364, 371, 374, 375 Danta, 273, 291 Davidoff, 56, 77, 266, 267, 291, 417, 447 Davidson, 236, 258, 260, 452, 523, 533 Davis, 348, 377, 414, 432, 447, 460, 462, 506, 521, 533 De Luca, 46, 48, 53 De Nobili, 474, 488 De Renzi, 279, 291 De Vries, 432, 447 De Waal, 484, 492 De Weerd, 74, 79 DeClerk, 364, 368, 371, 377 DeFrancesco, 278, 294 DeHaan, 326, 344, 350, 375 Dehaene, 334, 340, 375 Dehaene-Lambert, 334, 340, 375 Dehay, 56, 58, 78 Delaleu, 508, 532 Delis, 94, 121 DeMichele, 515, 533 DeMonasterio, 98, 121 D~monet, 303, 306, 310, 319 Denckla, 476, 493 Denes, 9, 28, 47, 51 DeOlmos, 506, 533 Derefeldt, 267, 293 deRegnier, 344, 352, 354, 373, 380 DeRenzi, 474, 477, 489 Desimone, 143, 157, 222, 230 Deutsch, 234, 258, 307, 308, 319, 320 DeValois, 5, 6, 16, 27, 127, 267, 291 DeYoe, 34, 50 Di Stefano, 280, 290 Diamond, 46, 50, 118, 120, 135, 154, 310, 319, 321, 401, 413, 447, 453, 455, 457, 465, 510, 533
546
N a m e Index
DiFranco, 289, 291 Dimond, 131, 150, 154, 273, 291, 523, 533 Dimov, 515, 533 Divenyi, 307, 319 Dixon, 410, 447 Dodd, 509, 537 Dodds, 481,489 Dodson, 477, 493 Dodwell, 289, 291 Donchin, 234, 258, 359, 375 Dorman, 329, 376 Doty, 501, 507, 511, 515-520, 522, 524, 529, 530, 532-535, 539, 542 Dowling, 236, 258, 307, 319 Drake, 410, 447 Dr~a, 484, 490 Dreher, 98, 121 Drislane, 46, 52 Driver, 222, 226, 229, 23 5, 258 Duda, 469, 480, 489 Duncan, 190, 193, 199, 215, 229, 258, 529, 534 Dumford, 273, 291 Dustman, 37, 50
E Eacott, 70, 77 Eccles, 418, 447 Edgar, 269, 291 Edmonston, 476, 493 Edwards, 274, 291 Efron, 94, 121, 162, 163, 166, 168, 171, 173, 177, 179, 183, 187-189, 191-195, 235, 259, 261, 264, 276, 279, 282, 285, 290-292, 294, 296 Egeth, 487, 488 Eglin, 278, 292 Ehrlichman, 499, 531, 534
Eichenbaum, 148, 154, 507, 516, 517, 534 Eilers, 329, 376 Eimas, 325, 329, 341, 342, 356, 376 Ekmann, 507, 534 Elbert, 327, 380 Elias, 481, 494 Elliot, 473, 489 Ellis, 347, 373, 475, 494 Elverland, 395, 396, 398, 400, 447 Elwyn, 384, 406, 462 Emmerson, 37, 50 Eng, 84, 121, 521, 534, 536 Engen, 521, 534 Ennis, 503, 506, 540 Enns, 185, 188, 199, 201, 228, 229, 230 Eriksen, 233-235, 245, 246, 250, 256-260 Erkelens, 274, 295 Erulkar, 388, 394, 416, 419, 426, 447, 457, 458 Eskenazi, 514, 515, 517, 518, 519, 531, 534, 535 Ettlinger, 484, 489, 490, 491 Etzweiler, 524, 532 Evans, 300, 306, 319, 321-323, 412, 458, 524, 531, 542 Evarts, 307, 320 Eviatar, 84, 121
F Fagg, 417, 448 Faglioni, 279, 291, 474, 477, 489 Fagot, 471, 479-484, 487, 490, 491 Faingold, 417-419, 431, 446, 448 Farah, 87, 122, 125, 129, 135, 154, 312, 314, 320
Farley, 419, 448 Farrar, 40, 49 Farrington, 273, 291, 523, 533 Fayolle, 479, 490 Fedio, 517 Feldman, 387, 450 Felleman, 127, 155 Fendrich, 11, 27, 60, 76, 77, 199, 229, 263, 292 Feng, 388, 398, 448, 464 Fennell, 472, 490 Ferguson-Segall, 529, 534 Fernandez, 392, 399, 400, 448, 450 Ferretti, 474, 488 Ferrier, 383, 384, 392, 395, 396, 400, 402, 405, 415, 439, 448 Fiess, 356, 375 Fiez, 127, 129, 157 Filbey, 266, 292 Filla, 515, 533 Findlay, 279, 292 Finger, 508, 535 Finke, 267, 292 Fiorentini, 8-10, 18, 19, 27, 36, 37, 40, 50, 52, 63, 67-69, 71, 73, 75-78, 267, 269, 290, 292 FitzPatrick, 302, 320, 433, 448 Flammino, 430, 448 Flanery, 481, 490 Flechsig, 391, 392, 405, 448 Foley, 274, 292 Fontenot, 474, 490 Forbes, 509, 532 Forgays, 281, 282, 285, 294 Forget, 472, 493 Forster, 412, 438, 460 Foster, 417, 448 Foulke, 476, 490 Fox, 127, 155, 321, 523, 537
Name Index 547 Fragaszy, 484, 491 Franco, 478, 483, 490 Franzel, 161, 163, 195, 199, 231 Freeman, 325, 354, 379 Friedman, 118, 120, 224, 230 Friend, 517, 534 Fruh, 95, 120 Frye, 519, 529, 530, 535 Fujii, 143, 156 Fujita, 60, 78 Fukuda, 58, 68, 78, 98, 121 Fukui, 44, 50 Fullerton, 414, 449 Fulton, 384, 449 Furukawa, 529, 535 Fuse, 392, 394, 396, 449
G Gabrieli, 84, 122, 293 Gaffan, 70, 77 Gage, 417, 449 Gainotti, 301, 320, 335 Galaburda, 46, 52, 316, 320, 321, 510, 535, 536, 538, 540 Galambos, 415, 437, 449, 459 Gallagher, 364, 371, 374 Gallaway, 283, 290 Ganz, 31, 33, 50, 54, 326, 345, 375 Garcha, 484, 490 Gardner, 4, 27 Garon, 278, 294 Gawryszewski, 266, 292 Gazzaniga, 11, 27, 52, 55, 60, 76-78, 87, 122, 124, 131, 154, 156, 185, 188, 205, 229, 230, 263, 266, 292, 308, 318,320, 322, 472, 490, 513, 53 5, 540 Geffen, 129, 155 Geisler, 388, 445, 459
Gelade, 190, 195, 199, 231, 257, 260, 276, 296 Gelfer, 340, 359, 376 Geminiani, 226, 231 Geschwind, 510, 535, 538 Gesteland, 501, 535 Getchell, 508, 535, 537 Ghent, 471, 477, 490, 493 Gianotti, 335, 377 Gibson, 131, 154, 470, 478, 490 Gidiuli, 476, 493 Gilbert, 13, 25, 27, 530, 535 Gill, 267, 293, 326, 356, 362, 365, 371, 373, 374, 379 Gish, 270, 296 Glass, 37, 54 Glendenning, 384, 387, 392, 394-400, 402, 404, 418, 419, 423427, 429, 434-437, 44 A, 449, 452, 455 Glick, 4, 28, 540 Goldberg, 388, 389, 390, 395-397, 400, 401, 430, 442, 449, 450, 452, 456, 459 Goldstein, 364, 371, 374 Golinkoff, 356, 376 Gollomp, 518, 534 Goodale, 127, 155, 158 Goodglass, 335, 375, 478, 492 Goodrich, 226, 228, 231 Gordinier, 384, 450 Gordon, 356, 376, 498, 511-514, 535 Gori, 279, 293 Gottlieb, 308, 320 Grabowska, 7, 9, 10, 28, 47, 51, 273, 292 Grafman, 118, 120, 476, 491 Gratton, 234, 258 Gravina, 69, 76 Green, 10, 28
