Growing Points in Developmental Science: An Introduction

Growing Points in Developmental Science

Growing Points in Developmental Science is an ISSBD publication based on the millennium symposia papers published in the International Journal of Behavioral Development in 2000. This collection of overview chapters summarises the state of the art and the way forward for this discipline. Experienced researchers as well as younger, cutting-edge scientists have contributed to this international collection. The topics range from early experience to old age, and include issues in both social and cognitive development. Particular interests are investigated, such as the biological substrates of behavioural development, early experiences in terms of both basic and applied science, and cross-cultural contexts of development. Personality, knowledge and the acquisition of memory are also considered. In each case, the authors survey the history and traditions that have marked their research areas, as well as the current status and outlook. Growing Points in Developmental Science represents expert wisdom rooted in a bird's-eye view of the trends and controversies that have helped to shape the discipline, its contributions to science, and its application. It is intended as a resource for scientists of different generations interested in developmental science, and will appeal to advanced students and young investigators as well as seasoned researchers. Willard W. Hartup is Professor Emeritus in the Institute of Child Development at the University of Minnesota, where he taught for many years. He has also been the Institute's Director and President of the International Society for the Study of Behavioural Development and the Society for Research in Child Development. Rainer K. Silbereisen is Professor and Chair of Developmental Psychology at the University of Jena, and Adjunct Professor of Human Development and Family Studies at the Pennsylvania State University. He is President of the German Psychological Association and the International Society for the Study of Behavioural Development as well as a member of the Academia Europaea and a Fellow of the American Psychological Association.

Growing Points in Developmental Science An Introduction

Edited by

Willard W. Hartup

Institute of Child Development, University of Minnesota, Minneapolis, USA and

Rainer K. Silbereisen

Department of Developmental Psychology, Friedrich-SchillerUniversity, Jena, Germany

First published 2002 by Psychology Press 27 Church Road, Hove, East Sussex, BN3 2FA Simultaneously published in the USA and Canada by Psychology Press 29 West 35th Street, New York, NY 10001 This edition published in the Taylor & Francis e-Library, 2005. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Psychology Press is part of the Taylor & Francis Group Ø 2002 ISSBD Cover design by Jim Wilkie All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Growing points in developmental science : an introduction / edited by William W. Hartup & Rainer K. Silbereisen. p. cm. Includes bibliographical references and index. ISBN 1-84169-311-1 1. Developmental psychology. I. Hartup, William W. II. Silbereisen, R. K. (Rainer K.) 1944± BF713.5 .G765 2002 155Ðdc21 ISBN 0-203-98926-0 Master e-book ISBN

ISBN 1-84169-311-1 (Print Edition)

2002072796

Contents

List of ®gures List of tables List of contributors Preface 1

Motor development as foundation and future of developmental psychology

vii viii ix xi

1

ESTHER THELEN

2

The early experience assumption: Past, present, and future

24

H. RUDOLPH SCHAFFER

3

Behavioural genetics in the 21st century

47

ROBERT PLOMIN

4

Nutritional de®ciencies as a biological context for development

64

THEODORE D. WACHS

5

Prospects for research on the quality of pre-school experience

85

EDWARD C. MELHUISH

6

Development of children's knowledge about the mental world

102

JOHN H. FLAVELL

7

Domain-speci®c constraints of conceptual development

123

GIYOO HATANO AND KAYOKO INAGAKI

8

Research on memory development: Past and present WOLFGANG SCHNEIDER

143

vi

Contents

9 Mindreading, emotion understanding, and relationships

167

JUDY DUNN

10 Lifespan personality development: Individual differences among goal-oriented agents and developmental outcome

177

CORNELIS F.M. VAN LIESHOUT

11 Perspectives on gender development

202

ELEANOR E. MACCOBY

12 Agency in development

223

TODD D. LITTLE

13 The development of aggressive behaviour during childhood

241

RICHARD E. TREMBLAY

14 Cross-cultural views on human development in the third millennium

266

PIERRE R. DASEN AND RAMESH C. MISHRA

15 Transitions in adolescent research

287

NANCY L. GALAMBOS AND BONNIE J. LEADBEATER

16 The impact of longitudinal studies on understanding development from young adulthood to old age

307

K. WARNER SCHAIE

17 Growing points in developmental science: A summing up

329

WILLARD W. HARTUP

Author index Subject index

345 358

Figures

4.1 An expanded conceptual framework encompassing functional isolation processes 6.1 Overview of research directions in the area of theory-of-mind development 12.1 Relations among the three constituents of human action, both when the self is the agent and when others are the agent. Means can vary from personal attributes such as effort and ability to external aids such as friends and teachers 12.2 An organismic classi®cation of coping behaviours in terms of their action-control aspects 12.3 Structural relations and decompositions among the forms and functions of aggressive behaviour 13.1 A descriptive model of aggression 13.2 Cumulative age of onset of different forms of aggression in the oldest sample of the Pittsburgh Youth Study 13.3 Male age±crime curve 13.4 Age at onset of physical aggression: attacks physically 13.5 Trajectories of physical aggression for boys age 6 to 15 16.1 A conceptual model for in¯uences affecting late-life psychological functioning in the sample case of cognition 16.2 Changes in level and rate of ageing across ®ve cohorts for the mental abilities of inductive reasoning and spatial orientation

74 108

227 228 229 245 247 248 250 252 312 321

Tables

4.1 Summary of ®ndings on nutritional de®cits and human behavioural development 10.1 The big-®ve personality dimensions, activation of problem solving, and self-regulated problem solving in four developmental domains

67 194

Contributors

Pierre R. Dasen FPSE, University of Geneva, 9 route de Drize, CH-1227 Carouge, Geneva, Switzerland Judy Dunn Institute of Psychiatry, King's College London, Social, Genetic and Developmental Psychiatry Research Centre, 111 Denmark Hill, London SE5 8AF, UK John H. Flavell Department of Psychology-Jordan Hall, Stanford University, Stanford, CA 94305-2130, USA Nancy L. Galambos Department of Psychology, Box 3050, University of Victoria, Victoria, BC, Canada, V8W 3P5 Willard W. Hartup Institute of Child Development, University of Minnesota, 51 E. River Road, Minneapolis MN 55455, USA Giyoo Hatano The University of the Air, Wakaba, Mihama-ku, Chiba, Japan 261-8586 Kayoko Inagaki School of Education, Chiba University, Yayoicho, Inageku, Chiba, Japan, 263-8522 Bonnie J. Leadbeater Department of Psychology, Box 3050, University of Victoria, Victoria, BC, Canada, V8W 3P5 Todd D. Little Schiefelbusch Institute for Life Span Studies and Department of Psychology, University of Kansas, 1052 Dole, Lawrence, KS 66045, USA Eleanor E. Maccoby Department of Psychology, Stanford University, Building 420, Stanford, CA 94305-2130, USA Edward C. Melhuish Institute for the Study of Children, Families and Social Issues, School of Psychology, Birkbeck College, Malet Street, London WC1E 7HX Ramesh C. Mishra Department of Psychology, BHU, NewG/5 Jodhpur Colony, Varanasi 221005, India

x

Contributors

Robert Plomin Social, Genetic and Developmental Psychiatry Research Centre, Institute of Psychiatry, De Crespigny Park, 111 Denmark Hill, London SE5 8AF, UK H. Rudolph Schaffer Department of Psychology, University of Strathclyde, 40 George Street, Glasgow G1 1XQ, UK K. Warner Schaie The Pennsylvania State University, 134 E. Nittany Ave., Suite 405, State College, PA 16801, USA Wolfgang Schneider Department of Psychology, University of WuÈrzburg, Wittelsbacherplatz 1, D-97074, WuÈrzburg, Germany Esther Thelen Department of Psychology, Indiana University, 1101 East Tenth Street, Bloomington, IN 47405, USA Richard E. Tremblay University of Montreal 3050, Edouard-Monpetit, 210 Montreal, Quebec, Canada H3T 1J7 Cornelis F.M. van Lieshout Department of Psychology, University of Nijmegen, PO Box 9104, 6500 HE Nijmegen, The Netherlands Theodore D. Wachs Department of Psychological Sciences, Purdue University, West Lafayette, IN 47907, USA

Preface

As the twentieth century was nearing its close, Rainer K. Silbereisen, then editor of the International Journal of Behavioral Development, planned a series of special articles focusing on the past and future of developmental science to be published during the year 2000. The idea was to combine taking stock of the most important issues in developmental science over the past century with informed speculations concerning the near future of the discipline. To this end, experts from various ®elds, older and more experienced ones as well as scientists from younger generations, were asked to write about their views concerning trends in theoretical concepts, methodologies, and, of course, substantive insights. Owing to the wealth of material available, it was decided to approach the project in two ways: First, a series of lead articles would be written by senior developmentalistsÐthe generation that had shaped the vision and produced the major results on psychological development over the second half of the twentieth century (to be edited by Willard W. Hartup). Second, these contributions would be supplemented with articles, empirically based but more speculative in nature, written by younger scientists (to be edited by Rainer K. Silbereisen). When we looked at the articles submitted, it became clear that we actually had assembled more than a series of excellent papers; we had in our hands the makings of a sourcebook comprising leading views on the past, current status, and future of many major ®elds in developmental science. Once we realised this, a long-conceived but never-realised plan occurred to us. We have both served the International Society for the Study of Behavioural Development (ISSBD) for a long time, a learned society whose explicit aim is promoting research on behaviour development in an interdisciplinary and international fashion, and for whom the International Journal of Behavioral Development (IJBD) is the main publication. One Society endeavour that has been discussed for many years is a book (or even a series) that would help researchers of all backgrounds to ®nd out as easily as possible (still at an advanced level) the essentials of developmental science. Essentials here are not seen in the sense of a set of theorems, but as a collection of exemplary topics and leading frameworks that, taken

xii

Preface

together, represent an integrative and authoritative view on human development. Consequently we invited the authors of the millennium articles submitted to the IJBD to expand on their ideas (wherever possible and necessary) and to submit their work for inclusion in this book. The 19 individuals contributing to this book were asked to write about the growing points in developmental science that they know best. Most have the bene®ts of long careers to guide them. Their topics range from early experience and development to old age, and include issues in both social and cognitive development. In each case, authors were asked to survey the history and traditions that have marked their research areas as well as current status and outlook. Although one can easily identify topics not encompassed (e.g., emotional development), the contributors display the range necessary for them to convey that developmental science is integrated in topic, ideas, methodology, and conclusions to a considerable degree. Although many developmentalists like to pretend that they have little in common with research areas other than their own, there is more sharing than one might suspect. Among these contributors are investigators whose main interests are in biological substrates of behavioural development, early experiences in terms of both basic and applied science, personality and social development, cross-cultural contexts of development, knowledge acquisition and the development of memory, and certain developmental epochsÐin this instance, adolescence and adulthood. Considering that the contributors come from eight different countries, the boundaries of this book are international rather than national. Readers will discover that we did not ask authors to write about certain issuesÐfor example, the social and historical contexts that are linked to changes in developmental ideas over the last century, or matters of the societal relevance of developmental research and applied developmental science. Complex stories can be told about these matters. Most relationships between scientists, on the one hand, and policy-makers, on the other, have been extraordinarily complex; so, too, the relationships among scientists, practitioners, and consumers. The history of research and reform in such areas as child care and education is a tangle of empirical effort, ideology, personal belief, values, and politics. Rarely has developmental science ``solved'' social problems all by itself. At the same time, there continues to be strong trust in science as a means of human betterment. Except for a view on day care not dealt with directly in this volume, these issues nevertheless motivate its publication along with the desire to describe growing points themselves in developmental science. The increasing importance of the biological and neurosciences has resulted in new challenges for the behavioural sciences, and psychology is a case in point. The new insights in the structure and functional role of the brain enable us to realise what in the past was merely good willÐresearch

Preface

xiii

on the mutual relationships between the various levels of biological activity, from the genome to behaviour. This is not to say that the more social science-oriented thinking is obsolete, but rather that a new synthesis is on the horizon. Adolescence is a case in point: Since the beginning of empirical developmental studies, many people have thought that the young show distinctive behaviours during this period, such as heightened stimulus-seeking or pronounced moodiness. Research in the 1980s modi®ed this view by demonstrating that most of the earlier studies did not refer to normal population samples, and thus overrated the prevalence of these problems. As recent research on brain changes during adolescence shows, however, neural structures particularly related to stimulus-seeking grow during adolescence, show heightened activity, and decline thereafter. Given this insight, one can now establish relations to characteristic social activities and physical movements during adolescence that can be assessed by much more ecologically valid means today than was previously possible (by systematic observation in situ based on satellite-guided positioning, for example). As mentioned earlier, the intention of the editors of this book is to provide a resource for scientists of many different generations interested in developmental science. As a rough guide to this population, consider that ISSBD has members in about 60 countries. Most reside in North America and Western Europe, but there are also substantial numbers in Asia (particularly Japan, China, and India), in Eastern Europe (for instance Poland, Hungary, the Baltic republics, and Russia), South America (especially Brazil), and Africa (unfortunately too few as yet). Developmental science is being taught, research conducted, and applications made in all of these regions. Many of these countries are ``developing'' according to United Nations criteria although, in terms of size, they actually represent the ®rst world; many have scienti®c infrastructures that are still emerging and scienti®c activity exists under the pressure of economic hardship. In these regions, the chance to read about developmental science in the leading international journals, or to contribute to these outlets, is still hampered in many ways. We hope that this source book in developmental scienceÐpast, present, and futureÐwill help readers gain quicker access to the larger research community than would be possible without it. These considerations also apply to advanced students and young investigators as well as seasoned researchers in the countries of North America and Western Europe, where developmental science has undergone exceptionally rapid growth in the last 50 years. The chapters in this book represent expert wisdom rooted in a bird's-eye view over the trends and controversies that helped shape the discipline, its contributions to science, and its application. We thank many individuals for their assistance with this volume, especially the contributors for their willingness to share their thoughts and visions. We also appreciate the support of Kenneth H. Rubin, ISSBD's

xiv

Preface

outgoing President, for his interest in the book, the help of Caroline Osborne, Lucy Farr and Imogen Burch at Psychology Press, the secretarial support of Katrin Mueller, and the conscientious attention of Verona Christmas-Best, previously the editorial assistant for the International Journal of Behavioral Development. RKS and WWH

1

Motor development as foundation and future of developmental psychology Esther Thelen

Indiana University, Bloomington, USA

Human infants are born with very little control over their bodies. Yet within a year or so, they are able to sit, stand, walk, reach, manipulate objects, feed themselves, gesture, and even speak a few words. A year later, toddlers are adept at running, climbing, scribbling, riding a tricycle, and talking in simple sentences. For parents, these new motor skills are the most dramatic and visible changes in the ®rst few years of life, frequently noted and commented upon. For those interested in studying developmental processes, this sequential unfolding of motor milestones is like a open window on change. Whereas children's mental lives must be measured indirectly, their movements are continuously observable. Each new pattern is there to see and describe. There are no hidden processes between the control of the movement and its actual execution. MOTOR DEVELOPMENT: THE GOLDEN AGE It is no surprise, therefore, that the emergence of motor skills has ®gured so prominently in the ®rst scienti®c studies of human development, truly laying the foundation of the ®eld. Rich description of infant movement dates to the last century with Darwin's (1877) well-known ``biography'' of his own child and the pioneering work of the German physiologist Preyer (1888), and the tradition was continued into the twentieth century with the narratives of Millicent Shinn (1900). The full ¯owering of the descriptive work in motor development began in the 1920s with the publication of the ®rst of Arnold Gesell's pathbreaking research and popular monographs (e.g., Gesell, 1928; Gesell & Thompson, 1934). The golden age continued through the 1930s with Mary Shirley's (1931) exquisite longitudinal descriptions of 25 infants, Myrtle McGraw's (1935) well-known and still contentious study of the twins Jimmy and Johnny, and Nancy Bayley's Berkeley Growth Study (1935). In 1946, Carmichael's Manual of child psychology contained two seminal articles by Gesell (1946) and McGraw (1946), attesting to the theoretical status of motor studies in the ®eld.

2

Thelen

There were three important and related legacies from this golden age of motor development research. The ®rst legacy from these pioneers is their theoretical contributions, and especially their strong grounding of human development in biology. The second is empirical through the introduction of detailed and rich description, novel methods for capturing human movement, and clever natural experiments. The third, and perhaps the most enduring, was the establishment of developmental norms. I discuss each of these legacies in turn.

The legacy of development grounded in biology The two most important theorists of this era were Arnold Gesell and Myrtle McGraw. Both provided us with massive descriptions of early motor development. But for both scientists, the motor catalogues were not the ends in themselves. Rather both Gesell and McGraw saw these descriptions as a way to understanding the most general, and profound, developmental principles. Gesell and the dynamic principles of growth Gesell left us the most well-articulated and comprehensive developmental theory, one that is still insightful and attractive (for many details, see Thelen & Adolph, 1992). From the start of his career, Gesell claimed deep roots in the science of biology. He viewed development as a unitary process, requiring description at many levels encompassing evolution, embryology, comparative psychology, neurophysiology, and anthropology. He was especially in¯uenced by Charles Darwin, whom he considered the founder of the scienti®c study of the child, and by G.E. Coghill, an early behavioural embryologist. Gesell attributed to Darwin his core belief that mental life is continuous with, and impelled by, the same processes that drive all organic growth. Moreover, Gesell admired Darwin's naturalistic methods, and he saw himself, too, as a naturalist, ``tirelessly'' seeking ``ideological order'' though relentless observation and comparison (Gesell, 1948, pp. 36±37). From Coghill, Gesell learned, ®rst, that development was a morphological process, that is a change in the form of behaviour. Gesell believed that, just as movements and postures provided a read-out of the nervous system, so even mental development had this morphological character, that is, it could be understood though observable behaviour. Second, Gesell was convinced that principles of growth and development illustrated by Coghill's (1969) studies of the salamander were the same for all species, including humans. ``We believe that the growth processes which mold the body and behavior of the human infant are in essence comparable with those which are being successfully analyzed by experimental embryology,'' he claimed (Gesell & Thompson, 1938, p. v). This led to Gesell's third legacy from Coghill, the view that behavioural changes followed biologically driven neural

1. Motor development

3

maturation, and not the other way around. Gesell's maturationist views stemmed directly from Coghill's discoveries that, in the salamander, changes in movement patterns emerged from neural events that happened before sensory connections were made1. Movements, therefore, were a product of autonomous neural changes, not from the sensory input. The practical importance of Gesell's theory was to act as a counterweight to the popular notions of behaviourism that were fashionable at that time. To parents who were told that infants were totally shaped by their environments, Gesell offered a different view, one of an autonomous unfolding of potential. The role of the environment was to support this unfolding, but it did not engender it. Thus, there was no point of training or teaching children until they were developmentally ready. McGraw and the biology of development Myrtle McGraw was also a developmentalist of great sophistication and subtlety who used human motor skill development as her principal empirical data. Although McGraw believed strongly in the ultimate inseparability of structure and experience, she expressly designed her study of the twins Jimmy and Johnny (McGraw, 1935) to ask which of their early motor skills could be trained and which were more ®xed by developmental design. One twin, Johnny, was exercised daily in a variety of skills, both universal (sitting, walking) and cuturally speci®c (swimming, roller-skating). She discovered that while training had some effects on the quality and initial performance of Johnny's movements, in the long run, intensive training did not make a big difference. These results were taken up by the popular press as refuting associationism and supporting the primacy of maturation. Bergenn, Dalton, and Lipsitt (1992) claim that McGraw's legacy as a maturationist oversimpli®es her more sophisticated view of development. They are probably correct, but the twin study, plus her subsequent work on the development of locomotion, put the role of maturation into the forefront.

The legacy of methodology From our technology-rich vantage point, we can only stand in awe of the detailed data so cleverly and painstakingly collected by the early movement pioneers. In addition to photographs and movies, analysed in great detail, they also recorded movement directly. For instance, Burnside (1927) and Shirley (1931) recorded infants' and toddlers' footprints by oiling the children's feet and allowing them to walk on paper. Many critical parameters

1 Bergenn et al. (1992) argue that Gesell's interpretation of Coghill is more ``maturationist'' than Coghill himself espoused.

4

Thelen

of gait can be measured by this simple technique, and indeed it is still being used today (Adolph, 1997). Another methodological advance, important in the establishment of developmental norms, was repeated testing of the same children in a large number of standardised tasks. Gesell took this testing to high art, with highly standardised equipment, structured interviews, and detailed instructions to the testers (Gesell & Thompson, 1938).

The legacy of developmental norms The legacies of theory and of measurement are closely tied to the third, and perhaps most lasting, in¯uence, that of developmental norms. Although mental testing of children dates back to the early part of the century, largely through the work of Alfred Binet, it was Gesell who brought the concept of developmental norms for infants and young children into the mainstream of developmental psychology, into the popular psyche, and into the homes of millions of parents. The test battery that Gesell and his staff perfected in the 1920s and 1930s still forms the basis of the most widely used infant tests today, the Bayley Scales of Infant Development and the Denver Developmental Screening test. Today such tests are so universally accepted that their origins are little discussed. But it was Gesell's theoretical insights and methodological rigour that led to the vast catalogue of motor milestones upon which these tests were based. Gesell, like Coghill, saw in the postural and movement forms a direct re¯ection of the internal, lawful processes of growth. The link to collecting normative data on infant movement was a direct one (Gesell & Thompson, 1938, p. 4): the underlying concepts of the normative study may be summed up as follows: Behavior grows. Growth expresses itself in ordered patterns. Behavior growth, like physical growth, is a morphosis. It is a process which produces a progressive organization of behavior forms. This morphogenesis can be investigated by morphographic methods and especially by analytic cinematography. By these methods we can ascertain the lawful sequences and norms of psychological growth for the purpose of genetic research. These norms may also be used as standards of reference for the analytic appraisal of development status. The idea that by a particular age, an infant or toddler ``should'' have achieved a particular motor milestone has not only become a standard developmental diagnosis, but it has also become completely entrenched in our cultural beliefs about child-raising. For instance, I own a used copy of Gesell's The ®rst ®ve years of life (1945), which was originally a gift from ``Mother Dingle'' to the ``Dunlaps'' with the loving inscription, ``Please read from cover to cover.'' This book is one of many written by Gesell which sets normative standards of behaviour based on his observations,

1. Motor development

5

and we can imagine the Dunlaps scrutinising their child at each age for ``typical behaviour'' in each ``typical day''. In sum, the golden age of motor development left us with a rich heritage: a deep understanding of growth and form, an appreciation for the interweaving and nonlinear course of development, a sense of our biological continuity, and exact scienti®c methods. What happened to that legacy? MOTOR DEVELOPMENT: THE DORMANT TIMES After more than two decades of extraordinary theoretical and empirical contributions to our understanding of development, the study of motor systems had declined by 1950 and then lay dormant for nearly 30 years afterward. This dramatic reversal of fortunes can be attributed both to the state of the ®eld itself and to changes occurring in psychology as a whole. As I have suggested previously, it may be that the successes of the early pioneers also contributed to the decline of the ®eld (Thelen, 1995). They produced vast and widely published catalogues of motor milestones and richly detailed descriptive studies. The norms were incorporated into developmental tests and became family lore, at least in middle-class North American families. There seemed to be little left to do. Moreover, both Gesell's and McGraw's theoretical positions appeared to lead to dead ends in terms of further empirical studies, but for different reasons. Once Gesell showed, through the descriptive topologies of behaviour, that human development obeyed universal principles, the case was effectively closed. For McGraw, the situation was more complex. She admitted in her 1962 foreword to The neuromuscular maturation of the human infant that she did not achieve her goal of relating the development of function to the maturation of structure. Part of the problem was that methods for studying changes in the brain were inadequate. But she also recognised that the earlier theoretical formulations were too simplistic, especially the division of behaviour into cortical and subcortical in¯uences, and the neat separation of ``instincts'' vs. ``acquired traits'' or ``maturation'' opposed to ``learning'' (McGraw, 1945/1973). The decline of motor development must also be understood in light of the other forces in the ®eld during the 1950s, 60s, and 70s. During this time, learning theorists and experimental psychologists dominated academic departments of psychology. Descriptive and normative data, no matter how detailed, could not be compared in apparent ``scienti®c rigour'' to the tightly controlled experimental methods practised by these disciplines. At the same time, the focus in developmental psychology shifted dramatically. The work of Jean Piaget became increasingly well known and highly in¯uential, igniting fervent interest in the inner mental life of children (as opposed to the form of their overt behaviour). Piaget offered descriptive stages, but he also inspired brilliant experimentation, something

6

Thelen

that the motor development theorists did not do. The lesson may be that even the most elegant theory survives only as it generates new and copious empirical research. MOTOR DEVELOPMENT: TWO DECADES OF REBIRTH Beginning in about 1980, the tide began to turn again, and interest in movement gradually gained momentum. Just as multiple factors contributed to its decline, the ®eld's revitalisation has come from a number of converging in¯uences. These included several theoretical advances: new ideas in movement science and biomechanics, insights from ecological psychology, and the import of dynamic systems theory. But the ®eld has also bene®ted greatly from a new understanding of the plasticity of the brain, and from technological advances in recording movement and brain activity.

The importance of N. Bernstein The legacy from the neuroembryologists, through Gesell and McGraw, was that movement was a direct read-out of the maturational status of the nervous system. The reverse was scarcely consideredÐthat the developmental course of the nervous system may be moulded by the nature of the body and how it moves. Bernstein's (1967) seminal insight was to pose the control issue differently. He started with the body, which, he noted, had hundreds of bones and joints and millions of muscle ®bres. Yet every movement is a coherent, coordinated event. How can the brain accomplish this feat of coordination, given so many possible combinations, or degrees of freedom? The answer, according to Bernstein, was that movements were organised in synergies, that is, a functional linking together of muscles into ensembles that worked together. What the brain recruited, according to Bernstein, was not individual muscles, but an appropriate pattern to accomplish a functional task. This vastly simpli®ed the control problem. Indeed, Bernstein showed that movement was function speci®c and not muscle speci®c: his classic example is that you can sign your name using a pen on paper or using a broomstick on a blackboard, but the signature remains the same. Moreover, as the motor system is assembled for functional action, it actually exploits the mechanical properties of the limbs and body. For example, limbs have spring-like properties because of the elastic qualities of the muscles and the anatomical con®guration of the joints. When an ordinary spring is stretched and let go, it oscillates in a regular trajectory. The movement pattern need not be con®gured explicitly because it arises from the natural properties of the spring. Likewise, many aspects of human movements need not be detailed in the nervous system because they arise by

1. Motor development

7

themselves from the natural properties of the body. One good example is walking: as people step on one leg and then shift their weight forward over it, the back leg is stretched and stores energy, like a spring. As that leg then swings forward, it uses little active muscle contraction, but rather relies on the potential energy gained by stretching the muscles. In his writing on the development of movement, Bernstein turned the old theories on their heads. It was not, he claimed, so much the nervous system instructing the muscles, as the dynamics of the movement instructing the nervous system. Children must learn the biodynamics of their bodies, the changing forces that produce and accompany each movement in each situation. To learn to walk, for instance, infants must deal with the complex interactions between the movements of the legs, the centre of gravity of the body, and the support surface. These cannot be instructed beforehand, but must be individually assembled through experience. Eventually, infants learn not just to control their movements, but to make them ef®cient by exploiting their biomechanics, that is, using what the system can provide ``for free''. The impact of Bernstein on the study of motor development was to shift the focus from thinking exclusively of the central nervous system as the sole contributor to the emergence of new skills to considering the contributions of the biomechanics of the moving limbs. In my early work on the coordination and control of spontaneous leg movements in infants, for example, I came across two puzzling results. The ®rst involved the kinematic (timespace) organisation of the common kicking movements seen throughout the ®rst 6 months. These movements are well-coordinated, displaying a nearly simultaneous ¯exion and extension of the hips, knees, and ankles. Moreover, the durations of the parts of the moments were not random, but showed structure as well (Thelen & Fisher, 1983). How did this structure arise? Surprisingly, when my colleagues and I looked at the patterns of muscle activity that produced kicking movements, we found that the underlying electromyographic activity was far less patterned than the resulting movement. Indeed, infants appeared to just contract all the ¯exors and extensors at the beginning of the ¯exion phase of the kick and use very little muscle activation thereafter. Extension was largely passive. This meant that the extension was not directly programmed by the central nervous system, but rather emerged as a consequence of the elastic properties of the legs and their stored energy when ¯exed. In short, the precise patterns of these movements came about from the interplay between the neural command and the peripheral properties of the body. The second puzzle involved the well-known newborn ``stepping'' re¯ex. Newborn infants will perform alternating step-like movements when held upright with their feet on a surface. Within a few months, such movements can no longer be elicited. Since the time of Preyer (1888), the common explanation has been that the patterns are a primitive re¯ex, which becomes inhibited as the cortex, and voluntary movement, matures. However,

8

Thelen

although stepping while upright does disappear, kinematically identical kicking movements performed while infants are supine do not, and indeed increase in frequency. To account for this strange disparityÐmovements that are a function of body postureÐwe looked again to the peripheral contribution. Here again, we discovered that there was a relation between infants' abilities to step and the mass of their legs. As their legs became heavier through the deposition of subcutaneous fat in the ®rst few months, their ability to lift them in the biomechanically demanding upright posture decreased. Simply laying infants down, or reducing the effective mass by holding them in a tank of water, restored the movement (Thelen & Fisher, 1982). Again, performance depended on all elements of the moving system. Because body masses, lengths, centres of inertia, and so on are continually changing as infants grow, and because new postures bring on new biomechanical challenges, skill acquisition is a continually interactive process. Infants must discover how to produce the appropriate coordinative pattern and modulate it to ®t the task. Furthermore, the addition of each new skill opens different opportunities for these interactions to occur. Development progresses through each new achievement, setting the next set of challenges. Thus, Bernstein's impact was to reintroduce the child as an active movement problem-solver, much as Piaget had replaced the child as passive stimulus-responder with one who actively seeks stimuli in the world.

The importance of ecological psychology Bernstein's message of the child as an active explorer meshed beautifully with the second powerful in¯uence on the renaissance of motor development research. This was the theoretical approach loosely known as ``ecological psychology'', based largely on the work of psychologists Eleanor Gibson (E.J. Gibson, 1969, 1988) and James Gibson (J.J. Gibson, 1966). The basic assumption of Gibsonian psychology is that people, and other animals, are able to directly perceive structured information in the environment that enables them to act functionally within it. (This is in contrast, of course, to the view that the environment does not have meaning until it is reconstructed within the brain.) The goal of development, in the Gibsonian view, therefore, is for infants and children to discover progressively the affordances for action in the environment, a process of matching the abilities of the actors with the opportunities in the world around them. Ecological psychology has shaped the ®eld of contemporary motor development studies in several important ways. First and foremost is the notion that perception and action are inseparable in the formation of skills. Perception is essential for movement, but movement also informs perception. Movements of the head and eyes, for instance, enable the perceiver to sample the visual array. Movements of the arms, hands, and ®ngers are necessary for haptic exploration of new objects and surfaces. Locomotion is essential for understanding the quality of surfaces and the layout of the

1. Motor development

9

spatial surroundings. Indeed, we can even cast movement as a form of perception, a way of knowing the world by moving in it. Second, research in the ecological tradition has made researchers aware that perception in the service of action is always multimodal, probably right from birth. For example, a study by Rochat and Morgan (1995) has shown that very young infants are aware of the correspondences between their visual perception of their moving limbs and their proprioceptive and haptic senses of them. In these studies, infants were shown two side-by-side televised displays of their own legs, clothed in distinctively striped stockings. On one display, the video image was concordant with the infants' views of their legs as they sat in the infant seat. On the second, the legs were reversed. Infants showed by their preferential looking that they could distinguish the two displays. Clearly, they must have, through experience looking and moving, mapped the correspondences between what they saw and what they felt. The third lasting contribution from ecological psychology to current motor development research is the emphasis on exploration as an important force for developmental change. E.J. Gibson (1988), for instance, describes three overlapping phases of exploration in the ®rst year. At ®rst, infants expore events with vision and hearing. Next, they explore objects with reaching, grasping, and mouthing, and, later, they explore the large layout with self-produced locomotion. Changing motor skills contribute to infants' exploratory behaviour, but the behaviour itself also leads to new motor skills.

The introduction of dynamic systems Both Bernstein and the Gibsons were concerned with a similar issue: How to avoid the ``homunculus'' problemÐsome entity in the head of the actor that represents the world and makes decisions to act in it. For Gibson, the solution was direct perception. For Bernstein, the solution was reducing the executive decisions by the synergistic organisation of the brain and body. In the 1980s, a group of movement theorists centred around the Haskins Laboratories in New Haven, Connecticut, merged these two theoretical traditions with recent advances in the physics and thermodynamics of complex systems to produce a radical change in the theory of motor control and development. In two landmark papers, Peter Kugler, Scott Kelso, and Michael Turvey (Kelso, Holt, Kugler, & Turvey, 1980; Kugler, Kelso, & Turvey, 1980) considered Bernstein's synergies in a new light, as self-organising systems, and described by the same dynamic principles that governed complex, so-called dissipative structures, in physics and chemistry. (Dissipative systems absorb energy to maintain themselves in organised states far from thermodynamic equilibrium. All biological systems are dissipative, but so are some other natural systems such as cloud formations or ¯uid ¯ows.)

10

Thelen

The key insight here was that when a person assembles a motor synergy to do some task, the participating components cohere and produce patterns that have temporal and spatial organisation that is not the result solely of the detailed instructions from the nervous system. These investigators, and subsequently many others, demonstrated that simple, cyclical movements of the limbs in humans could be described by the mathematics used for coupled oscillators in general. In particular, they discovered that when people move their body parts in a rhythmical fashion, they have preferred coordination modes: spatial and temporal patterns that are comfortable and easily performed. Other possible patterns are unstable, such that the preferred mode seems to suck them in, or ``attract'' them. Under certain conditions, people may spontaneously shift coordination modes from one preferred regime to another. Again, using locomotion as an example, quadrupeds use different gaits depending on the speeds of their movements. The gaits performed are those that are the most energy-ef®cient for the particular speed (see Kelso, 1995). There were several important implications of dynamic systems ideas for conceptualising motor development, and for developmental theory in general. First is the restatement of Bernstein's ideas that every movement is a system-wide ensemble of all participating components, assembled in the context of a particular task at hand. Kugler and Turvey (1987) used the terms ``soft-assembly'' to describe this ¯exibility: Behaviour patterns are not prescribed, although some may be preferred. This meant that behaviour was not ``hard-wired'' into the brain, but emerges ``on-line'' in light of the person's available structure and energetic resources, and the nature of task to be done. Second, dynamic systems theory emphasised not only patterns in space, but also that behaviour has pattern over time as well. This is important because it changes the focus from taking a ``snapshot'' of behaviour at a particular age or skill level, to more serious considerations of how change occurs over time. Here, the time scale could be seconds, or minutes, or weeks, or months. Indeed, from a dynamic perspective, it makes little sense to consider changes at different time scales as different processes. What happens at an action time scale cascades into changes over the time scales of learning and development. Finally there is the idea that the coordinative state of a stable movement patternÐfor instance, walking or reachingÐbehaves like a dynamic attractor with varying levels of stability. For a pattern to change, something must disrupt the stability of the old pattern, so that the components can coalesce into new forms. In terms of development, this meant characterising behaviour in terms of its stability to identify the transitions into new forms and test the mechanisms that engender them (see Thelen & Smith, 1994; Thelen & Ulrich, 1991). Systems ideas are not new in developmental psychology. They have been proposed by Werner, Lewin, Piaget, and many others. For example, Piaget

1. Motor development

11

(1952) believed that new levels of mental equilibration could only be achieved through disequilibration of current stages, similar to the phase shifts described by contemporary dynamicists, and he was profoundly concerned with processes of change. Current dynamic systems theories have gone beyond the old formulations, however, in two important ways. First, we have achieved some level of success in re®ning the precision of our ideas through formal mathematical models (see Newell & Molenaar, 1998; Thelen, SchoÈner, Scheier, & Smith, 2001). But equally important, dynamic thinking has inspired a renewed interest in empirical work that is closer to Piaget's original agenda of looking at processes of change. (In the postPiagetian era, studies were more concerned with describing age-de®ned performance stages.) Process accounts often involve detailed longitudinal studies to identify times of transition combined with microgenetic methods that test possible mechanisms that move the child into new developmental phases. Finally, dynamic systems have also inspired studies that incorporate multiple levels of analysis from the cognitive to the biomechanical. For instance, I have used dynamic systems principles in my own work to uncover the multiple in¯uences on change in locomotor-movement (Thelen & Ulrich, 1991) and reaching (Thelen et al., 1993; Thelen, Corbetta, & Spencer, 1996).