Greenberg, 288, 292 Greenspan, 472, 473, 490 Greer, 507, 540 Gross, 58, 60, 61, 70, 75, 78, 79, 131, 143, 155, 157, 308, 319, 320, 459 Gruzelier, 524, 532 Guariglia, 47, 54 Guez, 274, 292 Guillemot, 67 Guinan, 388-391, 397, 450 Guitierrez, 97, 124 Gunderson, 343, 376 Gunter, 280, 292
tl Haberly, 506, 535 Hackman, 417, 447 Hagbarth, 506, 535 Hall, 498, 537, 538 Halpern, 313, 314, 315, 320, 323 Hamilton, 40, 49, 73, 78, 79, 88, 107, 122, 156, 320 Hammerschlag, 405, 450 Hammond, 274, 275, 296 Hamsher, 273, 292, 474, 488 Handwerker, 509, 541 Hannay, 487, 488 Hans, 328, 334, 339, 379 Hardyck, 8, 21, 29, 75, 79, 264, 292 Hari, 509, 536 Harley, 476, 491 Harmon, 33, 51 Harms, 235-237, 246, 250, 256, 259 Harnois, 40, 50 Harpman, 394-396, 400402, 465 Harrington, 3, 28 Harris, 5, 28, 432, 450
548
Name Index
Harrison, 387, 388, 394397, 402, 450, 451 Hartley, 234 Harvey, 8, 21, 29, 60, 75, 78, 79, 263, 292 Harwood, 307, 319 Hatta, 481, 491, 494 Haun, 60, 78 Haxby, 127, 155 Hazeltine, 109 Heffley, 359, 375 Heffner, 307, 320, 386, 410, 412, 439-441, 451 Heilman, 279, 295 Heine, 510, 536 Held, 384, 387, 391, 392, 394-396, 399, 400, 405, 448, 451, 464 Helfert, 387, 420, 423425, 436, 445, 451 Heller, 96, 120, 264, 293, 472, 476, 489, 491 Hellige, 4, 8, 10, 13, 15, 21-23, 27-29, 31, 33, 35, 51, 53, 84-87, 89-91, 95-97, 99, 106, 109, 110, 117,122, 124, 125, 131, 149, 150, 151, 154, 155, 188, 193, 194, 246, 253, 259, 260, 267, 268, 272, 286, 287, 291-293, 477, 485, 491, 521, 536 Helmholtz, 440, 451 Hendrickson, 170, 193 Hendrix, 524, 537 Henkel, 387, 396-400, 429, 451, 452, 457, 461, 462, 465 Henkin, 517, 536 Henning, 16, 28 Henschel, 351, 354, 380 Hermelin, 476, 491 Heron, 281, 292 Hertz, 16, 28
Hess, 46, 51, 268, 274, 290, 294, 325, 328, 330, 331, 342, 378 Heywood, 279, 292 Hillger, 107, 122 Hillyard, 73, 78, 205, 230 Hilpert, 332, 377 Hilton, 273, 291 Hind, 416, 429, 438, 452, 459 Hinton, 118, 121, 136, 142, 145, 151, 152, 155, 157 Hirsh-Pasek, 356, 376 Hoerster, 484, 491 Hoffman, 190, 193, 250, 259, 344-346, 353, 354, 376, 448 Hofmann, 387, 452 Holley, 508, 532 Holt, 18, 28 Holtzman, 87, 122, 224, 225, 229 Hommel, 236-238, 243, 247, 250, 253-255, 257, 259 Honda, 279, 292, 487, 491 Hood, 356, 375 Hopkins, 481, 482, 484, 490, 492, 493 Horn, 316, 322, 446, 536, 542 Hornung, 517 Horwitz, 88, 122, 155 Howard, 274, 294, 321 Howe, 387, 402, 450 Howell, 510, 537 Hoyer, 85, 99, 124 Hubel, 4, 5, 28, 34, 52, 96, 98, 107, 123, 124 Huber, 517, 536 Hudspeth, 432, 452 Hugdahl, 440, 452 Hughes, 14, 28, 199, 229, 236, 259, 449, 459, 507, 531 Hummel, 128, 155, 498, 499, 508, 509, 511,
522, 523, 524, 531, 536, 537 Humphreys, 190, 193, 199, 229, 258 Hunter, 394, 463 Hurtig, 518 Hurwitz, 525, 533 Hutsler, 318, 320 Hutson, 385-387, 398401, 415, 418, 4239426, 428, 429, 435438, 449, 452, 453, 456 Hutton, 279, 292 Huttunen, 509, 536 Hylkema, 268, 293 Hyson, 396, 446 Hyv~rinen, 170, 194
l Iaccino, 476, 491 Imig, 302, 320, 401, 447, 452, 456 Ingvar, 41, 54 lnhoff, 224, 230 Inokuchi, 509, 536 Ipata, 66, 78 Iriki, 471, 491 Irvine, 430, 443, 452 Irving, 394, 395, 397, 450 Ishikawa, 47, 51 Ito, 78, 413, 436, 453 lversen, 308, 320 I vry, 16, 28, 94, 95, 107-110, 117,,121, 222, 229 Ivy, 507, 541 Iwamura, 471, 491
J Jackson, 280, 292 Jacobs, 95, 115, 116, 118, 121, 122, 141, 142, 155 Jacobson, 235, 259 Jafek, 501, 539 Jane, 411, 453, 455
Name Index 549 Jasper, 341, 352, 357, 376 Jean-Baptiste, 394, 453 Jeffress, 436, 453 Jeffreys, 44, 51 Jenkins, 412, 413, 433, 437, 438, 439, 453 Jerison, 307, 320 Joanette, 187, 193 Johanson, 58, 77 Johnson, 510, 523, 533 Johnston, 417, 418, 453 Jones, 388, 417, 420, 422, 447, 453, 459, 472, 491, 523, 536, 537 Jones,Gotman, 517, 518, 520, 524, 537, 542 Jonsson, 13, 21, 28, 96, 121 Jordan, 136, 157 Julesz, 267, 272, 290, 293 Jungert, 414, 453 Jurko, 507, 531 Jusczyk, 325, 342, 376
K Kalina, 170, 193 Kamhi, 356, 376 Kamide, 529, 535 Kandel, 470, 471, 478, 483, 489, 491, 537 Kane, 388, 445, 453 Kaneko, 507, 541 Kaptein, 192, 193 Karapas, 392, 399, 400, 448 Karni, 179, 194 Karrer, 344-348, 352354, 372, 373, 376, 377, 380 Kashu, 478, 492 Kato, 44, 50 Kauer, 502, 541 Kaukoronta, 509, 536 Kavanagh, 412, 413, 438, 453
Kaye, 8, 11, 29, 37, 51, 75, 78, 267, 268, 293 Keegan, 20, 30 Keenan, 21, 28 Kellar, 336, 380 Kelly, 36, 51, 398, 412415, 431, 433, 438, 453, 465, 509, 537 Kendall-Reed, 542 Kennedy, 56, 58, 78 Kertesz, 510, 537 Kesslak, 525, 532, 542 Kettenmann, 509, 510, 536, 537 Kiang, 414, 432, 449, 453 Kidd, 431, 453 Kim, 191, 194, 432, 458 Kimmelmann, 509, 536 Kimura, 273, 291, 440, 453 Kinchla, 199, 230 King, 498, 537 Kingstone, 185-188, 192, 199, 201, 223, 228-230 Kinnamon, 508, 535 Kinsbourne, 226, 230, 231, 279, 293, 494, 535 Kircher, 37, 50 Kirsner, 282, 296 Kiss, 395, 454 Kitterle, 8-16, 22, 23, 27-29, 35, 37, 51, 64, 65, 75, 78, 95, 113, 120-123, 151, 154, 155, 188, 193, 194, 198, 205, 230, 237, 242, 254, 259, 260, 267, 268, 270, 273, 288, 293, 533 Kitzes, 402, 420-423, 425, 426, 433, 437, 454, 456, 457, 461 Klemm, 41, 44, 51, 524, 537 Klinke, 389, 390, 446 Klug, 418-419, 431, 454 Knapp, 18, 30, 127, 141, 155