Plasticity in the central nervous system There have been two other in¯uences on research in motor development that bear mention. First, just as our predecessors over 50 years ago were inspired by the current ®ndings in neurophysiology, so too has contemporary thinking been shaped by remarkable discoveries in neuroscience. First is a growing understanding of the systems-wide properties of the brain. Although much research is dedicated to pinpointing local areas that subserve different functions, it has also been discovered that no area works in isolation. Indeed, neural networks supporting perceptual, motor, and cognitive processes are widely and densely interconnected (see, for instance, Edelman, 1987). For instance, neurons responding to both spatial localisation of visual targets and to intended movements are found in many areas of the cortex. At the same time, a single neuron may be activated by the visual, planning, memory, and movement aspects of a task (reviewed in Thelen et al., 2001). These ®ndings reinforce the idea that perception and action, and its cognitive counterparts, are part of the same continuous and coupled process. Second, neuroscientists have discovered remarkable, dynamic plasticity in even the adult brain. By careful brain mapping in monkeys, they have established that experiences in the world both establish and maintain the functional connectivity of both cortical and subcortical areas. Old ideas about the ®xity of the adult brain have been overthrown (see, for example, Merzenich, Allard, & Jenkins, 1990). The implication for development is

12

Thelen

profound: Experience moulds the brain. What may previously have been considered as autonomous maturational changes in brain function may indeed be driven by children's everyday actions in the world. But the loop is still closed: Just as experience reorganises the brain, so also the resulting improvements in perceptual discrimination, memory, and motor control provide children with new opportunities for experience to further remap the brain. A ®nal major in¯uence on the ®eld has been the important theoretical work of Gerald Edelman, which is a synthesis of current neuroembryology, neurophysiology, and behavioural development consistent with Bernstein, Gibson, and dynamic systems. In his ``Theory of neuronal group selection'', Edelman (1987) proposes that adaptive behaviour emerges as the recurrent perceiving and acting in the world strengthens particular neural networks such that patterns are progressively selected from many, wider possibilities. In his view, genetic and neuroembryonic processes provide the rough outline of the neural anatomy. The functional mapping of the brain is subsequently experience-dependent, especially through perceptual-motor exploration.

Technological contributions to motor development New methods of video and computer-assisted movement analyses have contributed greatly to advances in our understanding of motor development, perhaps more than in any other area of developmental psychology. We now have the ability to trace the details of movements in time and space rapidly and accurately, providing a real-time read-out of the development of coordination and control. Sensitive force platforms record changes in posture in sitting, standing, and walking. Electromyographic recording provides detailed pictures of muscle activation patterns. These advances open new windows on how behaviour is produced and how it changes. It seems that our vision of what questions can be asked is, at least in part, tied to the means we have to answer them.

MAJOR THEMES IN CONTEMPORARY MOTOR DEVELOPMENT

Descriptions of the development of motor skill: Understanding coordination and control We still carry on the heritage of McGraw and Gesell by using detailed, longitudinal studies as the foundation for understanding motor skill development. Like our predecessors, we also use these studies to infer developmental changes in the underlying mechanisms. Moreover, from the point of view of dynamic systems, such studies are essential for providing

1. Motor development

13

the ``landscape'' of behavioural patterns: when they are stable and when they change. This is the ®rst step for identifying points of transition, where the system may be probed by experiments. Thus, longitudinal studies are further supplemented by experiments to test hypotheses about coordination and control. Notable examples of contemporary descriptive work are the pioneering studies of Claes von Hofsten on infant reaching (1979, 1982, 1984). Von Hofsten was the ®rst to reintroduce detailed kinematic measures of infant movement, after Halverson's (1931) efforts 40 years earlier. Especially compelling were von Hofsten's (1980) demonstrations of young infants' catching moving objects, a seemingly remarkable precocious ability. Von Hofsten's work has been extended by other longitudinal studies of reaching using even more dense sampling techniques and adding kinetic and electromyographic measure (Spencer & Thelen, 2000; Thelen et al., 1993, 1996), and following infants until they were several years old (Konczak, Borutta, & Dichgans, 1997; Konczak, Borutta, Topka, & Dichgans, 1995). Lower limb movements have also received attention. For instance, my colleagues and I were the ®rst to describe infants' spontaneous leg movements using kinematic techniques. Previously, infants' leg movements, because they were not apparently goal-directed, were thought to be disorganised or random. However, we demonstrated a high degree of coordination both within and between limbs and a pattern of developmental changes in that organisation. We followed this with a more in-depth look at the kinetics of infant leg movements, that is, how infants managed the forces that produce movements (Jensen, Ulrich, Thelen, Schneider, & Zernicke, 1994; Schneider, Zernicke, Ulrich, Jenzen, & Thelen, 1990). As in motor development's ``golden age'', the development of upright locomotion has been a primary focus in contemporary studies. Learning to walk is a dramatic developmental milestone, the transition from infancy to childhood. Moreover, there has been increasing recognition of the complexity of the task, and thus, in the question of how infants solve the problems involved (Thelen, 1984). For example, Sutherland (1984) reported gait measures on children up to 7 years of age, documenting kinematic changes in step parameters and joint excursions, as well as changes in muscle patterns. Bril and Breniere (1992; Breniere, Bril, & Fontaine, 1989) focused more directly on changes in newly walking infants, using a large force plate to provide detailed descriptions of weight shifts and propulsive forces. Clark and Phillips (1993) looked primarily at changes in inter-limb coordination over a similar time period. Moreover, there have been two recent longitudinal studies of inter-limb patterns in crawling, another topic well researched by earlier investigators (Adolph, Vereijken, & Denny, 1998; Freedland & Berntenthal, 1992). Here I do not dwell on the detailed results of these investigations. Rather, I suggest that taken together, these contemporary studies of reaching and walking have gone beyond the classic studies in two important ways,

14

Thelen

re¯ecting the in¯uence of Bernstein, Edelman, and dynamic systems. First, there is more explicit consideration of the biomechanical aspects of early movement. Theoretically, this means thinking about movement as a biomechanical problem to be solved by the nervous system: We cannot think about neural control distanced from what is controlled. For instance, learning to walk requires keeping the centre of mass of the body over a permissible base of support and controlling the ``fall'' as the infant steps forward. Crawling requires the correct limb combinations to maintain a dynamic base of support. Reaching involves stabilising one arm segment against the forces generated by the other moving segments. In this view, the patterns of muscle activation may well be the result of the biomechanical demands of the movement, not the cause of the movement. For example, Angulo-Kinzler, Ulrich, Chapman, and Thelen (2000) discovered that, as Bernstein predicted, the patterns of muscle activation underlying treadmill stepping, supported, and independent walking in infants were much more variable than the patterns of forces that moved the legs. And, second, contemporary researchers are considering new ideas about variation and individual differences. From the start, researchers in motor development have noticed, measured, and discussed individual differences. Shirley (1931), for instance, documented the differences in onsets of various motor skills in her 25 babies and related these differences to the infants' physical growth, muscle tone, and ``willingness to expend energy'' (p. 125). Physical dimensions and movement ``styles'' are still seen as an important part of the story (Adolph, 1997; Thelen et al., 1993). Additionally, today's researchers are less concerned with variability around age norms than in earlier times. Indeed, in many studies, children are compared on the basis of their skill levels rather than their ages. Experimental studies of early skills One important way to uncover processes of change is to experimentally manipulate variables to which the system is sensitive. For example, I mentioned previously my hypothesis that infants' limb masses were important in the disappearence of the stepping re¯ex. To test this, we changed the limb mass by submerging the infants' legs in water or adding weights (Thelen, Fisher, & Ridley-Johnson, 1984). More recently, the effects of biomechanics and movement have been evaluated by changing infants' postures (Jensen et al., 1994; Savelsbergh & van der Kamp, 1993) or by adding weights to their limbs and torsos (Adolph & Avolio, 2000; Thelen, Skala, & Kelso, 1987). To study the development of postural control, for example, researchers have commonly intervened experimentally by placing infants and children on a platform that perturbs their balance. By this means, researchers can test the limits of children's postural stability at various ages, and investigate the underlying neuromotor mechanisms that produce the response (Woollacott & Sveistrup, 1992).

1. Motor development

15

Studies in perception±action coupling Although all motor development studies are perception±action studies, experiments explicitly in the Gibsonian tradition continue to dominate the ®eld. Here the issue is not so much the motor control per se, but how action is modulated by perception, and in turn, how action informs perception. For example, early work by Lee and Aronson (1974) and Butterworth and Hicks (1977) established the dominance of vision in toddlers' postural control by using the famous ``moving room'' paradigm. This situation was later extended and re®ned by Bertenthal and Bai (1989) with younger infants and also used by Stoffregen, Schmuckler, and Gibson (1987) to show that infants were sensitive to peripheral visual ¯ow. Since then, there has been considerable debate over the relative importance of vision and proprioception in the control of posture. In their review, Bertenthal and Clifton (1998) concluded that the balance between multiple sources of information may be contextually determined, much like the soft-assembly of action I described earlier. The role of perception in prehension has also been a topic of intense study. Early researchers, including Piaget (1952), Bruner (1973), and White, Castle, and Held (1964) stressed the necessity of infants gradually learning to match the sight of their hands with the sight of the objects to be reached. Questions about this gradual visual±visual matching arose with von Hofsten's (1982) report of infants' directed reaches during the newborn period, before a long period of learning commenced. Recently, Clifton, Muir, Ashmead, and Clarkson (1993) further challenged the visual matching idea by showing that, at ®rst, infants reached as well in the dark to a lighted or sounding object as when they could see their hands. This work underscored the importance of learning the ``feel'' as well as the sight of the arm and hand. The studies I have mentioned so far have looked at how infants use perception to adjust movements. But the converse is also true: Children use actions to inform perception, primarily through the use of exploratory movements (E.J. Gibson, 1988). For instance, when confronted with an unfamiliar surface, infants, unsure of the suitability of the surface for locomotion, will use their hands to touch and pat (Adolph, 1997; E.J. Gibson et al., 1987 ). Bushnell and Boudreau (1993) provide an excellent illustration. They show how infants' perceptual detection of the properties of objects, such as weight, texture, or sounding abilities, develops only as their motor abilities are suf®cient to manipulate the objects appropriately. In sum, we have learned a great deal about development in general through the experimental study of perception and action. In particular, this work has emphasised the continual active role of children in exploring their environments, and the context-dependent, problem-solving process that constitutes developmental change. Many of the old ideas about developmental timetables can now be recast in different terms. Instead of a

16

Thelen

phylogenetically determined sequence of stages, development is better conceptualised as a changing landscape of patterns, whose stability depends not only on the organic status of the child, but also on its experiential history, and how those interact with the particular task at hand. Cognition and motor skill development A ®nal area gaining in prominence is the intersection between perception± action and cognition. Views on the relationship between these two domains have been complex. For many years, motor skill and cognition were believed to be unrelated because early studies showed only modest correlation, if any, between children's motor and intellectual development (e.g., Shirley, 1931). Today, there is little interest in using motor development to predict later mental status, but there is increasing agreement with Piaget over the tight linkage between movement and cognitive development. The work of Bertenthal and Campos and their colleagues has been especially in¯uential in this regard (e.g. Bertenthal & Campos, 1990; Bertenthal, Campos, & Barrett, 1984; Campos et al., 2000). These scholars argue that one setting eventÐthe onset of crawlingÐinitiates a developmental cascade that has consequences for changes in spatial cognition and emotional development. The mechanisms by which being able to move about changes the ways babies think is not fully understood, but may involve their increased attention to perceptual information as they move about. There is, perhaps, an even more basic way in which movement and cognition are tightly linked, a way close to what Piaget envisioned (Thelen, 2000). Infants, children, and adults are perceiving and moving all of their waking hours. Movement itself is a form of perception because the proprioceptive and haptic senses are continuously receiving information, information that is perfectly coupled with information from the external senses such as vision and hearing. Thus, movement is an integral part of the ensemble of all our experience, including the times when we are just looking at something, because looking involves movements of the eyes, head, and neck. If, in the Piagetian sense, higher cognition is built from sensorimotor experiences, then the movement that occurs with those experiences is remembered and recalled to the same degree as information from the other perceptual senses. AN AGENDA FOR THE FUTURE: IT'S NOT JUST MOTOR DEVELOPMENT ANY MORE Today, motor development is a robust ®eld, with strong theoretical bases and empirical work of great sophistication. The continued strength of the ®eld in the future, I believe, lies both in our abilities to pursue important issues within the ®eld and, at the same time, to tie motor development with

1. Motor development

17

advances in other, related disciplines. In this concluding section, I offer some thoughts about these future directions.

Multimodal perception and action We have made great strides in understanding the role of visual perception in the development of reaching, posture, and locomotion. But, as previously discussed, experience is continually multimodal, including the perception of movement. Much less is known about how infants and children use correlated information from multiple sources, especially in decisions to act (but see, for instance, Streri & PeÃcheux, 1986).

Formal models and robotics The processes involved in motor development are excellent candidates for a variety of types of computational and dynamic models, as well as implementation in robots. Modelling of any type offers the opportunity to think more precisely about the phenomenon in question, and to generate testable hypotheses about the processes involved. There is already considerable progress towards this end (see, for instance, Newell & Molenaar, 1998). Examples inspired explicitly by dynamic systems theory include Robertson's (1993; Robertson, Cohen, & Mayer-Kress, 1993) dynamic analysis of the time structure of foetal and infant movements; Bertenthal's (Bertenthal, Boker, & Rose, 1995) dynamic analysis of infant postural control; Gold®eld, Kay, and Watten's (1993) work on infants learning the Jolly Jumper; Fitzpatrick, Schmidt, and Lockman's (1996) analysis of children learning to clap; and Thelen et al.'s (2001) dynamic ®eld model of infant perseverative reaching.

Embodied cognition There is a great need for studying the role of movement in so-called ``higher'' cognitionÐmemory, decision-making, categorisation, and language. For example, Thelen, Smith and colleagues have demonstrated the role of body memory in a classic Piagetian task, the A-not-B error (Smith, Thelen, Titzer, & McLin, 1999; Thelen et al., 2001). They showed that when infants reach several times to one of two targets, they build up a location memory of the target that also includes the feel of the arms and infants' postural set, and that this memory in¯uences further decisions to reach. This is a clear demonstration that movement is not separate from remembering and deciding, which are traditionally considered ``cognitive'' processes. These authors suggest that movement must be considered as part of every task: What aspects of movements that accompany everyday actions are remembered and encoded as part of the task ensemble?

18

Thelen

Neural bases of motor skill development The ®eld of ``developmental cognitive neuroscience'' is just coming into its own (e.g., Nelson & Luciana, 2001). Current neuroimaging and direct recording techniques are not well adapted to studying large movements in normal young human subjects. However, scientists have used other approaches to better understand brain correlates of skill development. For example, several research programmes are following groups of infants, largely prematurely born, who have suffered well-characterised, perinatal brain lesions. Many show considerable recovery of function while others do not attain fully functional outcomes (see review in Elman et al., 1996). Such studies raise profound issues both about early plasticity and the effects of experience, but also about the old ideas of localisation of function, since it is apparent that when there is injury to one part of the brain, other areas can assume needed functions. It is hoped that, as collaborations between neuroscientists and developmentalists increase, and with inevitable technical advances, this exciting area of the brain±behaviour interface will grow at a rapid pace.

Learning and plasticity The explosion of interest in learning and plasticity in the neurosciences has already reverberated in our ®eld. There is already a large and detailed literature devoted to motor learning and plasticity in human adults and nonhuman primates. Unfortunately, as yet, few of these ®ndings have been extended to infants and children, although they are highly relevant. For example, Shadmehr and colleagues (e.g. Shadmehr & Mussa-Ivaldi, 1994) have conducted elegant studies examining how adults learn novel motor tasks, what is remembered from these newly learned tasks, and then how the newly learned task competes with previously learned ones. This is a good model for learning new skills as infants and children, yet much work needs to be done (but see Adolph, 1997, for a good example of competing skills.)