Knight, 94, 122, 123, 198, 205, 230 Kobal, 499, 508, 509, 522-524, 531, 534, 537 Kobayashi, 47, 51 Koelega, 520, 537 Koenig, 33, 34, 51, 84, 85, 89, 95, 122, 125, 141, 156, 293 Koemer, 60, 78 Kohn, 222, 229 Koisumi, 419, 447 Koivisto, 133, 155 Kolb, 510, 537 Kolinsky, 278, 294, 321 Kolliker, 384, 391, 394, 399, 400, 454 Koltuska, 9, 30 Komatsu, 143, 156 Konnerth, 509, 537 Kopala, 525, 533 Koslow, 327, 375 Kosslyn, 33, 34, 35, 49, 50, 51, 84-90, 93, 95, 96, 106-110, 115-118, 120-123, 125, 127, 132, 136, 137, 141, 142, 155157, 253, 259, 267, 270, 272, 292, 293, 313, 314, 320 Kostelyanets, 275, 293 Kramer, 235, 259 Kratskin, 506, 507, 538 Kreidl, 384, 406-408, 441, 454 Kreuzer, 332, 377 Kreysel, 509, 541 Kroger, 530, 539 Kroon, 6, 30, 170, 194, 267, 295 Kropfl, 267, 272, 290, 293 Krueger, 254, 259 Kryter, 409, 454 Kucharski, 498, 538 Kudo, 396, 399, 400, 401, 454 Kueck, 282, 285, 293 Kuhl, 342, 3.56, 377
550
Name Index
Kulikowski, 31, 32, 40, 41, 51-53, 267, 294 Kurian, 473, 491 Kuroiwa, 44, 50 Kurtzberg, 332, 377 Kuskowski, 344, 345, 346, 353, 376 Kutas, 359, 375 Kuwabara, 387, 394, 454 Kuwada, 416, 421, 430, 433, 444, 448, 454, 457, 462 Kuypers, 471, 473, 479, 488
L La Villa, 387, 394, 454 LaBerge, 143, 156, 184, 185, 188, 194, 222, 225, 230, 234-236, 259, 260 Lacreuse, 471, 479-481, 484, 487, 490, 491 Ladavas, 279, 293 Laeng; 85, 90, 112, 113, 116, 117, 122 Lahey, 356, 375 Lamantia, 504, 538 Lamb, 14, 30, 94, 122, 123, 151, 157, 198, 199, 205, 230, 231, 253, 260, 264, 295 Lambert, 472, 493 Lamont, 267, 294 Landis, 19, 29, 95, 120 Lang, 347, 373, 508, 538 Lannou, 44, 53 Lassonde, 472, 493 Laurinen, 170, 194 Lawless, 521, 538 Lebby, 16, 28 Leblanc, 276, 290 Lechelt, 475, 491 Lederman, 469, 470, 472, 478-480, 488, 491, 493 Lehky, 136, 142, 156 Leicester, 58, 79 Leicht, 336, 340, 379
LeMay, 510, 535, 538 Lenneberg, 330, 356, 371, 377 Lennerstrand, 267, 293 Lennie, 13, 25, 29, 40, 52 Lentz, 278, 294 Leonard, 473, 491 Lepore, 472, 493 Leventhal, 58, 68, 79 Levi, 46, 52, 170, 194 Levin, 474, 488 Levine, 135, 156 Levitsky, 510, 535 Levy, 264, 282, 285, 293, 481, 489, 526, 531 Lewy, 394, 395, 399, 455 Lhermitte, 276, 290 Li, 414, 415, 431, 453, 507, 525, 533, 534 Liberman, 301, 307, 320, 328, 377, 432, 455 Li6geois-Chauvel, 316, 320 Lifter, 356, 375 Lindsay, 118, 120 Lines, 60, 79, 263, 293 Linnville, 336, 340, 379 Lipsitt, 514, 534 Lisker, 329, 377 Livingston, 432, 462 Livingstone, 34, 46, 52, 96, 98, 107, 123 Loomis, 470, 491 Lorente De No, 386, 455 Lorig, 524, 538 Lovegrove, 45, 46, 52, 97, 124, 199, 229 Lovell, 501, 539 Luchins, 510, 542 Luciani, 383, 455 Luck, 205, 206, 209, 214, 216, 221, 223, 227, 230 Ludorf, 237, 242, 254, 259, 260 Lulenski, 132, 156 Lumb, 510, 541
Lundh, 267, 293 Lutes, 524, 537 Lutzenberger, 327, 380 Lykins, 347, 355, 381 Lyman, 521, 538 Lynch, 507, 525, 532, 541 Lyon, 224, 229, 325, 377
M Macar, 327, 359, 380 MacCabe, 484, 490 MacCana, 267, 294 MacDonald, 127, 158, 321 Mackeben, 40, 53 MacLeod, 503, 538 Macnamara, 356, 377 MacNeilage, 364, 368, 371, 377 Macrides, 506, 533 Maeda, 143, 156 Maffei, 13, 25, 29, 31, 36, 40, 50, 52, 76 Maier, 65, 79 Majorossy, 395, 454 Malaspina, 526, 538 Malcus, 170, 194 Maljkovic, 88, 122, 156, 320 Malpeli, 96, 124 Mangun, 185, 188, 205, 230 Mannon, 507, 534 Matin, 308, 312, 320 Markovitz, 390, 455 Marotta, 540 Marr, 126, 156, 273, 293 Marrocco, 98, 123 Marrota, 515, 540 Marsh, 358, 375 Marshall, 135, 156 Marsolek, 33, 34, 51, 89, 95, 115-117, 122, 123, 132-134, 136, 141, 142, 144, 145, 147-149, 153, 154, 156, 157
Name Index 551 Martin, 45, 52, 94, 123, 199, 230, 321, 387, 390, 392, 420, 444, 464, 465 Martinez, 515, 517, 518, 538 Marzi, 280, 290 Massironi, 47, 54 Masterton, 307, 320, 384, 386, 398, 410, 412, 413, 426, 427, 432-434, 437, 438, 440, 441, 446, 449, 451-453, 455, 457, 459 Mathai, 515, 518, 531 Mattingly, 226, 229, 301, 307, 320 Mattson, 515, 517, 518, 535 Maunsell, 34, 52 May, 45, 52, 96, 120 Mazzoni, 136, 157 McBeath, 275, 294 McCabe, 85, 122 McChesney, 274, 292 McClelland, 118, 121, 127, 141, 145, 148, 155, 157 McClurkin, 98, 123 McConkie, 285, 295 McCrary, 359, 375 McDaniel, 521, 538 McDonnell, 289, 293 McGeer, 417, 418, 455 McGillis, 472, 473, 490 McKeever, 267, 293 McKenna, 312, 321 McKenzie, 410, 447, 473, 492 McKeown, 529, 534 McLean, 506, 539 McNaughton, 148, 157 Mecacci, 25, 35-38, 4042, 44, 46, 48, 5254, 268, 294-296 Medina, 522, 539 Merigan, 34, 52 Merzenich, 389, 391, 412, 438, 439, 444, 453, 455, 460
Mesulam, 514, 539 Metter, 481, 492 Mettler, 408, 445, 455 Mewhort, 254, 260 Meyer, 394, 455, 503, 539 Meyers, 524, 542 Miceli, 335, 377 Michimata, 21, 29, 8486, 88, 90, 91, 109, 110, 121, 123, 272, 292 Middlebrooks, 438, 455 Miezin, 127, 154, 155, 158 Millar, 476, 482, 492 Miller, 234, 250, 260, 322, 356, 377, 432, 460 Millodot, 267, 294 Mills, 6, 8, 27, 267-268, 290, 356, 366, 368, 369, 371, 373, 377 Milner, 127, 155, 263, 293, 307, 308, 310, 311, 321, 440, 455, 473, 478, 481, 484, 491, 492, 516, 539, 540 Minagawa, 478, 492 Minckler, 58, 77 Miossec, 278, 294 Mirza, 530, 539 Mishkin, 58, 60, 61, 70, 75, 78, 79, 127, 155, 158, 281, 282, 285, 294, 308, 320 Mitchell, 471, 494 Miwa, 529, 535 Moffet, 484, 490 Molfese, 325-342, 344, 349, 352, 354-365, 367, 368, 370, 371, 373, 374, 375, 377381 Moll, 364, 371, 375 Mommers, 476, 492 Monakow, 384, 391, 396, 398, 400, 401, 405, 448, 449, 455, 456