Cultural and individual differences Individual and cultural differences in learning motor skills are as important as the commonalities that children share. Such differences inform us about the plasticity of developmental pathways, and the in¯uences and limits of daily experience in shaping them. All cultures and all intact individuals learn to sit, walk, reach and manipulate, and speak, but what are the diverse means of attaining similar ends? This remains an understudied area of motor development, probably due to the dif®culties of collecting crosscultural data, but one with great potential for helping us understand deep developmental issues.

1. Motor development

19

CONCLUSION The study of motor development is alive and vigorous at the turn of the century, perhaps reviving and reliving its old golden age. It has again returned to the mainstream of developmental psychology, but the ®eld also has extensions, as I have shown, into new ®elds, including neuroscience, cognitive science, and motor science. Moreover, the theoretical and empirical work of the last two decades is reshaping clinical practice dealing with perceptual-motor disorders in infants and children. I believe, as in other ®elds of psychology, real progress will be made in the next millennium only as we continue to combine our traditional naturalistic and experimental approaches with insights from other disciplines. This will work in two ways. First, we will learn more about the development of perception and action as we bring in information and techniques from neuroscience, cognitive science, clinical practice and the like. But I also think we will take leadership in the twenty-®rst century as we continue to show the centrality of movement in other domains of psychological interest. Finally, we offer models of how development can truly be studied as a time-dependent process. REFERENCES Adolph, K.E. (1997). Learning in the development of infant locomotion. Monographs of the Society for Research in Child Development, 62 (3, Serial No. 251). Adolph, K.E., & Avolio, A.M. (2000). Walking infants adapt locomotion to changing body dimensions. Journal of Experimental Psychology: Human Perception and Performance, 26, 1148±1166. Adolph, K.E., Vereijken, B., & Denny, M.A. (1998). Roles of variability and experience in development of crawling. Child Development, 69, 1299±1312. Angulo-Kinzler, R.M., Ulrich, B.D., Chapman, D., & Thelen, E. (2000). Context and control in the step patterns of newly walking infants. Manuscript submitted for publication. Bayley, N. (1935). The development of motor abilities during the ®rst three years: A study of sixty-one infants tested repeatedly. Monographs of the Society for Research in Child Development, 1, 26±61. Bergenn, V.W., Dalton, T.C., & Lipsitt, L.P. (1992). Myrtle B. McGraw: A growth scientist. Developmental Psychology, 28, 381±395. Bernstein, N. (1967). The coordination and regulation of movements. Oxford: Pergamon. Bertenthal, B.I., & Bai, D.L. (1989). Infants' sensitivity to optical ¯ow for controlling posture. Developmental Psychology, 25, 936±945. Bertenthal, B.I., Boker, S.M., & Rose, J.L. (1995). Dynamical analyses of postural development. Journal of Sport and Exercise Psychology, 17, 8. Bertenthal, B.I., & Campos, J.J. (1990). A systems approach to the organizing effects of self-produced locomotion during infancy. In C. Rovee-Collier & L.P.

20

Thelen

Lipsitt (Eds.), Advances in infancy research, Vol. 6 (pp. 1±60). Norwood, NJ: Ablex. Bertenthal, B.I., Campos, J.J., & Barrett, K.C. (1984). Self-produced locomotion: An organizer of emotional, cognitive, and social development in infancy. In R. Emde & R. Harmon (Eds.), Continuities and discontinuities in development (pp. 175±210). New York: Plenum. Berthenthal, B.I., & Clifton, R.K. (1998). Perception and action. In W. Damon (Ed.), Handbook of child psychology: Vol 2. Cognition, perception and language (pp. 51±102). New York: John Wiley. Breniere, Y., Bril, B., & Fontaine, R. (1989). Analysis of the transitions from upright stance to steady state locomotion in children with under 200 days of autonomous walking. Journal of Motor Behavior, 21, 20±37. Bril, B., & Breniere, Y. (1992). Postural requirements and progression velocity in young walkers. Journal of Motor Behavior, 24, 105±116. Bruner, J.S. (1973). Organization of early skilled action. Child Development, 44, 1±11. Burnside, L.H. (1927). Coordination in the locomotion of infants. Genetic Psychology Monographs, 2, 279±372. Bushnell, E.W., & Boudreau, J.P. (1993). Motor development and the mind: The potential role of motor abilities as a determinant of aspects of perceptual development. Child Development, 64, 1005±1021. Butterworth, G., & Hicks, L. (1977). Visual proprioception and postural stability in infancy: A developmental study. Perception, 6, 255±262. Campos, J.J., Anderson, D.I., Barbu-Roth, M.A., Hubbard, E.M., Hertenstein, M.J., & Witherington, D. (2000). Travel broadens the mind. Infancy, 1, 149±219. Carmichael, L. (Ed.). (1946). Manual of child psychology. New York: John Wiley. Clark, J.E., & Phillips, S.J. (1993). A longitudinal study of interlimb coordination in the ®rst year of independent walking: A dynamical systems analysis. Child Development, 64, 1143±1157. Clifton, R.K., Muir, D., Ashmead, D.H., & Clarkson, M.G. (1993). Is visually guided reaching in early infancy a myth? Child Development, 64, 1099±1110. Coghill, G.E. (1969). Anatomy and the problem of behavior. New York: Macmillan. (Original work published in 1929.) Darwin, C. (1877). Biographical sketch of an infant. Mind, 2, 285±294. Edelman, G.M. (1987). Neural Darwinism: The theory of neuronal group selection. New York: Basic Books. Elman, J.L., Bates, E.A., Johnson, M.H., Karmiloff-Smith, A., Parisi, D., & Plunkett, K. (1996). Rethinking innateness: A connectionist perspective on development. Cambridge, MA: MIT Press. Fitzpatrick, P., Schmidt, R.C., & Lockman, J.J. (1996). Dynamical patterns in the development of clapping. Child Development, 67, 2691±2708. Freedland, R.L., & Bertenthal, B.I. (1992). Kinematic analyses of the development of creeping in human infants. Infant Behavior and Development, 15, 300. Gesell, A. (1928). Infancy and human growth. New York: Macmillan. Gesell, A., (1940). The ®rst ®ve years of life. New York: Harper. Gesell, A. (1946). The ontogenesis of infant behavior. In L. Carmichael (Ed.), Manual of child psychology (pp. 295±331). New York: John Wiley. Gesell, A. (1948). Studies in child development. Westport, CT: Greenwood Press.

1. Motor development

21

Gesell, A., & Thompson, H. (1934). Infant behavior: Its genesis and growth. New York: McGraw-Hill. Gesell, A., & Thompson, H. (1938). The psychology of early growth including norms of behavior and a method of genetic analysis. New York: Macmillan. Gibson, E.J. (1969). Principles of perceptual learning and development. Englewood Cliffs, NJ: Prentice Hall. Gibson, E.J. (1988). Exploratory behavior in the development of perceiving, acting, and the acquiring of knowledge. Annual Review of Psychology, 39, 1±41. Gibson, E.J., Ricco, G., Schmuckler, M.A., Stoffregen, T.A., Rosenberg, D., & Taormina, J. (1987). Detection of the traversability of surfaces by crawling and walking infants. Journal of Experimental Psychology: Human Perception and Performance, 13, 533±544. Gibson, J.J. (1966). The senses considered as perceptual systems. Boston: Houghton Mif¯in. Gold®eld, E.C., Kay, B.A., & Warren, W.H. (1993). Infant bouncing: The assembly and tuning of action systems. Child Development, 64, 1128±1142. Halverson, H.M. (1931). An experimental study of prehension in infants by means of systematic cinema records. Genetic Psychology Monographs, 10, 107±286. Jensen, J.L., Ulrich, B.D., Thelen, E., Schneider, K., & Zernicke, R.R. (1994). Adaptive dynamics of the leg movement patterns of human infants: I. The effects of posture on spontaneous kicking. Journal of Motor Behavior, 26, 303±312. Kelso, J.A.S. (1995). Dynamic patterns: The self-organization of brain and behavior. Cambridge, MA: MIT Press. Kelso, J.A.S., Holt, K.G., Kugler, P.N. , & Turvey, M.T. (1980). On the concept of coordinative structures as dissipative structures: I. Empirical lines of convergence. In G. Stelmach & J. Requin (Eds.), Tutorials in motor behavior (pp. 49±70). Amsterdam: North-Holland. Konczak, J., Borutta, M., Topka, H., & Dichgans, J. (1995). The development of goal-directed reaching in infants: Hand trajectory formation and joint torque control. Experimental Brain Research, 106, 156±168. Konczak, J. Borutta, M., & Dichgans, J. (1997). The development of goal-directed reaching in infants: II. Learning to produce task-adequate patterns of joint torque. Experimental Brain Research, 113, 465±474. Kugler, P.N., Kelso, J.A.S., & Turvey, M.T. (1980). On the concept of coordinative structures as dissipative structures: I. Theoretical lines of convergence. In G. Stelmach & J. Requin (Eds.), Tutorials in motor behavior (pp. 1±47). Amsterdam: North-Holland. Kugler, P.N., & Turvey, M.T. (1987). Information, natural law, and the self-assembly of rhythmic movement. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Lee, D.N., & Aronson, E. (1974). Visual proprioceptive control of standing in human infants. Perception and Psychophysics, 15, 529±532. Lockman, J., Ashmead, D., & Bushnell, E. (1984). The development of anticipatory hand orientation during infancy. Journal of Experimental Child Psychology, 37, 176±186. McGraw, M.B. (1935). Growth: A study of Johnny and Jimmy. New York: Appleton-Century-Crofts. McGraw, M.B. (1945). Neuromuscular maturation of the human infant. New York: Hafner.

22

Thelen

McGraw, M.B. (1946). Maturation of behavior. In L. Carmichael (Ed.), Manual of child psychology (pp. 332±369). New York: John Wiley. Merzenich, M.M., Allard, T.T., & Jenkins, W.M. (1990). Neural ontogeny of higher brain functions: Implications of some recent neurological ®ndings. In O. Franzn & P. Westman (Eds.), Information processing in the somatosensory system (pp. 293±311). London: Macmillan. Nelson C., & Luciana, M. (Eds.). (2001). Handbook of developmental cognitive neuroscience. Cambridge, MA: MIT Press. Newell, K.M., & Molenaar, P.C.M. (Eds.). (1998). Applications of nonlinear dynamics to developmental process modeling. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Piaget, J. (1952). The origins of intelligence in children. New York: International Universities Press. Preyer, W. (1888). The mind of the child: Pt. 1. The senses and the will (H.W. Brown, Trans.). New York: Appleton. Robertson, S.S. (1993). Oscillation and complexity in early infant behavior. Child Development, 64, 1022±1035. Robertson, S.S., Cohen, A.H., & Mayer-Kress, G. (1993). Behavioral chaos: Beyond the metaphor. In L. Smith & E. Thelen (Eds.), A dynamic systems approach to development (pp. 119±150). Cambridge, MA: MIT Press. Rochat, P., & Morgan, R. (1995). Spatial determinants in the perception of selfproduced leg movements by 3- to 5-month-old infants. Developmental Psychology, 31, 626±636. Savelsbergh, G.J.P., & van der Kamp, J. (1993). The coordination of infants' reaching, grasping, catching and posture: A natural physical approach. In G.J.P. Savelsbergh (Ed.), The development of coordination in infancy (pp. 289±317). Amsterdam: Elsevier. Schneider, K., Zernicke, R., Ulrich, B., Jensen, J., & Thelen, E. (1990). Understanding movement control in infants through the analysis of limb intersegmental dynamics. Journal of Motor Behavior, 22, 493±520. Shadmehr, R., & Mussa-Ivaldi, F.A. (1994). Adaptive representation of dynamics during learning of a motor task. Journal of Neuroscience, 14, 3208±3224. Shirley, M.M. (1931). The ®rst two years, a study of twenty-®ve babies: I. Postural and locomotor development. Minneapolis, MN: University of Minnesota Press. Spencer, J.P., & Thelen, E. (2000). Spatially speci®c changes in infants' muscle coactivity as they learn to reach. Infancy, 1, 275±302. Smith, L.B., Thelen, E., Titzer, R., & McLin, D. (1999). Knowing in the context of acting: The task dynamics of the A-not-B error. Psychological Review, 106, 235± 260. Stoffregen, T.A., Schmuckler, M.A., & Gibson, E.J. (1987). Use of central and peripheral optical ¯ow in stance and locomotion in young walkers. Perception, 16, 113±119. Streri, A., & PeÃcheux, M. (1986). Vision-to-touch and touch-to-vision transfer of form in 5-month-old infants. British Journal of Developmental Psychology, 4, 161±167. Sutherland, D.H. (1984). Gait disorders in childhood and adolescence. Baltimore, MD: Williams & Wilkins. Thelen, E. (1984). Learning to walk: Ecological demands and phylogenetic

1. Motor development

23

constraints. In L.P. Lipsitt (Ed.), Advances in infancy research, Vol. III (pp. 213± 250). Norwood, NJ: Ablex. Thelen, E. (1995). Motor development: A new synthesis. American Psychologist, 50, 79±95. Thelen, E. (2000). Grounded in the world: Developmental origins of the embodied mind. Infancy, 1, 3±28. Thelen, E., & Adolph, K.E. (1992). Arnold L. Gesell: The paradox of nature and nurture. Developmental Psychology, 28, 368±380. Thelen, E., Corbetta, D., Kamm, K., Spencer, J.P., Schneider, K., & Zernicke, R.F. (1993). The transition to reaching: Mapping intention and intrinsic dynamics. Child Development, 64, 1058±1098. Thelen, E., Corbetta, D., & Spencer, J.P. (1996). The development of reaching during the ®rst year: The role of movement speed. Journal of Experimental Psychology: Human Perception and Performance, 22, 1059±1076. Thelen, E., & Fisher, D.M. (1982). Newborn stepping: An explanation for a ``disappearing re¯ex''. Developmental Psychology, 18, 760±775. Thelen, E., & Fisher, D.M. (1983). The organization of spontaneous leg movements in newborn infants. Journal of Motor Behavior, 15, 353±377. Thelen, E., Fisher, D.M., & Ridley-Johnson, R. (1984). The relationship between physical growth and a newborn re¯ex. Infant Behavior and Development, 7, 479± 493. Thelen, E., SchoÈner, G., Scheier, C., & Smith, L.B. (2001). The dynamics of embodiment: A ®eld theory of infant perseverative reaching. Behavioral and Brain Sciences, 24, 1±34. Thelen, E., Skala, K., & Kelso, J.A.S. (1987). The dynamic nature of early coordination: Evidence from bilateral leg movements in young infants. Developmental Psychology, 23, 179±186. Thelen, E., & Smith, L.B. (1994). A dynamic systems approach to the development of cognition and action. Cambridge, MA: MIT Press. Thelen, E., & Ulrich, B. (1991). Hidden skills: A dynamic systems analysis of treadmill stepping during the ®rst year. Monographs of the Society for Research in Child Development, 56 (1, Serial No. 223). von Hofsten, C. (1979). Development of visually guided reaching: The approach phase. Journal of Human Movement Studies, 5, 160±178. von Hofsten, C. (1980). Predictive reaching for moving objects by human infants. Journal of Experimental Child Psychology, 30, 369±382. von Hofsten, C. (1982). Eye±hand coordination in the newborn. Developmental Psychology, 18, 450±461. von Hofsten, C. (1984). Developmental changes in the organization of prereaching movements. Developmental Psychology, 20, 378±388. von Hofsten, C. (1991). Structuring of early reaching movements: A longitudinal study. Journal of Motor Behavior, 23, 280±292. White, B., Castle, P., & Held, R. (1964). Observations on the development of visually directed reaching. Child Development, 35, 349±364. Woollacott, M., & Sveistrup, H. (1992). Changes in the sequencing and timing of muscle response coordination associated with developmental transitions in balance abilities. Human Movement Science, 11, 23±36.

2

The early experience assumption: Past, present, and future H. Rudolph Schaffer

University of Strathclyde, Glasgow, UK

INTRODUCTION One of the most basic and widespread assumptions about the nature of psychological development is summarised in the phrase ``early experience is important''. It is ®rmly based in popular views of childhood, as seen in phrases such as, ``As the twig is bent the tree's inclined'' (Alexander Pope), ``The Child is Father of the man'' (William Wordsworth), and ``Give me a child for the ®rst seven years and you may do what you like with him afterwards'' (attributed to the Jesuits). All such sayings express the belief that children at the beginning of life are more malleable than at any period later on, that initially they can absorb whatever experiences they are exposed to and retain their effects inde®nitely, and that the clues to later personality formation are therefore to be found primarily in those ®rst encounters of the child with its environment. For a major part of the twentieth century, developmental psychologists took over this view and based their theories on the same assumptions. Thus socialisation was seen as a clay moulding process: The newborn infant, that is, was thought of as an unformed lump of clay that adults can then mould into any shape they choose. In time the shape will set, and the form it takes will then be wholly explicable in terms of whatever caretakers did during the child's impressionable years. Children were thus seen as passive; developmental change was explained solely in terms of environmental input to the child; and the earlier the input the more crucial it was thought to be. Yet what seemed so simple and straightforward a proposition at one time has turned out to be an extraordinarily complex and controversial problem. This became increasingly apparent once efforts were made, by means of empirical research, to spell out the terms of the belief that early experience is important. How earlyÐare there critical periods beyond which children can no longer be affected, and if so what are these limits? What experiencesÐeverything, or only certain kinds, and if so which? Important in what wayÐbecause irreversible effects are produced so that the origins of all individual differences in personality functioning are to be found at the earliest stages?