Monti, 344, 345, 352, 353, 372, 377 Moore, 143, 154, 329, 376, 386, 387, 398, 400-402, 411-413, 417, 418, 421, 423, 424, 426, 429, 450, 456 Morais, 278, 294 Moran, 222, 230, 501, 539 Morel, 401, 456 Morelli, 10, 27, 77 Morest, 386-388, 394, 422, 445, 448, 452, 453, 456, 457, 463 Mori, 507, 540 Morihisa, 510, 542 Morikawa, 275, 294 Morris, 254, 260, 321 Morrone, 13, 25, 29, 37, 48, 50, 53 Morse, 326, 342, 343, 356, 362, 367, 368, 370, 371, 374, 375, 379 Morton, 516, 534 Moscovitch, 36, 52, 106, 123, 226, 231 Moulton, 511, 541 Mousty, 476, 492 Movshon, 40, 52 Moxson, 3, 29 Mugnaini, 386, 422424, 443, 456, 464 Muir, 289, 291 Mukuno, 47, 51 Mulder, 280, 292 Mullen, 274, 290 Munson, 327, 359, 380 Murasugi, 274, 294 Murphy, 510, 533 Murray, 40, 53, 267, 294 Musolino, 320 Myers, 70, 73, 77, 476, 492
N Nachson, 475, 492 Nadal, 525, 539
552
Name Index
Naegele, 276-279, 291 Nakayama, 40, 53 Nass, 46, 54 Nathan, 472, 488 Navon, 199, 230 Nebes, 481, 492 Neff, 307, 320, 411-413, 445-447, 449, 450, 456, 457, 464 Nejat-Bina, 507, 541 Nelson, 45, 52, 326, 327, 344, 346-352, 354, 373, 375, 380, 416, 419, 457, 462 Nettleton, 4, 27, 135, 154, 472, 488 Neville, 366, 368, 377 Nicholas, 116, 123, 148, 149, 156, 157 Nichols, 163, 173, 189, 193, 195, 235, 261, 276, 291, 292, 296 Niebauer, 8-10, 22, 27, 29, 91-93, 113, 114, 117, 123, 270-272, 285, 289, 294 Niemer, 396, 398, 457 Niimi, 401, 454 Nikkel, 347, 380 Nishizawa, 473, 492 Noell, 170, 194 Nonneman, 510, 538 Norcia, 326, 345, 375 Nordeen, 392, 394, 402, 421, 457 Novelly, 514, 515, 517519, 531, 534, 535, 538 Nowicka, 7, 28 Nowycky, 507, 540 Nudo, 432, 449, 455, 457
O O'Adams, 469, 480 O'Boyle, 273, 291 O'Connor, 476, 491, 515-518, 542 O'Doherty, 517 O'Reilly, 136, 148, 157
Oda, 280, 294 Oesterreich, 412, 462 Ogawa, 187, 188, 195, 536 Ohnisi, 398, 457 Ohta, 127, 158 Oldfield, 357, 363, 380 Oliver, 388, 392, 394, 402, 418, 420-422, 424, 425, 456, 457, 461 Orbach, 285, 294 Orban, 75, 79 Osaka, 280, 294 Oscar-Berman, 478, 487, 492 Osen, 386, 391, 392, 397, 444, 456-458 Ostapoff, 386, 423, 448, 458 Ostrosky-Solis, 163, 189, 194, 282, 285, 294, 539 Owen, 11, 30
P Packer, 170, 193 Palermo, 325, 354, 379 Palet, 279, 292 Pandya, 308, 319, 321 Pantev, 439, 458 Papez, 398, 399, 458 Parham, 432, 458 Park, 431, 453, 454, 458 Parr, 484, 492 Parsons, 335, 379 Pashler, 179, 194, 205, 223, 230, 236, 260 Passafiume, 474, 494 Pauli, 523, 531, 536, 537 Paus, 300, 321 Payer-Rigo, 335, 377 Payne, 266, 294 Pearce, 471 Pendse, 520, 539 Penfield, 412, 458 Penhune, 302, 316, 317, 321 Pennal, 269, 294
Pepe, 21, 28 Peppel, 509, 531 Perani, 226, 231 Perecman, 336, 380 Peretz, 308, 321 Perez, 118, 120 Perry, 308, 312, 320, 321, 323, 476, 491 Peschanski, 509, 531 Peters, 85, 90, 113, 116, 117, 122, 319, 326, 356, 362, 367, 368, 370, 371, 374, 379 Petersen, 127, 129, 154, 157, 158, 301, 302, 306, 310, 315, 321 Peterzeli, 8, 21-24, 29, 75, 79 Petrides, 303, 308, 311, 321 Pettigrew, 438, 455 Pfeffer, 412, 465 Pfeiffer, 518, 534 Phillips, 399, 400, 401, 413, 414, 415, 416, 437, 444, 458 Piccolino, 36, 50 Pierce, 529, 539 Pierson, 472, 488, 492 Pierson-Savage, 472 Pinger, 285, 291 Pinker, 236, 258 Piotrowski, 18, 29 Pirchio, 63, 78 Pisoni, 331, 380 Pizzamiglio, 47, 54, 474, 494 Plaut, 142, 157 Pointer, 46, 51, 268, 294 Poirier, 412, 439, 458 Polich, 184, 188, 194, 278, 294 Poljak, 392, 394, 395, 398, 458 Polk, 510, 537 Pollak, 390, 431, 454, 455, 458 Pollatsek, 280, 295 Porfert, 478, 492
Name Index 553 Posner, 197, 222, 224, 225, 230, 235, 236, 257, 258, 260, 280, 294, 321 Potter, 507, 516, 534, 539 Powell, 394, 447, 458, 503, 504, 506, 539 Prazdny, 273, 294 Prete, 275, 294 Previc, 11, 26, 29, 185, 188, 194, 264-266, 268, 269-274, 278, 283, 286, 288-290, 295 Price, 306, 319, 321, 503, 506, 535, 539 Prinzmetal, 109 Probst, 400, 458 Pruszewicz, 517, 536 Pusakulich, 477, 493
R Rabin, 521, 539 Rademacher, 316, 321 Rafal, 222, 224, 226, 229, 230, 294 Raichle, 127, 154, 155, 158, 197, 222, 224, 225, 230, 300, 321 Raisman, 503, 504, 539 Raj, 529, 534 Rakic, 504, 538 Ramon y Cajal, 386, 387, 391, 392, 394, 395, 399, 405, 458, 459 Ranson, 384, 406, 444, 459 Rao, 6, 8, 29, 267, 268, 295, 315, 319, 321 Rapcsak, 279, 295 Rasmussen, 395, 396, 459 Rausch, 517, 518, 528, 539, 540 Ravizza, 401, 440, 459 Raymond, 227, 231, 274, 275, 295
Rayner, 188, 189, 254, 260, 280, 285, 295 Read, 508, 540 Reale, 302, 319 Rebai, 37, 41, 44, 53, 268, 295 Reeh, 509, 541 Regan, 35, 46, 53, 63, 79, 272, 274, 295 Regg, 524, 532 Rehbein, 478, 492 Reid, 389, 391, 455 Reiser, 313, 320 Rentschler, 19, 29 Retherford, 356, 380 Reuter-Lorenz, 199, 226, 229, 231 Reyes, 503, 504, 506, 540 Rhode, 432, 459 Rhodes, 129, 157, 471, 492 Ribak, 418, 459 Ribaupierre, 400, 460 Rice, 471, 494 Richards, 272, 295 Richardson, 515, 528, 533, 542 Richie, 286, 295 Riege, 481, 492 Rijsdijk, 6, 30, 170, 194, 267, 295 Riolo-Quinn, 473, 492 Risberg, 41, 54 Risse, 514, 535, 540 Ritter, 359, 375 Rizzolatti, 58, 72, 77, 266, 292, 295 Roberts, 418, 459 Robertson, 14, 30, 94, 95, 107-110, 117, 121-123, 151, 157, 198, 199, 205, 230, 231, 253, 260, 264, 295, 465 Robin, 307, 322 Robinson, 307, 319 Robison, 266, 290 Robson, 5, 27 Rocchetti, 44, 53