2. The early experience assumption

25

Research in the last few decades has taught us a number of lessons. It has shown that the assumptions of the clay moulding model are mistaken: Children are not passive; their individual characteristics interact in complex ways with experience to determine outcome. There is thus no direct relationship between age and the impact which experience has on the individual. Effects are not irreversible: Later experiences can modify the consequences of earlier ones. Considerable variability therefore exists in outcome, which cannot be predicted from early experiences alone, for these do not exert their effect in isolation but in interaction with other in¯uences. The connection between early and late is consequently far from the straightforward one which previous developmental accounts had proposed. PARENTAL REARING PRACTICES AS ANTECEDENTS In the search for the kinds of experiences which impinge on children early on and which could have a formative in¯uence on them, parental child-care practices are an obvious candidate, and especially so in the light of the emphasis that psychoanalytic theory placed upon them. The belief in the irreversibility of early experience was powerfully bolstered by Freud's claim that the origins of neuroses are to be found in the child's initial encounters with the adult world, with particular reference to the way in which caretakers meet speci®c bodily needs. Variations in rearing practices thus account for the child's developing view of the world and for the course of its emotional life; overgrati®cation, undergrati®cation, frustration, punishment, and inappropriate timing are said to constitute infantile traumata, the nature of which tend to be associated with speci®c personality characteristics in later life. Freud's developmental propositions have, however, not fared well in the light of empirical examination (Caldwell, 1964; Orlansky, 1949). There are no indications that particular infant care practices have an inevitable and unvarying effect, either at the time or later in life. There is, in any case, much controversy as to whether and how young children can retain early experiences and give an account of them in subsequent years (Brewin, Andrews, & Gotlieb, 1993). And most of all, there is doubt whether the particular somatic experiences singled out by Freud are indeed capable of having such profound emotional consequences, in the short-term as well as in the long-term. However, when subsequently the search for antecedents spread to broader, more encompassing aspects of parenting such as permissiveness, punitiveness, and warmth, the success of relating these to outcome variables was no greater. Take the classical study by Sears, Maccoby, and Levin (1957), which set out to ®nd associations between types of child rearing in a large sample of American mothers and the developing personality characteristics of their young children. These associations, however, turned out to

26

Schaffer

be markedly elusive: Only 3% of the variance in the child outcome measures, it has been calculated (Yarrow, Campbell, & Burton, 1968), could be accounted for by maternal practices. As Maccoby (1992) subsequently put it, so few signi®cant correlations between parental and child measures emerged that virtually nothing could be published relating these two sets of data, and the report became instead a descriptive account of the child-rearing practices of one particular group of American mothersÐnot an achievement that the authors had originally set out to obtain. There are a number of reasons, both methodological and conceptual, to account for this particular failure, yet it has become apparent from subsequent research that attempts to relate aspects of early parenting directly to child development outcomes generally meet with only limited success. The correlations obtained have mostly been unimpressive, whatever their signi®cance level, giving little ground for a belief in the all-powerful, clearcut in¯uence that parents were supposed to exert on their children's lives. Above all, however, the lack of success lay in the unidirectional view of the socialisation process, i.e., the assumption that parent±child correlations could be interpreted in terms of cause and effect, that early rearing practices have uniform effects on all children, and that a child's individuality plays no part in development (Schaffer, 1999). Ever since Bell's (1968) landmark paper on child effects and the subsequent adoption of a bidirectional view of socialisation, research on the formative in¯uences on children's development has taken a more complex form. As recent reviews of the literature on temperament show (e.g., Sanson & Rothbart, 1995), attempts to de®ne the individuality that children bring to any social interaction in which they are involved are still continuing; there is as yet no general agreement as to the speci®c constituents of temperament nor ®rmly established knowledge regarding the genetic origin of this construct or its later developmental course. On the other hand, there is agreement that the consequences of parental practices can only be understood by also taking into account aspects of children's individuality which act as mediators for these practices, and that children with different temperamental characteristics will thus respond differently to similar environmentsÐthe principle of ``organismic speci®city'', as Wachs and Gandour (1983) termed it. Relating particular aspects of parenting, however early and however aberrant in nature, to child outcome variables is thus bound to result in limited success when no account is taken of the kind of child on which these experiences impinge: Variability in outcome is considerable and indicates the importance of respecting the child's individuality as an in¯uence in its own development. Parenting Ò temperament designs are thus more useful vehicles for learning about parenting. As yet there are only relatively few studies which have taken this course, though the results they have produced are promising. For example, the impact of maternal depression was found by Wolkind and De Salis (1982) to be greater on children classi®ed as dif®cult in temperament, in that they

2. The early experience assumption

27

were more likely to develop behaviour problems than children with easy temperament. According to Bates, Pettit, Dodge, and Ridge (1998), lax maternal discipline predicted children's externalising symptoms only when combined with children's temperamental resistance to control. And in a series of reports dealing with the development of conscience in children, Kochanska (e.g., 1993; Kochanska, Murray, Jacques, Koening, & Vandegeest, 1996) showed that individual differences in children's internalising of rules can best be accounted for by a combination of parental rearing style and measures of child individuality: Thus for fearful children gentle disciplinary techniques produce optimal outcome whereas for fearless children different parental strategies are required. Such varied ®ndings do not as yet constitute a ®rm body of knowledge concerning the operation of early parental in¯uences; they do, however, indicate that it is the interaction of parent characteristic and child characteristic that is likely to provide the greatest explanatory power rather than either of these two alone. Specifying the kinds of combinations that produce particular kinds of outcomes, as well as explaining the processes whereby this is brought about, is an enormous task for future research. In the meantime it has become apparent that the role which parents play in their children's psychological development is by no means as dominant as it seemed at one time, and that the experiences they provide by way of their rearing practices do not have the unequivocal effect formerly assumed. Given the ubiquity of parenting it is perhaps ironic that we are still unable to spell out with any degree of precision the actual extent of parental in¯uence; in the meantime it is at least noteworthy that Scarr (1992) felt propelled, as a result of her examination of the parent±child literature, to suggest that as long as children receive ``good enough'' parenting their development is determined primarily by inherited characteristics, and that only families that fall outside the average expectable environment are likely to affect psychological development. DEPRIVATION AND THE ISSUE OF IRREVERSIBILITY Scarr's statement is, of course, highly controversial, but it does suggest that in the search for the kinds of early experiences that could leave a lasting impression one should examine extreme aberrations rather than more subtle variations of parental care. This is in fact the course that much of the early literature took, as seen in the concern with the effects of maternal deprivation, i.e. where parental care as a warm and supportive relationship is entirely absent and the child is brought up under the impersonal conditions of unfeeling and unstimulating institutions. Given the highly pathological nature of such an upbringing it was generally expected to produce permanent, irreversible consequences, overriding all variations produced by child individuality.

28

Schaffer

Early studies, as those by Spitz (1945) and Dennis (1973), demonstrated very clearly the devastating effects produced on children by such environments. Findings, initially con®ned to cognitive aspects of development because of the ready availability of measures such as the DQ and the IQ, showed that children of apparently good inherent potential could quickly deteriorate when lacking environmental and personal stimulation, failing to reach developmental milestones at the usual times and ending up with all the signs of gross mental retardation. Given the severity of their condition, it is perhaps not surprising that it was generally assumed that this state was a permanent, irreversible one. However, to come to such a conclusion it is, of course, essential that children are followed up beyond their early years, and moreover that an active attempt is made to reverse their condition by moving them to a new, bene®cial environment. It is only as a result of the gradual accumulation of studies meeting these requirements that it has become apparent that, under certain conditions at least, reversibility even of gross de®ciencies is possible. Skeels (1966), for example, found that by moving infants, severely delayed in their development, from an inadequate orphanage to a much more stimulating environment, considerable improvement could be brought about in their intellectual functioningÐan improvement that they were able to maintain into adult life. Likewise Dennis (1973) observed that children, brought up in their early years in an institution where they received minimal attention and showed gross retardation, were able to function within normal limits following either adoption or removal to a much more stimulating institutionÐeven when that move did not occur till the age of six. And in a series of reports Koluchova (see Clarke & Clarke, 1976, 1998) described the progress of twin boys whose stepmother hid them in a cellar for a large part of their ®rst 7 years, where they grew up in almost total isolation, never taken out of the house and deprived of toys, books, and other such sources of stimulation. When discovered the boys were grossly retarded, lacking any speech and having few social skills, unable to understand pictures, and showing great fear of any new experience. Yet, after being placed with foster parents who devoted themselves wholly to their care and with whom they stayed until adulthood, the boys gradually caught up intellectually and socially, were able to attend school and mix with other children, and as adults became skilled technicians, married and had children, and in general appeared to lead normal, stable lives. A number of other case studies indicate similar reversals of early misfortune (see Skuse, 1993); together these reports fail to support the belief that early adversity inevitably leaves the individual permanently marked. Admittedly, most of these earlier reports can be criticised on methodological grounds, such as the nature of the samples on which they are based, the unreliability of some of the measures used, the indirect nature of parts of the evidence, the contaminating in¯uence of malnutrition and other

2. The early experience assumption

29

physical conditions, and so forth. However, more recent studies, adopting more sophisticated methods able to avoid most of these criticisms, have con®rmed the basic message of reversibility provided by the earlier ones. Thus Rutter and the ERA Study Team (1998) followed up a large group of Romanian children, reared from birth in highly unsatisfactory orphanages but brought to the UK before the age of 2 and placed there in adoptive families. At the time of transfer the children were severely impaired in their development, as seen in both their physical growth and their cognitive level, in comparison with a group of within-UK adoptees. Yet when assessed at age 4 the children showed a remarkable degree of catch-up: Those who came to the UK after the age of 6 months had progressed from a Denver Scale quotient of 48 to one of 97, while those arriving before 6 months progressed from a quotient of 76 to one of 116. Even the disproportionately large number of children with autistic-like symptoms showed marked improvements (Rutter & the ERA Study Team, 1999). When re-assessed at age 6 (O'Connor & the ERA Study Team, 2000), the children were found to have maintained their progress, though there was no evidence of further catch-up. An additional group of later-adopted children, i.e. after the age of 2, and assessed only at age 6, also showed considerable catch-up, though their developmental progress was not as marked as earlier-placed children. Age at the time of move from adverse to bene®cial environment is clearly an important consideration and sets limits to the extent of recovery; nevertheless, the pattern of these ®ndings as a whole underlines the remarkable resilience of development that can become apparent after early and severe deprivation. While the catch-up phenomenon has been documented primarily with respect to early malnutrition (Tanner, 1978), studies such as those mentioned earlier suggest that the equivalent can also be found in psychological growth, giving rise to considerable optimism about the future of children rescued in time from early adversity. However, what ``in time'' amounts to is still dif®cult to spell out on the basis of available data. This is partly because in ``experiments of nature'' investigators have only a limited range of ages to investigate, and partly because age tends to be inextricably confounded with the length of deprivation, making it dif®cult to tell what role each of these variables plays. Thus at some unknown point it may well become too late to reverse damage, whatever the ameliorative action taken. The other reservation to bear in mind concerns the outcome measures: As we shall see below, where these take a global form such as those obtained from developmental or intelligence tests, differences among psychological functions in vulnerability to adversity may well be obscured. Nevertheless, it does appear that the potential of children to recover from early environmental stressors of even a quite severe and lasting nature has been underestimated in the past and that a self-righting tendency analogous to that found in physical growth can, under certain circumstances at least, be seen in psychological development too.

30

Schaffer

ATTACHMENT FORMATION AND THE CRITICAL PERIOD HYPOTHESIS The effects of deprivation on children's social development became a topic for empirical research rather later than the effects on cognitive development, largely because sophisticated measures were not available until appropriate advances in attachment theory had taken place. Yet in his book Maternal care and mental health, published in 1951, Bowlby had singled out the consequences for children's ability to form interpersonal relationships as the main issue in the debate on institutionalisation. To be brought up impersonally in the early years, without a consistently available and loving mother ®gure, will, he maintained, produce permanent damage to children's capacity to establish emotionally meaningful relationships with other people, resulting in a syndrome he referred to as the ``affectionless character''. Children, that is, must be involved in an attachment relationship if they themselves are to form attachments with other people; what is more, that involvement has to occur in a sharply delimited period early on in development. As Bowlby put it in an oft-cited quotation: ``Even good mothering is almost useless if delayed until after the age of 2Ý years''. Inspired by the work of Lorenz (1935), Bowlby believed that there is an early critical period beyond which no amount of the ``right'' kind of experience is effective if the child has been deprived before its end; whatever occurs during this critical period will produce permanent effects. Critical periods, as Lorenz originally formulated them to account for the imprinting phenomenon in precocial birds, represent developmental phases of maximum susceptibility to certain kinds of in¯uence. That animals (and humans too) learn more readily at some stages than at others is not a matter of contention; that this is entirely a function of maturationally determined susceptibility, and that the limits of these periods are sharply de®ned in an all-or-none manner, has not been con®rmed even with the species on which Lorenz based his formulation. In particular, the limits of critical periods have not been found as immutable as Lorenz suggested, nor is the learning that then occurs as permanent as he thought. The term sensitive periods has thus come to be preferred, for as Hinde (1963) put it, it seems that we are merely concerned with changing probabilities of certain forms of learning, so that periods of maximum probability may be surrounded by periods of reduced probability. In any case, the notion that the boundaries of the periods are solely a function of the individual's genetically determined timetable can also not be sustained in the light of evidence; the nature of relevant experience plays a part too. However attractive the critical period hypothesis initially appeared as an explanatory device for the social consequences of children's deprivation, subsequent empirical investigations failed to sustain it. Indeed, one of the ®rst of such investigations was reported by Bowlby himself (Bowlby, Ainsworth, Boston, & Rosenbluth, 1956), and concerned a group of

2. The early experience assumption

31

children who had experienced prolonged separation from home in their ®rst 4 years on account of tuberculosis. During this time they had minimal contact with their parents and were provided with impersonal care, with no substitute mothering being available in the sanatorium where they spent their time. When examined subsequently, aged between 7 and 14 years, there was no indication that these children's capacity to form relationships with other people, parents or peers, was in any way damaged, despite the age when the period of separation had taken place. As the authors were forced to conclude: ``Statements implying that children who are brought up in institutions or who suffer other forms of serious privation or deprivation in early life commonly develop psychopathic or affectionless characters are seen to be misplaced''. A more recent test of the hypothesis that early impersonal upbringing produces irreversible effects on children's capacity to form relationships is provided by Tizard's (1977; Hodges & Tizard, 1989) study of late-adopted children. All these children had been admitted to an institution in the early weeks of life, and were cared for there by such a large number of different people that they had virtually no opportunity to form an attachment to anyone. Subsequently, at an age ranging from 2 to 7, the children were adopted and for the ®rst time given the opportunity to form attachments. As follow-up assessments at ages 8 and 16 showed, the great majority of the children were able to make use of this opportunity: Following adoption the children soon began to show real affection to their new parents and this developed quickly into close emotional bonds. Even those who had spent their ®rst 6 or 7 years in institutions were said to be deeply attached to both of their adoptive parents, none resembling the stereotype of the maternally deprived child originally described by Bowlby. Admittedly, the adoptive parents emerged as a highly motivated group, eager to devote a great deal of time and effort to their children. The fact remains that their efforts were largely successful: There was no indication that prolonged and total emotional deprivation in the early years necessarily leads to the formation of an affectionless character and that lack of close attachments in infancy is bound to bring about an inability to form such attachments later on. Nevertheless, some ill-effects of these children's early experience were noted by Tizard. For one thing, the children showed dif®culties in relationships with adults outside the family: At age 8 they were said to be overfriendly and even affectionate towards strangers, and though this was no longer a problem when seen again at age 16, the children still tended to be more oriented towards adult approval and attention than was the case in the control (nonadopted) group with whom they were compared. And for another, dif®culties with peers also emerged, in that they tended to be less popular, less likely to have a special friend, more quarrelsome, and more often reported to bully other children than their controls. While relationships within the family appeared satisfactory, those outside it were less so, suggesting the children had not emerged entirely unscathed from their

32

Schaffer

experience and that some aspects of behaviour are more likely to show long-term consequences than others. This conclusion is borne out by ®ndings recently reported by Chisholm (1998; see also Chisholm, Carter, Ames, & Morison, 1995), who, like Tizard, was interested in the effects of early deprivation on children's subsequent ability to form attachments. Her investigation concerns another group of Romanian orphans who, as those described by Rutter et al. (1998), had spent their infancy under conditions of severe deprivation, but at ages ranging from 8 months to 5Ý years were adopted by Canadian families. Assessed between 4 and 9 years and having spent a minimum of 2 years in their adoptive homes, all children were found to have formed attachments to their new parents, thus bearing out the conclusion that being kept emotionally ``on ice'' during the alleged critical period need not prevent subsequent development of social bonds. However, when Chisholm also investigated the quality of the attachments formed, she found that signi®cantly more insecure attachment patterns occurred in the ex-institutional group than in a comparison group of never-institutionalised children or, for that matter, in a group of Romanian orphans who had been adopted before the age of 4 months. In addition, children of the ex-institutional group also showed more indiscriminate friendliness to strangers than those in the other two groupsÐa pattern particularly evident in those with the greatest degree of attachment insecurity. These results suggest that conclusions about reversibility may depend on the measures obtained, and that the effects of early adversity do not manifest themselves uniformly across all domains of behaviour. Some psychological functions, such as attachment formation to parents, appear to be better buffered than others such as peer relationships; even with respect to attachments there may be a price to pay in that they are of a less secure nature than those formed in the absence of early deprivation, indicating the need to ask questions not only about the existence of attachments but also about their quality. The possibility that some behaviour systems are better buffered than others has been discussed before (MacDonald, 1985). Thus, according to Scarr (1976), sensorimotor intelligence is an example of a wellbuffered system because it develops under a wide range of rearing environments, even of an aberrant nature. Similarly Skuse (1993) has concluded that perceptuomotor skills are less vulnerable to the effects of deprivation than language development, and that even within language development there are variations from one aspect to another, with expressive speech being more retarded than comprehension. For that matter, there are indications that recovery of physical well-being takes precedence over that of psychological well-being: Thus the case study of Genie, the severely deprived child who was not rescued from her con®nement till the age of 13 years (Curtiss, 1977), shows that she was able to make rapid gains in height and weight but remained severely retarded in all aspects of social adjustment and language development. Also Rutter et al. (1998), in their study of

2. The early experience assumption

33

Romanian orphans, found that physical catch-up in terms of height, weight, and head circumference was very substantial following adoption; the degree of cognitive catch-up, though also marked, was nevertheless somewhat less impressive in comparison. We are badly in need of more data from multiassessment studies before we can reach ®rm conclusions about the differential vulnerability of different functions, and until we have such additional information we can only speculate about the reasons for this variability. One possibility is that some functions are inherently more robust because they are older in an evolutionary sense and more basic in terms of the needs of the individual: Relationships to parents, one can argue, serve more fundamental requirements than relationships to peers and the capacity to establish the former must therefore be more resilient. The other possibility, advanced by Skuse (1993), is that the difference is accounted for by the extent to which a particular function requires a social context in which to develop, for it is this that is most likely to be lacking under conditions of deprivation: Perceptuomotor skills, for example, are less dependent on interpersonal interaction than language development. Whatever the explanation, it is apparent that the previous dependence on just a single outcome measure (usually an all-encompassing one such as is used to assess intelligence or behavioural adjustment) can be misleading by obscuring important differences among speci®c psychological functions in their vulnerability to early events. This is a similar conclusion to that now being reached in the study of attachments. Continuity between infant attachments and social functioning later in life is one of the basic tenets of Bowlby's theory, but it has become increasingly apparent that the continuity is far from absolute. Thus the ``inoculation'' theory of development, as Waters, Posada, Crowell, and Lay (1993) have labelled it, has had to be abandoned in the face of contradictory evidence: Secure attachments in infancy do not necessarily inoculate a child from adverse outcomes throughout development, nor will early attachment dif®culties inevitably place the child at risk for subsequent adjustment problems. Intervening experiences play a part too. What has also become apparent, however, is that the connection between early attachment and later social competence varies according to what aspect of competence is assessed. As Sroufe, Egeland, and Carlson (1999) have shown on the basis of longitudinally tracing such connections, these tend to be more evident with regard to certain aspects of later social behaviour than with others. Thus, while some functions are primarily dependent on early input and are relatively independent of later and concurrent in¯uences, others on the contrary show little trace of events in the distant past and draw more on recent and ongoing experiences. It appears that the ``differential effects hypothesis'' (Verschueren & Marcoen, 1999), expressing the belief that a given type of experience may have different effects on different aspects of behaviour, has wide applicability, and that the use of multiple measures is therefore essential if one is to obtain an accurate picture of the consequences of early experience.

34

Schaffer

VARIABILITY IN OUTCOME Much of the earlier literature gives the impression that all children are equally affected by adversity. This is partly because the experiences then studied were of such severity that all children were indeed severely disturbed by them: Conditions of infant care such as those described by Spitz (1945) inevitably marked every child encountering them. However, it is also because reports at that time were generally expressed in terms of group means, with little attention paid to individual variability. It is only relatively recently that we have come to realise that there are survivors as well as victims, that there are children who go through horri®c experiences of deprivation, abuse, and neglect and yet emerge apparently unscathed. Fergusson, Horwood, and Lynskey (1994), for example, in a study of children brought up in seriously dysfunctional and disadvantaged families, found a large proportion to display multiple behaviour problems in their adolescence; clearly such an upbringing constitutes a considerable risk factor for later maladjustment and delinquency. Yet the association was by no means wholly deterministic: 13% of the teenagers reared under such conditions were problem-free. Similarly Werner's in¯uential follow-up study of children born on the Hawaiian island of Kauai (1989, 1993; Werner & Smith, 1982), showed that there is nothing inevitable about the outcome of unfavourable upbringing. Many of these children were reared under conditions of considerable disadvantage, including such risk factors as poverty, pre- and peri-natal complications, parental mental illness, and disruption of the family unit, and subsequently developed behaviour problems of various types and degrees of severity, Yet others, from the same kind of background, developed normally, showing no signs of maladjustment throughout the follow-up period from early childhood into adulthood. What accounts for such differences? There is now a considerable literature on the vulnerability±resilience continuum along which children can be ranged (e.g. Masten, 1994; Rutter & Rutter, 1993), based on attempts to assemble lists identifying the conditions which act either as protective or as risk factors (Schaffer, 1996). Some of these are characteristics of the child, others those of the rearing environment. Among the former, the child's temperament is prominent; as we saw previously, it represents an essential feature to take into account if one is to understand the differential impact of identical experiences. For example, Werner found ``easy'' children to develop greater stress resistance than ``dif®cult'' childrenÐprobably because the former could more readily establish supportive relationships with adults which would stand them in good stead when faced with adversity, whereas the latter had effects on others which were often described in negative terms and as a result were less likely to elicit approval and help. Temperament has also been considered in relation to children's reactions to parental divorce: Thus dif®cult children, being less adaptable to change, tend to be more badly affected by family

2. The early experience assumption

35

disruption and in particular by the negative behaviour of parents that often accompanies it than easy children (Hetherington & Stanley-Hagan, 1995). Or take activity level, another aspect of children's inherent nature that has been found to affect children's experience of depriving environments: Highly active infants tend to be less damaged under such conditions because they are able to supply their own stimulation than inactive children who are more dependent on adults in this respect (Schaffer, 1966). That there is a link between children's individuality and the varying outcome of exposure to stress cannot be doubted; the mechanisms underlying the link, however, remain to be clari®ed. It may be that some children are more susceptible to adversity because of a lowered tolerance to stress; or that they are more likely to create their own adverse environments; or that they attract more parental hostility and criticism and less emotional supportÐin short, the link either may be a direct one or may operate indirectly through the child's effect on others. More likely than not, it may turn out that several such mechanisms usually operate in tandemÐan indication of the complexity that confronts any attempt to unravel the effects of adversity. Among the factors operating in the child's environment, the nature of family relationships as they exist before, during, and after the adversity is prominent. This is well illustrated by studies of children's long-term responses to acute stressÐhighly traumatic events, that is, which impinge suddenly on the child at one particular point of time and which cause much upset in the short term. Here too, however, long-term consequences can be highly variable. Let us quote three examples. The ®rst concerns one of the worst experiences a child can encounter, namely a parent's suicide. As Shepherd and Barraclough (1976), in a study of such children, have shown, all were much affected at the time and there was no doubt that each child experienced this event as a major trauma. Seen some years later, some children remained considerably disturbed; others, however, were functioning normally, without any sign of maladjustment. The reason for this difference, according to the authors, could be found in the general pattern of the children's lives in which that speci®c event was embedded: Where the suicide was the culmination of considerable previous disturbance affecting the whole family, due to parental mental disturbance, violence, or alcoholism, the children were much more likely to show emotional scars years later than in those cases where the parent's troubles remained relatively self-contained. The traumatic experience, that is, had to be seen in the context of the child's long-term quality of life. Our second example leads to a similar conclusion. This concerns the impact on children of natural disasters, such as earthquakes, ¯oods, hurricanes, and volcanic eruptions (Saylor, 1993)Ðagain highly traumatic events causing much fear and upset at the time. Yet once again there is variability in long-term outcome: Some children recover while in others the problems persist. This is well illustrated in a study by McFarlane (1988) of the effects on children of a large-scale, highly

36

Schaffer

destructive brush®re in Australia that caused much upset at the time in all the children who experienced it. Yet 2 years later only some of the children still showed psychological disturbances that could be traced to this event whereas the remainder functioned normallyÐa difference attributed by the author to the way in which the children's parents reacted to the disaster: the greater the parents' disturbance (as a result of loss of livelihood, home, income, etc.), the greater the likelihood of the child's adverse reactions persisting. The third example, concerning the effects on children of separation from parents, again draws attention to the importance of taking into account the child's general family situation. Separation, particularly during the early years of life and particularly where it involves care by unfamiliar people in an unfamiliar environment, constitutes a considerable trauma with well-documented short-term reactions (Schaffer, 1998). Yet once more we ®nd that the event on its own does not enable one to predict longer-term outcome, for this depends more on the reason for the separation than the separation per se. As Quinton and Rutter (1976) showed, where the child's removal from home is linked to a disturbed family situation representing discord and insecurity, and especially where the separation is part of a chain of other removals from home, later psychological disturbance is far more likely than when it is a single event occurring against the background of a settled and secure home life. Thus repeatedly we ®nd that, however severe the stressful experience may be at the time, its eventual consequences for the individual must be seen in the context of the child's family relationships as they exist both before and after the event.

TRACING DEVELOPMENTAL PATHWAYS In the light of these ®ndings a different methodological approach has had to be adopted in ascertaining the after-effects of early adversity. Instead of jumping straight from that experience to the point at which the outcome is to be assessed, it has become clear that the experience needs to be seen in the context of individual children's life paths, and that instead of expecting particular traumata uniformly to lead to pathological outcomes it is necessary to recognise the modifying role of other events, both before and after the adversity. A simple cause-and-effect model is clearly inadequate; whatever the adversity, it impinges on children of differing personalities, in differing life situations, and with differing possibilities open to them subsequently. Particular attention has of late been given to the last of these sources of variability, i.e., the multiple pathways that lead from early adversity to later personality functioning. The need for longitudinal studies that trace the whole sequence from one to the other and so do justice to the modulating effects that intervening experience may have is now widely acknowledged. Where there is long-term continuity of adversity one can

2. The early experience assumption

37

expect subsequent events to reinforce the effects of earlier ones. But what of individuals who are provided with more positive opportunities, and what kinds of events constitute such opportunities? Longitudinal studies, particularly of a prospective nature, are timeconsuming and results from only a few are as yet available. These have, however, already shown their potential to provide some interesting insights into the nature of developmental change. Take, for example, the studies by Brown and his colleagues (Brown, 1988; Brown, Harris, & Bifulco, 1986; Harris, Brown, & Bifulco, 1986) on the speci®c link between loss of a parent in childhood and the development of depression in adult women. Such a link had been proposed on theoretical grounds, yet empirical investigations that attempted to associate the one directly with the other had come up with contradictory results. By examining other, intervening experiences Brown was able to demonstrate that the link occurred under certain conditions only, i.e., where lack of adequate parental care following loss was experienced by the childÐsomething that was more likely when the loss involved the mother rather than the father, as under the former circumstances instability of care arrangements was more probable. Thus loss of a mother in childhood may predispose the individual to later depression, but this is brought about only in conjunction with the unsatisfactory parenting arrangements to which that event can lead. For that matter, these arrangements in turn were found to make other unfortunate experiences more likely, such as premarital pregnancy and an unsatisfactory marriage, each of which contributed to the ®nal outcome, i.e. the development of clinical depression. The importance of tracing such chains is also shown by Quinton and Rutter's (1988) in¯uential study of the link between a disrupted childhood and subsequent parenting ability. It has often been asserted that a deprived child will become a depriving parent, and indeed the children investigated by Quinton and Rutter, having spent a major part of their early years in institutions, were found as adults to be less sensitive, less supportive, and not as warm in their relationships with their children as those in a comparison group of parents who had been brought up under normal conditions of family life. Yet behind these group averages there was considerable individual variability: By no means all those from a deprived background ended up as de®cient parents. Tracing individual children's developmental trajectories provided examples of the kinds of links making up the chain of circumstances leading from childhood experience to adult behaviour. For instance, positive school experience was found to counteract some of the adverse effects of the children's home experiences; this in turn led to greater planning ability in such matters as choice of career and of marriage partner; the chances of a satisfying job and a supportive marital relationship were thereby greatly increased; and as a result an individual capable of good social functioning, including the provision of sound parental care, emerged from such a sequence.

38

Schaffer

Sequences can take many forms, their course depending to a large extent on the way in which the individual negotiates the turning points that everyone reaches periodically in the course of development. There is growing interest in the nature of such turning points (Rutter, 1996), for decisions taken then may well have considerable implications for an individual's ability to escape from early adversity. To stay on at school or to leave; to obtain an unskilled job or to undertake further study or training; to marry a particular partner or notÐthese are some examples, and although the individual may or may not have control over the choice made, the course taken can reinforce or, on the contrary, help to minimise the consequences of previous experiences. Turning points can thus modify life trajectories (Caspi, 1997); they are an essential part of any explanation of the connection between childhood events and adult behaviour and help to account for the great variety of outcomes that follow identical early experience. The advantage of longitudinal studies is that they can disentangle the role that different though intertwined experiences play in producing particular outcomes. Recent ®ndings on the effects of child abuse illustrate this well (Stevenson, 1999). Abused children tend to come from particular kinds of families, characterised by such features as marital con¯ict, insensitive parenting, social isolation, and poverty. It is known that each one of these characteristics is associated with children's maladjustment, yet much of the earlier literature on child abuse attributed any adverse consequences found in such children to the abuse itself, disregarding the social context in which the maltreatment occurred. However, follow-up studies of such children (e.g. Fergusson & Lynskey, 1997) have shown that their risk status derives more from that context than from the abuse itself; thus in cases of malfunctioning families where there is no abuse the long-term ill-effects are almost as severe as in those where early abuse had taken place. The practical implications of such ®ndings are obvious: Attending to the abuse without rectifying the associated family variables will be of limited help. Thus, longitudinal studies can help to sort out the part played by these various past and current in¯uences; they show that, however traumatic the early maltreatment may have been, it is the continuing in¯uences that help to maintain children's maladjustment. CONTINUITY AND TIMING The problem of early experience and its consequences raises issues concerning the continuity of development and the capacity for individual change. Much of the initial work on early experience was conducted in a climate of opinion that was ®rmly opposed to the idea of change, as seen in the longstanding conviction that IQ constancy was a universal phenomenon and part and parcel of human nature. Given such a belief it is not surprising that recovery from early adversity was considered unlikely; the idea

2. The early experience assumption

39

that individuals were propelled along a ®xed path by whatever happened to them in the ®rst few years of life ®tted in more easily with preconceptions about the working of the mind. No wonder that the rescue of children from conditions of severe deprivation was initially a matter of chance rather than of deliberate planning (Dennis, 1973; Skeels, 1966). It is, of course, an essential part of the early experience assumption that what happens early in life is regarded as important just because it is early. The younger the child the more impressionable it is said to be; the more immature the individual the greater the impact of events encountered. However, as we have gradually come to realise, such a claim cannot be sustained. There is no simple correlation between age and sensitivity to experience. For example, separation from parents is often said to have its greatest impact in the earliest years, but in fact separation as such is not a meaningful event until lasting attachments are formed in the second halfyear of life; until then a safe period exists when infants show no distress in such a situation (Schaffer, 1998). Or consider the impact of parental divorce on children of different ages: No sweeping generalisation linking age to vulnerability can be upheld in the light of contradictory evidence (Hetherington & Stanley-Hagan, 1999); a more likely conclusion is that age groups differ qualitatively rather than quantitatively, i.e., reactions vary in kind among younger and older children and not in severity (Wallerstein, Corbin, & Lewis, 1988). The picture of the highly vulnerable infant and pre-schooler, who gradually grows into a mentally much more sturdy school child and adolescent, cannot be borne out. Particular experiences have an impact depending on whatever mental organisation exists at that particular developmental stage; it is the nature of that organisation rather than age as such that determines how the child responds. The age± vulnerability relationship can therefore be expected to vary according to type of experience: There are indications, for example, that shared rearing in¯uences on personality development are greater at younger than at older ages, whereas nonshared in¯uences on the contrary increase with age right into adulthood (Caspi, 1997). Timing matters, but not in terms of a simple ``the earlier the greater the effect'' type of generalisation. All depends on the kind of experience and the child's ability at any given age to interpret and incorporate it. The notion that continuity of development is guaranteed by the potency of early experience is clearly invalid. Where early maladjustment is followed by later maladjustment, other explanations need to be evoked, and of these the most obvious one is in terms of environmental continuity. Thus, children subjected to unfavourable upbringing in their ®rst years may very likely also experience unfavourable upbringing in later childhoodÐeither because the child remains in the same adverse rearing environment or because of the tendency for one misfortune to lead to another (the ``conveyor belt'' phenomenon). Individual continuity is thus a function of environmental continuity: It is the cumulative effect of all links in the chain

40

Schaffer

rather than the ®rst link on its own that may then be responsible for any long-term damage. However, children are not just passive recipients of others' treatment; as is now generally acknowledged, individuals tend to shape and select their environment to be compatible with whatever characteristics they have been endowed with, so that continuity of environment may in fact re¯ect stable, enduring characteristics of the child (Scarr, 1992; Scarr & McCartney, 1983). Environment-based explanations are, of course, not incompatible with child-based explanations; both mechanisms may be operative simultaneously in a transactional sequence. The child can, however, be a source of continuity in another sense too, namely by virtue of the internal representations of the social world, including of themselves and of their relationships with signi®cant others, that children form in the course of development (Bowlby, 1973). In the early years in particular such cognitive models are by no means closed systems but can be in¯uenced by environmental change; nevertheless, eventually they do come increasingly to act autonomously and in this way become capable of maintaining the individual's expectations and response patterns. Abused children, for example, are known to have poorer selfconcepts as a result of their treatment, and while certain positive experiences such as the opportunity to form close relationships with peers can subsequently moderate this effect (Bolger, Patterson, & Kupersmidt, 1998), such children's mental models of self and others resulting from past treatment may well continue to act as a source of maladaptive behaviour even when the original abuse lies in the past. Tracing developmental pathways is thus not just a matter of plotting the external events impinging on a child but should also include the exploration of children's internal representations left by those events. In so far as these representations will subsequently gain a life of their own, not necessarily directly re¯ecting the child's current experiences, they too require to be examined in the course of the follow-up period so that their contribution to outcome measures can also be taken into account. Developmental continuity, we can conclude, is far from the straightforward process suggested by the favourite metaphor employed in the past to depict it, i.e., that of a house being built where the soundness of the foundations determines the stability of the structure as a whole. As we have seen, direct connections between early experiences and later personality functioning have proved dif®cult to demonstrate. In the past the ``allure of infant determinism'', as Kagan (1998) has referred to it, prevailed as a simple matter of faith; in view of the growing body of evidence, however, a different conception is coming to be recognised, one that sees development (to quote Clarke & Clarke, 1998) ``as a series of linkages in which characteristics in each period have a probability of linking with those in another period. But probabilities are not certainties, and de¯ections of the life path, for good or ill, are possible, although always within the powerful limits imposed by genetic, constitutional and social trajectories''.