Rocha-Miranda, 131, 155 Rockel, 388, 417, 420, 422, 459 Rockstroh, 327, 380 Rodieck, 98, 121 Roemer, 36, 53 Rogers, 476, 477, 488, 491 Rose, 388, 389, 391, 416, 430, 449, 452, 459 Rosen, 41,46, 52, 54, 510, 54O Rosenwasser, 530, 535 Rosenzweig, 384, 413, 415, 437, 439, 459, 460 Ross, 18, 28 Roth, 97, 103, 124, 390, 394, 396, 398, 402, 420, 421, 426, 427, 444, 460 Rouiller, 397, 400, 460 Rourke, 6, 8, 29, 267, 295 Rovamo, 6, 30, 170, 194 Rovee-Collier, 343, 381 Rowley, 501, 539 Roy, 473, 492 Rubens, 279, 296 Ruchkin, 327, 359, 380 Rudel, 476, 493 Rueckl, 137, 157 Rumelhart, 118, 121, 128, 136, 141, 145, 155, 157 Rusconi, 226, 231 Ryan, 432, 460 Rybash, 85, 99, 124 Ryugo, 387, 397, 402, 460, 465
S Sabin, 392, 460 Sackett, 343, 376 Sagi, 179, 194 Saint Marie, 425, 460 Sakaguchi, 143, 156 Sakoda, 270, 296
554
Name Index
Salapatek, 326, 327, 344-346, 349, 353, 354, 376, 380 Samson, 302, 308, 312, 322, 323 Sanchez-Longo, 412, 438, 460 Sanders, 278, 296 Sanes, 398, 460 Sanford, 477, 493 Sanides, 316, 320 Santhakumari, 473, 491 Santorelli, 515, 540 Sartucci, 73 Saslow, 473, 492 Satz, 472, 490 Saul, 414, 447, 460 Sawada, 540 Schacter, 116, 123, 127, 143, 148, 154, 156, 157 Scheibel, 26, 30, 387, 460 Schein, 98, 121 Schiavetto, 472, 493 Schiessl, 508, 536 Schiller, 96, 124 Schmidt, 46, 51, 328, 336, 337, 340, 342, 379, 447, 499, 530, 540 Schneider, 499, 530, 540 Schofield, 386, 461 Schottler, 507, 541 Schuck, 420, 444 Schultz, 234, 245, 259 Schwartz, 143, 157, 282, 296, 356, 380, 387, 440, 449, 451, 461, 470, 471, 491, 492, 524, 537, 540 Scobey, 170, 194 Scotti, 474, 477, 489 Scoville, 516, 540 Seacord, 61, 70, 75, 79 Searock, 326, 335, 341, 342, 373, 379 Sechi, 46, 52 Segalowitz, 330, 340, 359, 375, 378, 379, 380, 472, 491
Sejnowski, 136, 142, 144, 154, 156 Seldon, 318, 322 Selig, 9, 10, 29, 95, 113, 121, 151, 155, 188, 194 Semenza, 9, 28, 47, 51, 476, 493 Semmes, 471, 472, 477, 493 Semple, 390, 415, 416, 420-423, 425-427, 429, 433, 437, 438, 446, 454, 456, 461, 462 Serafetinides, 517, 518, 528, 539, 540 Sereno, 254, 260 Sergent, 6, 7, 13, 20, 21, 23, 30, 31, 33, 35, 49, 53, 84, 85, 91, 93-95, 117, 121,124, 127, 129, 151, 152, 157, 158, 246, 253, 260, 264, 270, 296 Sersen, 471, 494 Shagass, 36, 53 Shaman, 520, 534, 535 Shaner, 387, 461 Shankweiler, 328, 377 Shapley, 96, 124 Sharma, 473, 491 Shedlack, 507, 534 Shepherd, 502, 507, 540, 541 Sherman, 58, 79, 510, 540 Shipley, 503, 504, 506, 540 Shiu, 179, 194 Shneiderman, 397, 398, 399, 400, 424, 429, 461 Shore, 356, 375 Shulman, 270, 296 Sidtis, 307, 322 Silver, 327, 359, 380, 511, 541 Silverman, 79, 127, 158, 445
Simada, 417, 418, 419, 426, 464 Simos, 328, 331, 332, 335, 342, 380, 381 Singer, 143, 153, 268, 296 Siqueland, 325, 342, 376 Sivian, 432, 461 Skandries, 266, 296 Slaghuis, 45, 52 Sloan, 170, 193 Slotnick, 507, 541 Smith, 269, 274, 275, 291, 296, 358, 375, 394-397, 432, 433, A4A, 448, 450, 459, 461, 476, 493, 502, 529, 534, 535, 541 Smoak, 413, 446 Snow, 509, 510, 536, 537, 541 Snowdon, 342, 380 Snyder, 236, 260, 356, 375, 381 Spalten, 476, 493 Spangler, 386, 395-397, 399, 400, 429, 452, 462 Sperling, 515, 516, 517, 518, 542 Sperry, 55, 70, 73, 78, 79, 263, 455, 472, 478, 483, 490, 498, 511-513, 535 Spinelli, 36-38, 40-42, 44, 46-48, 52-54, 63, 78, 268, 296 Spirou, 432, 433, 462 Spitzer, 390, 462 Sprague, 74, 75, 77, 79, 319, 508, 541 Squire, 115, 123, 127, 132, 143, 148, 154, 156, 158, 504, 541 St. James, 250, 260 Stanford, 438, 444, 454, 462 Starr, 432, 462 Staubli, 507, 541 Steen, 509, 541 Stefan, 509, 537
Name Index 555 Stein, 287, 296 Steinschneider, 306, 322 Stellar, 508, 541 Stenberg, 41, 54 Stephenson, 18, 30 Stepien, 308, 322 Stern, 518, 534 Stevens, 432, 462 Stewart, 502, 541 Stokes, 400, 462 Stone, 58, 68, 79, 508, 541 Stotler, 387, 389, 394, 395, 398, 462 Streeter, 329, 381 Streitfeld, 473, 481, 493 Strominger, 392, 395, 402, 412, 462 Strong, 384, 406, 462 Studdert-Kennedy, 328, 377 Stumpf, 440, 462 Sugishita, 60, 79 Sullivan, 270, 296, 343, 381, 477, 493 Summers, 469, 478-480, 488, 493 Suomi, 326, 343, 375 Sutherland, 510, 538 Sutton, 327, 359, 380, 415, 460 Switkes, 79, 127 Szelag, 9, 10, 30 Szmeja, 517, 536
T Talairach, 317, 319, 322 Tallal, 318, 322, 440, 461 Tan, 510, 541 Tanabe, 507, 541 Tanaka, 60, 78, 79, 471, 491 Tang, 84, 122, 293 Tarr, 128, 158 Tasaki, 47, 51 Tassinari, 58, 77, 280, 290 Taylor, 478, 481, 492, 536
Tei, 11, 30 Temme, 170, 194 Testa, 9, 28, 47, 51 Teuber, 60, 78, 471, 477, 493 Teuting, 359 Theeuwes, 192, 193 Thibault, 472, 473, 493 Thomas, 347, 355, 381 Thompson, 88, 122, 156, 320, 384, 392, 394, 395, 400, 418, 419, 423, 429, 436, 438, 446, 447, 449, 463 Thornton, 40, 50 Thurauf, 509 Thurber, 478, 481, 483, 493 Tieger, 33, 54 . Toet, 170, 194 Tolbert, 394, 463 Tolhurst, 31, 32, 52 Tootell, 60, 68, 79, 127, 158 Toulouse, 519, 541 Toumoux, 317, 322 Toyama, 143, 156 Tramo, 308, 321, 322 Treisman, 184, 190, 191, 194, 195, 199, 205, 223, 231, 234, 257, 260, 276, 296 Tressoldi, 520, 542 Tsal, 280, 296 Tschermak, 384, 394, 463 Tsuchitani, 388-391, 397, 430, 432, 445, 463 Tueting, 327, 375 Tulving, 148, 157 Tunturi, 437, 463 Turner, 384, 392, 394396, 400, 402, 415, 448, 463 Tyler, 40, 54, 268, 296 Tynes, 343, 381