2. The early experience assumption

41

CONCLUSIONS Our thinking about early experience and its effects on personality development has gone through three phases in the course of the twentieth century. The ®rst was one of profound pessimism: young children, it was believed, are highly impressionable, and accordingly little hope can be held out to the victims of adversity encountered at that vulnerable period. Early experience was considered to have an overriding effect: Once singled out, the individual was irreversibly marked and beyond help. In the second phase the pendulum swung to the opposite extreme: In¯uenced by a number of reports of normal development following rescue from even quite profound misfortune, marked optimism came to be expressed about the possibility of reversing ill-effects. Early experience was no longer considered to have a privileged position; remedial action initiated even in later childhood appeared capable of undoing earlier effects. The third phase is one that we have entered only quite recently and is characterised by a midposition: On the one hand we realise that early experience does not have a dominant role on its own and that reversal of ill-effects may therefore be feasible; on the other hand we are leaving open the possibility that individuals might, in certain speci®c respects at any rate, remain scarred. In short, it is impossible to give a straightforward yes/no answer to the question, ``Does early adversity leave irreversible effects?'', for all depends on a considerable range of associated conditions over and above the early experience itself. It has become amply apparent in recent years that the search for the effects of early experience needs to be a much more complex enterprise than was ®rst realised. To make use of a simple antecedent±consequent design, in which a group of children undergoing some early adversity is seen again many years later and contrasted with a group who had not undergone such an experience, is clearly insuf®cient. In the ®rst place, it is necessary to take into account the ever greater number of in¯uences that research has shown to be relevant over and above the early experience itself. These include the child's individuality, the social context in which the child lived before and that to which it returned after the experience, the efforts made to remedy the ill-effects, the life opportunities provided to the child subsequently and the particular alternatives chosen, the age at which assessments are made to test for after-effects and the speci®c aspects of behaviour then examined, and also the various dimensions of the original experience such as the child's age at commencement and end, the total length of time involved, the severity of adversity, and the precise components that de®ne the experience. In the second place, exclusive reliance on group averages has been shown to be misleading: It is a feature of every report that has examined individual differences as well as averages to point to the capacity of at least some children to escape the fate that was apparently waiting for them, highlighting the great variability of outcome following identical experience.

42

Schaffer

Third, to jump the time gap between early experience and outcome assessment (often a matter of many years), disregarding the sometimes quite varied experiences making up that gap, is to neglect the modulating effect that subsequent events may exercise: Prospective longitudinal studies are required to trace the varied developmental paths taken and enable one to identify the nature of ameliorative occurrences. Fourth, we need to take into consideration children's subjective interpretation of their experiences: that is, not merely to describe these events in objective terms but also to examine how children construe what is happening to them and the role such constructs play in further development. And ®nally, it has become apparent that much greater precision is required in de®ning both the original adversity and the nature of subsequent outcome. Thus concepts like ``deprivation'' or ``maltreatment'' are too global to be of use for anything more than umbrella terms; each covers a range of different aspects that can have varying impact on children in both the short- and the long-term. Similarly, outcome measures need to be more speci®c than the all-embracing single assessment tools employed in the past, and although this is one major area where more work is required, it is already clear that the effects of adversity differ from one behavioural domain to another and that multiple assessment is thus called for. Such greater speci®city will then enable us to formulate much more precise hypotheses linking early experience to its consequences, taking also into account the varied mediating and modifying conditions referred to earlier. In short, a much more dynamic approach is gradually emerging in attempts to understand what part particular experiences at particular stages of life play in shaping particular aspects of personality. It is an approach that sees even the youngest child involved in a series of transactions with the environment and never merely at the mercy of external events, thus requiring a programme of intensive assessments over time if one is to do justice to the continuing process of reciprocity between child and environment. REFERENCES Bates, J.E., Pettit, G.S., Dodge, K.A., & Ridge, B. (1998). Interaction of temperamental resistance to control and restrictive parenting in the development of externalizing behavior. Developmental Psychology, 34, 982±995. Bell, R.Q. (1968). A reinterpretation of the direction of effects in studies of socialization. Psychological Review, 75, 81±95. Bolger, K.E., Patterson, C.J., & Kupersmidt, J.B. (1998). Peer relationships and selfesteem among children who have been maltreated. Child Development, 69, 1171± 1197. Bowlby, J. (1951). Maternal care and mental health. Geneva: World Health Organisation. Bowlby, J. (1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. London: Hogarth Press.

2. The early experience assumption

43

Bowlby, J., Ainsworth, M., Boston, M., & Rosenbluth, D. (1956). The effects of mother±child separation: A follow-up study. British Journal of Medical Psychology, 29, 211±247. Brewin, C.R., Andrews, B., & Gotlieb, I.H. (1993). Psychopathology and early experience: A reappraisal of retrospective reports. Psychological Bulletin, 113, 82±98. Brown, G. (1988). Causal paths, chains and strands. In M. Rutter (Ed.), Studies of psychosocial risk: The power of longitudinal data. Cambridge: Cambridge University Press. Brown, G.W., Harris, T.O., & Bifulco, A. (1986). The long-term effects of early loss of parent. In M. Rutter, C.E. Izard, & P.B. Read. (Eds.), Depression in young people. New York: Guilford Press. Caldwell, B.M. (1964). The effects of infant care. In M.L. Hoffman & L.W. Hoffman (Eds.), Review of child development research. New York: Russell Sage Foundation. Caspi, A. (1997). Personality development across the life course. In W. Damon (Ed.), Handbook of child psychology, Vol. 3. New York: John Wiley. Chisholm, K. (1998). A three-year follow-up of attachment and indiscriminate friendliness in children adopted from Romanian orphanages. Child Development, 69, 1092±1106. Chisholm, K., Carter, M.C., Ames, E.W., & Morison, S.J. (1995). Attachment security and indiscriminately friendly behaviour in children adopted from Romanian orphanages. Development and Psychopathology, 7, 283±294. Clarke, A.M., & Clarke, A.D.B. (1976). Early exerience: Myth and evidence. London: Open Books. Clarke, A.M., & Clarke, A.D.B. (1998). Early experience and the life path. The Psychologist, 11, 433±436. Curtiss, S. (1977). Genie: A psycholinguistic study of a modern-day ``wild child ''. New York: Academic Press. Dennis, W. (1973). Children of the creche. New York: Appleton-Century-Crofts. Fergusson, D.M., Horwood, L.J., & Lynskey, M. (1994). The childhood of multipleproblem adolescents: A 15-year longitudinal study. Journal of Child Psychology and Psychiatry, 35, 1123±1140. Fergusson, D.M., & Lynskey, M.T. (1997). Physical punishment during childhood and adjustment in young adulthood. Child Abuse and Neglect, 21, 617±630. Harris, T., Brown, G., & Bifulco, A. (1986). Loss of parent in childhood and adult psychiatric disorder: The role of lack of adequate parental care. Psychological Medicine, 16, 641±659. Hetherington, E.M., & Stanley-Hagan, M. (1995). Parenting in divorced and remarried families. In. M. Bornstein (Ed.), Handbook of parenting. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Hetherington, E.M., & Stanley-Hagan, M. (1999). The adjustment of children with divorced parents: A risk and resiliency perspective. Journal of Child Psychology and Psychiatry, 40, 129±139. Hinde, R.A. (1963). The nature of imprinting. In B.M. Foss (Ed.), Determinants of infant behaviour, Vol. 2. London: Methuen. Hodges, J., & Tizard, B. (1989). Social and family relationships of ex-institutional adolescents. Journal of Child Psychology and Psychiatry, 30, 77±98.

44

Schaffer

Kagan, J. (1998). Three seductive ideas. Cambridge, MA: Harvard University Press. Kochanska, G. (1993). Toward a synthesis of parental socialization and child temperament in early development of conscience. Child Development, 64, 325± 347. Kochanska, G., Murray, K., Jacques, T.Y., Koening, A.L., & Vandegeest, K.A. (1996). Inhibitory control in young children and its role in emerging internalization. Child Development, 67, 490±509. Lorenz, K. (1935). Der Kumpan in der Umwelt des Vogels. Journal fuÈr Ornithologie, 83, 137±213, 289±413. Maccoby, E.E. (1992) The role of parents in the socialization of children: An historical overview. Developmental Psychology, 28, 1006±1017. MacDonald, K. (1985). Early experience, relative plasticity, and social development. Developmental Review, 5, 99±121. Masten, A.S. (1994). Resilience in individual development: Successful adaptation despite risk and adversity. In M.C. Wang & E.W. Gordon (Eds.), Educational resilience in inner-city America. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. McFarlane, A.C. (1988). Recent life events and psychiatric disorders in children: The interaction with preceding extreme adversity. Journal of Child Psychiatry and Psychology, 29, 677±690. O'Connor, T.G., & the English and Romanian Adoptees Study Team (2000). The effects of global severe privation on cognitive competence: Extension and longitudinal follow-up. Child Development, 71, 376±390. Orlansky, H. (1949). Infant care and personality. Psychological Bulletin, 46, 1±48. Quinton, D., & Rutter, M. (1976). Early hospital admissions and later disturbances of behaviour. Developmental Medicine and Child Neurology, 18, 447±459. Quinton, D., & Rutter, M. (1988). Parental breakdown: The making and breaking of intergenerational links. Aldershot, UK: Gower. Rutter, M. (1996). Transitions and turning points in developmental psychopathology: As applied to the age span between childhood and mid-adulthood. International Journal of Behavioral Development, 19, 603±626. Rutter, M., & the English and Romanian Adoptees Study Team (1998). Developmental catch-up and de®cit following adoption after severe global early privation. Journal of Child Psychology and Psychiatry, 39, 465±476. Rutter, M., & the English and Romanian Adoptees Study Team (1999). Quasiautistic pattrerns following severe early global privation. Journal of Child Psychology and Psychiatry, 40, 537±550. Rutter, M., & Rutter, M. (1993). Developing minds. Harmondsworth, UK: Penguin. Sanson, A., & Rothbart, M.K. (1995). Child temperament and parenting. In M.H. Bornstein (Ed.), Handbook of parenting, Vol. 4. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Saylor, C.F. (Ed.). (1993). Children and disasters. New York: Plenum. Scarr, S. (1976). An evolutionary perspective on infant intelligence: Species, patterns, and individual variations. In M. Lewis (Ed.), The origins of infant intelligence. New York: Plenum. Scarr, S. (1992). Developmental theories for the 1990s: Development and individual differences. Child Development, 63, 1±19. Scarr, S., & McCartney, K. (1983). How people make their own environments: A theory of genotype-environmental effects. Child Development, 54, 424±435.

2. The early experience assumption

45

Schaffer, H.R. (1966). Activity level as a constitutional determinant of infantile reaction to deprivation. Child Development, 37, 592±602. Schaffer, H.R. (1996). Social development. Oxford: Blackwell. Schaffer, H.R. (1998). Making decisions about children (2nd ed.). Oxford: Blackwell. Schaffer, H.R. (1999). Understanding socialization: From unidirectional to bidirectional conceptions. In M. Bennett (Ed.), Developmental psychology: Achievements and prospects. Philadelphia, PA: Psychology Press. Sears. R.R., Maccoby, E.E., & Levin, H. (1957). Patterns of child rearing. Evanston, IL: Row, Peterson. Shepherd, D.M., & Barraclough, B.M. (1976). The aftermath of parental suicide for children. British Journal of Psychiatry, 129, 267±276. Skeels, H. (1966). Adult status of children with contrasting early life experiences. Monographs of the Society for Research on Child Development, 31 (No. 105). Spitz, R.A. (1945). Hospitalism: An inquiry into the genesis of psychiatric conditions in early childhood. Psychoanalytic Study of the Child, 1, 53±74. Skuse, D.H. (1993). Extreme deprivation in early childhood. In D. Bishop & K. Mogford (Eds.), Language development in exceptional circumstances. Hove, UK: Psychology Press. Sroufe, L.A., Egeland, B., & Carlson, E.A. (1999). One social world: The integrated development of parent±child and peer relationships. In W.A. Collins & B. Laursen (Eds.), Relationships as developmental contexts. Minnesota symposia on child psychology, Vol. 30. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Stevenson, J. (1999). The treatment of the long-term sequelae of child abuse. Journal of Child Psychology and Psychiatry, 40, 89±112. Tanner, J.M. (1978). Fetus into man: Physical growth from conception to maturity. Cambridge, MA: Harvard University Press. Tizard, B. (1977). Adoption: A second chance. London: Open Books. Verschueren, K., & Marcoen, A. (1999). Representation of self and socioemotional competence in kindergartners: Differential and combined effects of attachment to mother and to father. Child Development, 70, 183±201. Wachs, T., & Gandour, M.J. (1983). The relationship of temperament and environment to cognitive development at 6 months: A test of the organismic speci®city hypothesis. International Journal of Behavioral Development, 6, 135±152. Wallerstein, J.S., Corbin, S.B., & Lewis, J.M. (1988). Children of divorce: A 10 year study. In E.M. Hetherington & J.D. Arasteh (Eds.), Impact of divorce, single parenting, and stepparenting on children. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Waters, E., Posada, G., Crowell, J., & Lay, K.-L. (1993). Is attachment theory ready to contribute to our understanding of disruptive behavior problems? Development and Psychopathology, 5, 215±224. Werner, E.E. (1989) High-risk children in young adulthood: A longitudinal study from birth to 32 years. American Journal of Orthopsychiatry, 59, 72±81. Werner, E.E. (1993). Risk, resilience, and recovery: Perspectives from the Kauai longitudinal study. Development and Psychopathology, 5, 503±575. Werner, E.E., & Smith, R.S. (1982). Vulnerable but invincible: A longitudinal study of resilient children and youth. New York: McGraw-Hill. Wolkind, S., & de Salis, W. (1982). Infant temperament, maternal mental state and child behaviour problems. In R. Porter & G. Collins (Eds.), Temperamental

46

Schaffer

differences in infants and young children. CIBA Foundation Symposium 89. London: Plenum. Yarrow, M.R., Campbell, J.D., & Burton, R.V. (1968). Child rearing: An inquiry into research and methods. San Francisco: Jossey-Bass.

3

Behavioural genetics in the 21st century Robert Plomin

Institute of Psychiatry, London, UK

The controversy that swirled around behavioural genetics research during the 1970s has largely faded. During the 1980s and especially the 1990s, the behavioural sciences became much more accepting of genetic in¯uence, as can be seen in the growing number of behavioural genetic articles in mainstream behavioural journals and in research grants. In my view, this is one of the most dramatic shifts in the modern history of the behavioural sciences. The purpose of this chapter is to consider new directions for research in the ®eld of behavioural genetics. Behavioural genetics is the genetic study of behaviour, which includes quantitative genetics (twin and adoption studies) as well as molecular genetics (DNA studies) of human and animal behaviour broadly de®ned to include responses of the organism, from responses measured in the brain, such as functional neuroimaging, to self-report questionnaires. Although animal models permit more powerful methods of analysis because both the genome and the environment can be manipulated, the present chapter focuses on genetic research that can be conducted in the human species. No crystal ball is needed to predict that the momentum of recent developments will carry the ®eld of behavioural genetics well into the new millennium. This momentum is propelled by new ®ndings, methods, and projects in quantitative genetics and especially in molecular genetics. The momentum will grow stronger as genetics continues to ¯ow into the mainstream of behavioural research. Instead of only a hundred or so behavioural geneticists working on personality and cognition (and again as many in psychopathology), more behavioural scientists will incorporate these techniques in their research. This ``giving away'' of behavioural genetics is already happening, with some well-known behavioural scientists leading the way (Plomin, 1993). This is an important trend because the best behavioural genetic research will be done by behavioural scientists who are not primarily geneticists and who use these methods to address theorydriven issues in their speciality ®elds. For this reason, a major motivation for writing this piece is to entice other behavioural scientists, and especially the next generation of behavioural scientists, to incorporate genetics in their research.