U
Umeda, 529, 535 Umilth, 267, 296 Underwood, 265, 290 Ungerleider, 127, 155, 158 Uribe, 472, 488
u Vallar, 47, 49 Valler, 226, 231 Van Canffort, 4, 27 van der Heijden, 192, 193
van der Wildt, 6, 30, 170, 194, 267, 295 Van Essen, 34, 50, 56, 79, 96, 98, 124, 127, 155 Van Gehuchten, 391, 392, 394, 396, 400, 405, 463 van Kan, 170, 194 Van Kleeck, 14, 30, 94, 124, 151, 158, 270, 296 Van Noort, 392, 395, 396, 398-400, 402, 463 van Toiler, 499, 522, 524, 531, 542 Vandenbussche, 75, 79 VanDeventer, 267, 293 Varney, 474, 487, 488, 493 Vaschide, 519, 541 Vater, 398, 463, 464 Vauclair, 471, 479-483, 487, 490, 491, 493 Vaughan, 332, 377 Vella, 37, 54 Vellutino, 45, 54 Verhaart, 387, 464 Verjat, 478, 481, 493 Vetter, 423, 464 Victor, 46, 54 Viggiano, 37, 41, 44, 46, 48, 52-54, 268, 295 Vigorito, 325, 342, 376 Vincent, 274, 295
556
Name Index
Vincenzi, 394, 464 Virsu, 6, 30, 170, 194 Vitek, 58, 79 Volkmann, 276, 290 Volpe, 307, 322 Volterra, 356, 374, 375 von Bekesy, 542 von Brunn, 508, 542 Von Economo, 316, 322 von Skramlik, 499, 542
W Walker, 410, 464 Wallace, 129, 155 Wallen, 484, 490 Walsh, 44, 54 Wang, 60, 79, 336, 340, 379 Ward, 226, 228, 231, 477, 493 Warr, 386, 392, 394400, 402, 413, 423, 443, 451, 456, 462, 464, 465 Warrenburg, 524, 537 Watanabe, 78, 417, 418, 419, 426, 464 Webster, 246, 259, 392, 395, 396, 398, 400, 404, 420, 429, 443, 445, 464, 478, 481, 483, 493 Wegner, 509, 541 Weinberger, 510, 542 Weiner, 114, 116, 117, 124 Weinstein, 273, 295, 471, 472, 477, 493, 494 Weiskrantz, 308, 320 Well, 280, 295 Wendler, 508, 536 Wendt, 41, 54 Wenthold, 398, 418, 419, 425, 443, 445, 449, 451, 464, 465 Werntz, 530, 542 West, 386, 465, 515518, 542
Wetzel, 326, 336, 340, 344, 352, 354-356, 362, 365, 371, 373, 374, 379 Whishaw, 510, 538 White, 423, 432, 461, 465 Whitfield, 306, 322 Whitley, 387, 399, 400, 429, 465 Whitman, 6, 8, 20, 21, 28-30, 267, 295 Whitworth, 199, 229 Wiesel, 4, 5, 13, 25, 27, 28, 98, 124 Wijers, 280, 292 Wiles, 471, 472, 494 Wilkinson, 476, 494 Willard, 387, 392, 460, 465 Williams, 96-98, 120, 124, 136, 157, 481, 492, 508, 541 Wilson, 329, 376, 472, 488, 535, 54O Winer, 401, 447, 457, 465 Winkler, 383, 406, 465 Winter, 509, 532 Wise, 303, 306, 310, 315, 319, 321, 322, 472, 490 Witelson, 478, 482, 494 Wojtascek, 348, 377 Wolfe, 161, 163, 191, 193, 195, 199, 230, 231 Woods, 187, 188, 195 Woollard, 394-396, 400402, 465 Worrall, 282, 285, 296 Worsley, 300, 319, 321, 322 Wortis, 412, 465 Wu, 398, 433, 465, 526, 542 Wurtz, 224, 231 Wyatt, 510, 542 Wyers, 415, 460
Y Y amamoto, 481, 494 Y andell, 481, 494 Y ates, 432, 465 Yeats, 509, 532 Y eh, 234, 246, 256, 257, 260 Yiannikas, 44, 54 Yin, 391, 416, 420, 429, 430, 436, 437, 446, 454, 457, 461, 465 Y oshida, 398, 465 Young, 98, 123, 475, 494 Y oungentob, 498, 520, 542 Y ousem, 507, 534 Y und, 6, 27, 162, 163, 166, 168, 171, 173, 177, 179, 183, 187189, 191-195, 235, 261, 276, 282, 285, 290-292, 294, 296
Z Zatorre, 301-303, 306, 307, 308, 310, 312315, 321-323, 517, 518, 520, 524, 537, 542 Zeier, 524, 532 Zeki, 56, 79, 225, 231, 265, 296 Zihl, 46, 51 Zimba, 236, 259 Zipser, 136, 158 Zoccolotti, 47, 53, 54, 474, 494 Zola, 285, 295, 541 Zola-Morgan, 504 Zook, 387, 395, 396, 399, 400, 438, 454, 455, 465, 466 Zopello, 476, 493 Zucco, 520, 542 Zwaardemaker, 522, 542
557
Subject Index A acoustic chiasm, 384, 433, 434-436, 441 acuity, 76, 255, 256, 267, 270, 272, 283, 286, 413 (see also resolution) adaptation, 11, 36, 40, 516 agnosia, 47, 129 alpha, 524 amnesia, 143 amusia, 308 amygdala, 503, 506, 507, 509, 516 analytic/holi stic dichotomy, 4, 135, 487 anisotropies, 272, 285 anosmia, 511 anterior commissure, 56, 60, 70, 72, 73, 498, 504, 511, 513, 514 anterolateral system, 470, 509 aphasia, 301, 335, 336 attention, 14, 23, 88, 89, 92, 107, 111, 143, 161, 163, 165, 166, 176, 184, 185, 190, 191, 198, 199, 205, 206, 214, 216, 222227, 233, 234-236, 253, 257, 258, 278281, 283, 285, 286, 289, 303, 353, 472, 473, 482, 486, 513 (see also spatial attention) auditory detection, 410, 411, 438, 441 auditory discrimination, 302, 303, 306, 307, 325, 328-331, 335343, 354, 357, 359, 362, 364, 367, 370,
373, 441 auditory 329, auditory 308, 354 auditory 308, 355, 441
374, 411, 440, identification, 342, 343 memory, 300, 311, 312, 343, recognition, 326, 343, 354, 364, 373, 440,
B band-pass, 7, 19-21 binocular, 58, 60, 273, 284 blur, 7, 19-21, 97, 104, 110, 111, 117 Braille, 475, 476, 482, 486 brightness, 96, 221, 222 (see also luminance) Broca's area, 302, 303, 306, 307
C callosal agenesis, 63, 73 categorical relations, 33, 36, 83-99, 101, 104, 106-109, 111-117, 141, 270, 272, 275, 281, 283, 286, 289 categorical perception, 329-331, 336 chili peppers, 509 chimeric faces, 264 chromatic, 48, 73, 74 coactivation, 64 coarticulation, 364, 367, 368, 371, 374 cochlea, 383, 390, 395, 405, 408-410, 412, 423, 432, 434, 437 color, 48, 73, 97, 99103, 107, 111, 117119, 164, 187, 188,
192, 235, 246, 256, 257, 269, 287 colour constancy, 73 commissure of Probst, 400, 401, 405, 407, 411, 413, 429 commissure of the inferior colliculus, 401, 404, 405, 411, 413-415 commissurotomy, 60, 308, 472, 478, 513 (see also split brain) contrast sensitivity, 7,8, 11, 33, 36, 37, 46, 47, 70, 75, 267, 268 coordinate relations, 8399, 101, 104-117, 141, 269-272, 275, 281, 283, 286, 289 corpus callosum, 55, 56, 58, 60, 70, 72, 73, 75, 76, 131, 149, 150, 197, 201, 206, 209, 214, 215, 222, 401, 411, 413, 440, 471, 472, 498, 504, 511, 513, 514, 519 critical flicker fusion, 268 cutaneous, 470-473, 485487
D dendritic fields, 373, 388 dichhaptic, 476, 479, 481, 482, 486 dichotic listening, 91, 439, 440 dopamine, 498, 518 dorsal ('where') pathway, 127, 265, 285 duration, 25, 41, 235, 276 dyslexia, 46, 287 (see also reading disability)
558
Subject Index
E emotion, 118, 523 epilepsy, 313, 498, 511516, 518, 522, 523, 528 equiluminant, see isoluminant evolution, 4, 146, 151153, 273, 312, 497 exposure duration, 7, 19, 20, 21, 23-25, 31, 33, 209, 215, 264, 266, 269, 275, 278, 279, 282 eye dominance, 37, 44, 288 eye movement, 47, 202, 206, 209, 222, 224, 274, 278, 279, 358
F facial processing, 5, 6, 7, 13, 20, 21, 23, 31, 33, 129, 130, 344, 3 54 feature detection, 5, 6, 22 Feature Integration, 191, 221, 222, 257, 278 flanker compatibility effect, 234-238, 241257 flicker, 31, 32, 268, 287 fornix, 504, 506 Fourier analysis, 5, 6, 20 frontal lobes, 302, 303, 308, 311, 312, 314, 344, 355, 358, 359, 361, 362, 364, 369, 370, 371, 373, 473, 477, 498, 507, 510, 514, 524, 526 fundamental frequency, 6, 9, 13, 16, 23, 68, 74, 301
G geniculo-cortical, 60, 65, 70, 74
global, 5, 14, 94, 97, 151, 198, 199, 201, 203-205, 209, 220, 221, 237, 253, 254, 265, 270, 272, 276, 283, 286, 287 grouping, 199, 223, 235 Guided Search Model, 161-166, 171, 173, 175, 179, 184-192
H handedness, 35, 44, 90, 113, 362, 469, 472474, 487, 519, 520 haptic, 470, 478-485, 487, 488 harmonic frequency, 6, 9, 13, 16, 23, 67, 73, 74, 307 head tilt, 288, 472 hemiretina, 8, 263, 268, 288 high-pass, 103, 104, 105, 111 hippocampus, 56, 148, 504, 506, 507, 510, 516, 527 hippocampal commissure, 498, 506 holistic processing, 131, 135, 142, 145-148, 151, 152 horizontal meridian, 61, 170, 171, 190, 266, 275, 279, 281, 282, 284, 285, 288
1 illusory conjunction, 278 imagery (auditory), 310, 312-315 imagery (visual), 88, 114, 310, 312 inferior colliculus, 311, 385, 387-396, 399, 400-405, 412-431, 433, 435-438, 442
inferotemporal cortex, 56, 58, 60, 70, 75, 127, 143 inhibition of return, 280 interchannel inhibition, 13-15, 25, 256 interhemispheric, 55, 56, 61, 63, 66, 67, 69, 70, 71, 72, 73, 74, 75, 76, 147, 149, 150, 224, 513, 514 isoluminant, 48, 73, 74, 98, 99, 287
K kinesthetic, 470, 471, 473, 485, 488
L language, 3, 4, 26, 118, 190, 280, 285, 325, 326, 327, 329, 330, 335, 358, 368, 370, 371, 372, 374, 483, 510, 521, 522, 528 language acquisition, 4, 325, 356, 368, 371 (see also word acquisition) lateral geniculate nucleus, 58, 76, 107 learning, 61, 71, 73, 74, 75, 115, 134, 177, 179, 180, 181, 183, 189, 289, 481 lemniscal system, 470, 471 limbic, 507, 526 linguistic, 4, 26, 302, 368, 476, 478, 482, 485, 521, 527 local, 14, 94, 95, 97, 144, 151, 197, 198, 199, 201, 203, 204, 205, 209, 220, 221, 237, 253, 254, 264, 270, 272, 283, 286, 287
Subject Index 559 localization (visual), 56, 83, 93, 95, 97, 112, 255-258, 267, 269, 273 location invariance, 143 low-pass, 19, 20, 103, 104, 105, 198 luminance, 5, 7, 9, 10, 12, 25, 33, 41, 48, 74, 85, 94, 166, 199, 236, 264, 269, 282 (see also brightness)
M magnocellular, 33, 34, 35, 40, 46, 48, 49, 96, 97, 98, 99, 103, 104, 105, 106, 107, 286, 287, 288, 289, 290 masking, 10, 21, 98, 216, 343 maze learning, 470, 477 medial geniculate nucleus, 302, 383, 385, 399, 401, 404, 408, 412, 424, 437, 438 melodic processing, 300, 307, 308, 309, 310, 311, 312 (see also music) memory, 35, 71, 131, 133, 147, 148, 279, 308, 326, 327, 343, 344, 345, 352, 353, 354, 355, 356, 373, 374, 482 metacontrast, 97-100 missing fundamental, 16, 307 modularity, 125, 132, 141, 301 motion, 5, 31, 41, 96, 97, 98, 99, 100, 113, 227, 274-276, 283, 286, 287 music, 299, 307, 308, 309, 312, 313, 318, 440 (see also melodic processing)
N nasotemporal overlap, 58, 60, 75, 263 neglect, 47, 48, 222, 226, 279, 280, 283, 508, 514, 530
0 occipital lobe, 36, 37, 47, 49, 66, 127, 265, 344, 352, 353, 369, 510, 513, 526, 527 odor memory, 504, 507, 515-517, 520, 522, 525-529 olfactory, 497, 498, 499, 501, 503, 504, 506, 507, 508, 509, 510, 511, 512, 513, 514, 515, 516, 517, 518, 519, 520, 521, 522, 523, 524, 525, 526, 527, 528, 529, 530 olfactory bulb, 56, 498, 501, 503, 504, 506, 507, 510 olfactory detection, 499, 515-520, 529 olfactory discrimination, 499, 503, 507, 514517, 520 olfactory identification, 499, 507, 514, 515, 516, 518, 519, 526, 527, 529 olfactory localization, 530 olfactory recognition, 513, 514, 517, 520, 521, 527 optic chiasm, 70, 434, 435 optokinetic nystagmus, 274, 283 orientation, 11, 12, 14, 18, 66, 71, 75, 107, 164, 170, 179, 183, 188-190, 192, 222
orienting response, 224, 226, 281, 285, 508
P parietal lobe, 47, 60, 127, 225, 228, 265, 279, 280, 302, 308, 314, 331, 342, 352, 353, 355, 358, 360, 361, 369, 370, 371, 471, 477, 513, 524 parts-based, 131, 135, 142, 144, 146, 147, 151 parvocellular, 33, 34, 35, 40, 46, 48, 49, 97, 98, 103, 104, 286, 287, 288, 289, 290 pattern recognition, 18, 19, 276, 281, 282, 283, 343 (see also visual identification) phase, 5, 9, 10, 18, 19, 67, 70, 71, 72, 73, 74, 75, 94, 269 phonology, 127, 247, 299-303, 306, 307, 325, 326, 328-330, 337, 342, 521 pitch, 114, 116, 117, 302, 307-314, 318 place of articulation, 328, 329, 332, 335-342, 372 pressure, 470, 471, 473, 485, 486, 487 primate, 4, 56, 60, 66, 70, 73, 96, 107, 127, 343, 423, 438, 483, 484, 488, 507 priming, 92, 93, 115, 116, 132, 133, 143, 144, 145, 147, 148, 238, 257, 521
R raster delay, 169 reaching, 289, 484
560
Subject Index
reading, 189, 255, 282, 325, reading
45, 127, 188, 190, 236, 254, 258, 275, 280, 284-286, 290, 476, 482 disability, 44-46 (see also dyslexia) receptive field, 33, 34, 58, 60, 71, 75, 95, 96, 98, 106, 107, 110, 115, 116, 118, 143-145, 181, 222 reference frames, 264 relative auditory frequency, 16 relative spatial frequency, 15, 16, 95, 107-111, 115, 117 resolution, 25, 31, 170, 171, 189, 267, 269, 270, 272, 281, 283, 286, 472 (see also acuity) resource, 163, 165, 175, 208, 227, 228 response bias, 22, 23 response competition, 234, 236, 238, 255, 256 retinal eccentricity, 7, 8, 24, 69, 72, 264, 272, 278 retinotopic, 127, 136, 143
S saccade, see eye movement saccadic latency, 279 scan, 165, 166, 185, 187, 199, 219-221, 235, 238, 247, 254258, 264, 276, 278286, 289, 313 semantic, 26, 118, 370 sex differences, 189, 266, 362, 370, 474, 510, 52O simple reaction time, 266
size, 5, 7, 21, 23, 24, 36, 44, 46, 48, 87, 113, 143, 185, 188, 189, 272 size invariance, 75, 143 sound localization, 411415, 430, 433, 436438, 440-442 spatial attention, 197, 198, 199, 205, 215, 218, 220, 224, 225, 234, 279 (see also attention) spatial coding, 480 spatial frequency, 4-14, 16, 18-29, 31-36, 40, 41, 44-50, 52-54, 6365, 67-71, 73-79, 9497, 103, 107-111, 113, 114, 117, 120124, 151-155, 188, 190, 193, 194, 198, 229, 253, 259, 264270, 272, 273, 276, 282-284, 286, 288, 291,293, 294, 296 spatial nonuniformities, 161, 162, 163, 165, 166, 168, 169, 170, 173, 179, 186, 190, 191 speech, 89, 90, 91 , 299, 301 302, 303, 306, 307 310, 315, 318, 325 326, 327, 328, 330, 331, 332, 335, 336 337, 338, 339, 340, 341, 342, 354, 356, 358, 364, 367, 368, 371, 372, 373, 374 439, 513, 528 splenium, 56, 70, 506, 513 split brain, 11, 55, 73, 87, 185-187, 198, 201, 203-206, 209, 214,-216, 219, 220, 222-224, 227, 228, 440, 498, 499, 511514, 527, 528 (see also commisurotomy)
split-chiasm, 66, 70, 73, 75 stereoacuity, 272 stereopsis, 60, 272, 273, 283, 284 structural invariance, 128, 131, 134, 142, 144, 146, 149, 151, 272 subcortical, 55, 56, 73, 198, 215, 216, 218, 222, 224, 225, 311, 508, 526, 527 superior colliculus, 222, 224, 225 superior temporal gyrus, 302, 306-310, 314 supranuclear palsy, 224 sustained, 25, 31, 34, 40, 41, 97, 98, 275, 286, 388, 416, 418
T tactile, 469-488 tactile discrimination, 470-475, 477, 478, 480, 482-487 tactile identification, 478 tactile localization, 471, 472 tactile memory, 475 tactile orientation, 470, 474, 475, 487 tactual localization, 472 tactile recognition, 474, 476, 478-482 temporal frequency, 31, 32, 34-37, 40-49, 53, 61-65, 67, 74, 76, 107, 108, 114, 115, 268, 269, 284, 286 temporal lobe, 36, 37, 40, 44, 49, 60, 127, 222, 265, 302, 306, 307 308, 310, 311, 312, 313, 314, 315, 330, 337, 339, 341, 342 355, 357, 358, 359, 360, 369, 370, 371 405, 439, 440,
Subject Index 498, 507, 514, 515, 516, 517, 518, 523, 525, 527, 528 temporal phase, 430 temporal resolution, 96, 268, 269, 283, 286, 524 thalamocortical, 317, 471 thalamus, 222, 225, 383, 471, 498, 506, 509, 526 theta activity, 524 timbre, 307, 312 tonotopic, 388, 390, 391, 397, 420, 437, 442 transient, 25, 31, 34, 40, 41, 45, 46, 98, 286 trigeminal, 392, 406, 499, 508, 509, 511, 519, 522, 523, 530
u ventral ("what") pathway, 127, 265, 285 verbal, 24, 26, 91, 282, 315, 439, 440, 476, 488, 512, 520, 521, 527 verbal code, 247, 257, 483, 487 vertical meridian, 56, 58, 60, 61, 64, 67-71, 73, 74, 76, 165, 168, 170, 171, 173, 175, 181, 222, 226, 263, 266, 275, 276, 279281, 284, 285 vibration sense, 470473, 476, 485-487 visual detection, 8, 9, 11, 12, 22, 23, 32, 35, 64, 162, 163, 165, 168, 170, 173, 189, 220, 236, 266-268, 272, 273, 278, 285, 288 visual discrimination, 9, 11, 12, 19, 23, 35,
47, 61, 66, 67, 6971, 73, 75, 86, 96, 108, 109, 168, 170, 179, 257, 269 visual identification, 10, 11, 12, 23, 64, 9397, 133, 184, 197, 198, 205, 216, 220, 221, 223-227, 235, 237, 246, 255, 268, 278, 288 (see also pattern recognition) visual lobe, 170, 171, 173, 176, 177, 179, 278, 279 visual pathways, 5, 1214, 23, 25, 32, 40, 46, 48, 49, 88, 9598, 103, 110, 127, 265, 269, 273, 285290 visual recognition, 56, 71, 126, 127, 128, 129, 130, 135, 136, 142, 146, 150, 151, 152, 270, 273, 278, 283 visual search, 161-190, 197-229, 264, 265, 270, 276, 278-282, 283, 285, 286 visuomotor, 66, 265, 271, 274 voice onset time, 328, 329, 330, 331, 332, 335, 336, 337, 338, 339, 340, 341, 372 voicing, 329, 331, 335, 336, 357
W Wernicke's area, 439, 440 word, 13, 24, 115, 129, 165, 246, 282, 326, 356, 357, 358, 359, 360, 361, 362, 368, 371, 373, 374, 478, 521 word acquisition, 326, 343, 356, 360, 364,
561
368, 370 (see also language acquisition) word recognition, 128130, 281, 283 word-stem completion, 116, 132-134, 147, 148 working memory, 303, 307