48

Plomin

QUANTITATIVE GENETICS Twin and adoption research and genetic research using nonhuman animal models will continue to thrive. We have only scratched the surface of possible applications, even in the most studied domains of cognitive abilities and disabilities, personality and psychopathology. For example, for cognitive abilities, most research has focused on general cognitive ability (intelligence) and major group factors of speci®c cognitive abilities assessed from the psychometric perspective (Plomin & Petrill, 1997). The future of quantitative genetic research in this area lies in more ®ne-grained analyses of cognitive abilities, and in the use of information-processing approaches (Neubauer, Spinath, Riemann, Borkenau, & Angleitner, 2000) and brain measures such as neuroimaging (Kosslyn & Plomin, 2001). Personality is so complex that it can keep researchers busy for decades, especially as they go beyond self-report questionnaires to use other measures such as ratings by others (e.g., Riemann, Angleitner, & Strelau, 1997) and observational measures (Spinath, Riemann, Angleitner, & Strelau, 1999). For psychopathology, genetic research has just begun to consider disorders other than schizophrenia and the major mood disorders. Much remains to be learned about disorders in childhood, for example (Rutter, Silberg, O'Connor, & Simonoff, 1999). A rich territory for future exploration involves the links between psychopathology and personality (Nigg & Goldsmith, 1994). Cognitive abilities and disabilities, personality, and psychopathology have been the targets for the vast majority of genetic research because these areas have traditionally considered individual differences, which are the stuff of genetic analysis. Health psychology and ageing are two newer areas of psychological research that consider individual differences and are beginning to be explored genetically. Some of the oldest areas of psychologyÐ perception, learning, and language, for exampleÐhave not emphasised individual differences and as a result have yet to be explored systematically from a genetic perspective. Entire disciplines within the social and behavioural sciencesÐsuch as economics, education, and sociologyÐare still essentially untouched by genetic research. The greatest need is for quantitative genetic research that goes beyond heritability, that is, beyond asking whether and how much genetic factors are important in behavioural development. We need to address the mechanisms by which genetic factors have their ubiquitous effect on behaviour. How do genetic effects unfold developmentally? What are the biological pathways between genes and behaviour? How do nature and nurture interact and correlate? These three directions for genetic researchÐdevelopmental genetics, multivariate genetics, and ``environmental'' geneticsÐwill play an increasingly important role in future behavioural genetic research that goes beyond heritability.

3. Behavioural genetics in the 21st century

49

DEVELOPMENTAL GENETICS Developmental genetics considers change as well as continuity during development throughout the lifespan. Two types of developmental questions can be asked. First, do genetic and environmental components of variance change during development? The most striking example to date involves general cognitive ability (g). One of the most striking recent ®ndings about g is that when the results are sorted by age, heritability increases from about .20 in infancy to about .40 in childhood to .60 or higher later in life (McGue, Bouchard, Iacono, & Lykken, 1993), even for individuals 80+ years old (McClearn et al., 1997). This increase in heritability throughout the lifespan is interesting because it is counterintuitive in relation to Shakespeare's ``slings and arrows of outrageous fortune'' accumulating over time. The causes of this increase in heritability are not known but a likely culprit is genotype-environment correlation (discussed later): As they live their lives, people actively select, modify, and even create environments conducive to the development of their genetic proclivities (Bouchard, Lykken, McGue, & Tellegen, 1997; Scarr, 1992). For this reason, it may be more appropriate to think about g as an appetite rather than an aptitude. The second developmental issue concerns genetic in¯uence on age-to-age change and continuity in longitudinal analyses. Change in genetic effects from age to age does not necessarily mean that genes are turned on and off during development, although this does happen. Genetic change simply means that genetic effects at one age differ from genetic effects at another age. For example, the same genes could have different effects in the brains of 8-year-olds as compared to 4-year-olds. Longitudinal genetic analysis (Plomin, 1986) can address the etiology of age-to-age change and continuity. Using g again as an example, although most genetic effects on g contribute to continuity from one age to the next, evidence for genetic change has been found at two important developmental transitions (Fulker, Cherny, & Cardon, 1993). The ®rst is the transition from infancy to early childhood, an age when cognitive ability rapidly changes as language develops. The second is the transition from early to middle childhood, at 7 years of age. It is no coincidence that children begin formal schooling at this ageÐall theories of cognitive development recognise this as a major transition. Other longitudinal genetic analyses suggest that, despite such interesting examples of change, a surprising amount of genetic in¯uence on g in childhood overlaps with genetic in¯uence even into adulthood (Plomin, 1986). MULTIVARIATE GENETICS Longitudinal genetic analysis decomposes the covariance of measure across age into genetic and environmental sources of change and continuity. It is a subtype of multivariate genetic analysis that addresses the covariance

50

Plomin

between traits rather than the variance of each trait considered by itself. Multivariate genetic analysis yields a statistic called the genetic correlation, which is an estimate of the extent to which genetic effects on one trait correlate with genetic effects on another trait independently of the heritability of the two traits. That is, although all cognitive abilities are moderately heritable, the genetic correlations between them could be anywhere from 0.0, indicating complete independence, to 1.0, indicating that the same genes in¯uence different cognitive abilities. One of the most surprising results from genetic research on cognitive abilities is that multivariate genetic analyses have consistently found that genetic correlations among cognitive abilities are very highÐclose to 1.0 (Petrill, 1997). In other words, if a gene were found that is associated with a particular cognitive ability, the same gene would be expected to be associated with all other cognitive abilities as well. It is remarkable that g accounts for about 40% of the total phenotypic variance of diverse cognitive tests (Carroll, 1993; Jensen, 1998), but it is extraordinary that g accounts for nearly all of the genetic variance of cognitive tests. That is, what is in common among cognitive abilities is almost completely genetic in origin. This ®nding of genetic g has the interesting converse implication that what is speci®c to each cognitive test is largely environmentalÐwhat makes us good at all tests is largely genetic but what makes us better at some tests than others is largely environmental. Another g-related example of multivariate genetic analysis involves tests of school achievement, which also show substantial genetic in¯uence. Multivariate genetic analyses indicate that genetic effects on g account completely for genetic effects on tests of school achievement (Thompson, Detterman, & Plomin, 1991; Wadsworth, 1994). Conversely, differences between g scores and tests of school achievement are exclusively environmental in origin. This ®nding suggests two important implications. First, school achievement scores independent of intelligence test scores appear to be largely devoid of genetic in¯uence, and would thus provide a better measure of the effectiveness of the child's environment in fostering achievement. Second, the ®nding suggests that underachievement, de®ned as a discrepancy between ability and achievement, is largely due to environmental factors. The importance of genetic g as shown in multivariate genetic research also has important implications for understanding how the brain works from an individual differences perspective. Spearman, who ®rst described g in 1904, noted that ultimate understanding of g ``must needs come from the most profound and detailed study of the human brain in its purely physical and chemical aspects'' (Spearman, 1927, p. 403). The simplest brain model of genetic g is that there is a single fundamental brain process that permeates all other brain processing such as neural speed (e.g., myelination), power (e.g., number of neurons), or ®delity (e.g., density of dendritic spines). The opposite model is that there are many brain processes that are uncorrelated phenotypically and genetically but lead to a genetic correlation in

3. Behavioural genetics in the 21st century

51

performance on cognitive tasks because all of these brain processes are enlisted by the cognitive tasks. Although cognitive neuroscience seldom considers individual differences, current normative theories assume a model of this latter sort in which the brain consists of independent modules (Fodor, 1983; Pinker, 1994). A middle position is that multiple brain processes underlie g in cognitive tasks but these processes are correlated phenotypically and genetically. That is, genetic g might exist in the brain as well as the mind (Paya-Cano, Galsworthy, & Plomin, 2001). To test these different models about brain mechanisms responsible for g, it is necessary to identify reliable individual differences in brain processes and investigate the phenotypic and genetic relationships among these processes. For psychopathology, one of the most important multivariate genetic results is that the same genetic factors are involved in anxiety and depression (Kendler, Neale, Kessler, Heath, & Eaves, 1992), suggesting that this traditional division in affective disorders needs to be re-thought. In developmental psychopathology, such genetic comorbidity seems to be the rule rather than the exception (Rutter et al., 1999). Multivariate genetic research will contribute to a genetic (rather than symptom-based) nosology of psychopathology based on genetic constellations (comorbidity) and components (heterogeneity) of psychopathology. Unlike cognitive abilities or psychopathology, research on self-report personality generally yields a genetic structure that corresponds to its phenotypic structure (Loehlin, 1992). Two other general directions for multivariate genetic research are links between the normal and the abnormal and between behaviour and biology. A different type of multivariate genetic analysis can address the extent to which genetic and environmental effects on disorders are merely the quantitative extremes of the same genetic and environmental factors that affect the rest of the distribution (DeFries & Fulker, 1985; DeFries, Gillis, & Wadsworth, 1998). Or are disorders different in kind, not just in quantity, from the causes of the normal range of behaviour? The surprise from such research so far is that the former seems to be the rule (DeaterDeckard, Reiss, Hetherington, & Plomin, 1997) with a few exceptions (e.g., Dale et al., 1998). This ®ndingÐwhich implies that there may be no disorders, just the extremes of normal dimensionsÐis of great relevance to molecular genetic studies of behaviour, as discussed later. Finally, multivariate genetic analysis also can be used to explicate the mechanisms by which genetic factors in¯uence behaviour by assessing genetic overlap between behaviour and biological processes such as hormones, neurotransmitters, and brain imaging. For example, despite the increasing use of cortisol assays in developmental research, there has been no quantitative genetic research on cortisol or its links with behavioural adjustment. Similarly, in the cognitive realm, there has as yet been no quantitative genetic research on neuroimaging measures and their relationship to cognitive performance (Kosslyn & Plomin, 2001).

52

Plomin

``ENVIRONMENTAL'' GENETICS The present chapter's focus on genetics is not intended to denigrate the importance of environmental in¯uences or to imply biological determinism. To the contrary, it should be emphasised that quantitative genetic research is at least as informative about nurture as it is about nature. In the ®rst instance, quantitative genetic research provides the best available evidence for the importance of the environment in that the heritability of complex traits, including g, is seldom greater than 50%. In other words, about half of the variance cannot be explained by genetic factors. In addition, two of the most important ®ndings about environmental in¯uences have come from genetic research. The ®rst ®nding is that, contrary to socialisation theories from Freud onward, environmental in¯uences operate to make children growing up in the same family as different as children growing up in different families, which is called nonshared environment (reviewed by Plomin, Asbury, & Dunn, 2001). We know this, for example, because genetically unrelated children growing up in the same adoptive family scarcely resemble each other for personality, psychopathology, and cognitive abilities after adolescence. Other behavioural genetic results converge on this surprising conclusion that growing up in the same family does not make siblings similar for environmental reasons (Plomin, Chipuer, & Neiderhiser, 1994). Siblings are similar of course, but for genetic rather than environmental reasons. The environment is important, but environmental in¯uences operate to make children in the same family different, not similar. This topic is called nonshared environment because these environmental in¯uences are not shared by children growing up in the same family. g was thought to be an exception to this rule that shared environmental factors have little in¯uence in behavioural development. Model-®tting estimates of the role of shared environment for g suggest that it accounts for about 20% of the variance for parents and offspring analyses, about 25% for siblings, and as much as 40% for twins (Chipuer, Rovine, & Plomin, 1990). For example, adoptive siblings correlate about .30 (Loehlin, 1989). Because adoptive siblings are genetically unrelated children reared in the same adoptive family, their resemblance can only be due to shared environmental in¯uence in the absence of selective placement. In these studies, little attention was paid to the fact that the adoptive siblings were children until the ®rst study of older adoptive siblings found a correlation of ±.03 (Scarr & Weinberg, 1978). Other studies of older adoptive siblings have also found similarly low correlations for g. The most impressive evidence comes from a 10-year longitudinal follow-up study of more than 200 pairs of adoptive siblings. At the average age of 8 years, the g correlation was .26. Ten years later, their correlation was near zero (Loehlin, Horn, & Willerman, 1989). Thus, although shared environment is an important in¯uence on g during childhood when children are living at home, its importance fades after adolescence as in¯uences outside the

3. Behavioural genetics in the 21st century

53

family become more salient. Next to nothing is known about these mysterious nonshared environmental factors that make siblings growing up in the same family so different after adolescence. This ®nding does not mean that family environment is unimportant. What it means is that environmental in¯uences in development have their effect on an individual-by-individual basis rather than on a family-byfamily basis. The key question is why children growing up in the same family are so different. Nonshared environment is a crucible for environmental research. Unless an environmental variable can be shown to be experienced differently by siblings growing up in the same family, it cannot be an important environmental predictor of developmental outcomes. The discovery of the importance of nonshared environment has sparked research in psychology that studies more than one child per family and asks why they turn out so differently. The largest study of nonshared environment, begun over a decade ago, is the Nonshared Environment and Adolescent Development (NEAD) project (Reiss, Neiderhiser, Hetherington, & Plomin, 2000). The results from NEAD showed that differences in family experiences were strongly related to differences in adjustment outcomes in the adolescent siblings. However, one of the reasons why siblings growing up in the same family are so different is geneticsÐsiblings are 50% similar genetically but this also means that they are 50% different genetically. For this reason, the NEAD study was embedded in a genetically sensitive design that included identical and fraternal twins, full siblings, half siblings, and genetically unrelated siblings. NEAD results suggest that siblings are treated differently because adolescents' genetically in¯uenced patterns of adjustment elicit or evoke different responses from their parents. In other words, differences in parental treatment do not cause the differences in adolescent adjustment by means of nonshared environment. The big question for future research is this: If family environment is not responsible for nonshared environment, what is? One possibility is that nonshared environment lies outside the family. A reasonable candidate in adolescence is differential experiences with peers, which is the theme of a recent provocative book called The nurture assumption (Harris, 1998). Although it makes sense to investigate systematic sources of nonshared environment, we need to keep our minds open to the possibility that chance also contributes to nonshared environment in the sense of idiosyncratic experiences (such as the formative experience when Bill Clinton as an adolescent met John Kennedy) or the subtle interplay of a concatenation of events that might be thought to be chance (Dunn & Plomin, 1990). NEAD is one of a dozen studies that have found genetic in¯uence on measures of the environment, a second important discovery about the environment that has emerged from genetic research. Although most research on the ``nature of nurture'' has focused on family environment, genetic in¯uence has also been reported for life events, social support,

54

Plomin

peers, and work environment (Plomin, 1994). If ostensible measures of the environment are in¯uenced by genetic factors and if developmental outcome measures are also in¯uenced by genetic factors, the implication is that the links between such environmental measures and behavioural outcomes can be mediated by genetic factors. This is the main ®nding from NEAD (Reiss et al., 2000) and several other studies (reviewed by Plomin, 1994; see also Ge et al., 1996; O'Connor, Deater-Deckard, Fulker, Rutter, & Plomin, 1998; Pike, McGuire, Hetherington, Reiss, & Plomin, 1996). Most measures of the environment used in behavioural research show genetic in¯uence when the environmental measures are treated as phenotypes. For example, the most widely used measure of the home environment relevant to cognitive development is the Home Observation for Measurement of the Environment (HOME; Caldwell & Bradley, 1978). In an adoption study of the HOME, correlations for nonadoptive and adoptive siblings were compared when each child was 1 year old and again when each child was 2 years old (Braungart, Fulker, & Plomin, 1992). HOME scores are more similar for nonadoptive siblings than for adoptive siblings at both years (.58 versus .35 at 1 year and .57 versus .40 at 2 years), results suggesting genetic in¯uence on the HOME. Dozens of studies using diverse measures of the environment in addition to family environment such as life events and social supportÐand even television-viewing, accidents, and divorceЮnd consistent evidence for genetic in¯uence. Given that environmental as well as behavioural measures show genetic in¯uence, it is reasonable to ask whether associations between environmental measures and behavioural measures are mediated genetically. Multivariate genetic analysis can be used to analyse genetic and environmental contributions to the correlation between environmental measures and behavioural measures. For example, the adoption study just mentioned found that the HOME correlated substantially with children's g, which the HOME was designed to do (Braungart, Fulker, & Plomin, 1992). Socialisation researchers had reasonably assumed that such HOME measures as parental responsivity, encouraging developmental advance, and provision of toys were environmental causes of children's cognitive development. However, multivariate genetic analysis indicated that about half of the phenotypic correlation between the HOME and children's g is mediated genetically, similar to the ®ndings from the NEAD study. One possible explanation of this result is that parents respond to genetically in¯uenced g in their children. A general rule of thumb in research of this sort is that genetic factors are responsible for about half of the phenotypic correlation between measures of the environment and measures of behaviour. No longer can environmental measures be assumed to be entirely environmental just because they are called environmental. Taking this argument to the extreme, two recent books have argued that socialisation research has been fundamentally ¯awed because it has not considered the role of genetics (Harris, 1998; Rowe, 1994). A far-reaching implication of this research is

3. Behavioural genetics in the 21st century

55

that it supports a shift from thinking about passive models of how the environment affects individuals towards models that recognise the active role we play in selecting, modifying, constructing, and reconstructing our experiences in memory (Plomin, 1994). In summary, genetic research has made some of the most important discoveries about the environment in recent decades, especially about nonshared environment and the role of genetics in experience. More discoveries about environmental mechanisms can be predicted as the environment continues to be investigated in the context of genetically sensitive designs. Much remains to be learned about interactions and correlations between nature and nurture. MOLECULAR GENETICS There is a lot of life left in the old workhorse of quantitative genetics. Quantitative genetics research that goes beyond heritability to address developmental, multivariate, and environmental issues, as described in the preceding section, will be increasingly important in charting the course for molecular genetic analyses that attempt to identify speci®c genes responsible for the ubiquitous heritability of complex traits. For example, multivariate genetic analysis suggests that despite its apparent complexity, g is a good target for molecular genetic analysis because genetic effects on cognitive processes are general. Another example is that quantitative genetic research has been important in showing that genetic variation contributes importantly to differences among individuals in the normal range of variability as well as abnormal behaviour. Moreover, it is increasingly accepted that many complex disorders, especially common ones, may represent the quantitative extreme of the same genetic and environmental factors responsible for variation in the normal range. That is, genetic in¯uence on common disorders such as mild mental retardation may not be due to genes speci®c to the disorder but rather to genes that contribute to the normal range of individual differences in g. This view, known as the quantitative trait locus (QTL) perspective, has important conceptual implications because it implies that some common disorders may not be disorders at all but rather the extremes of normal distributions. This QTL perspective has far-reaching implications for molecular genetics and for neuroscience. If many genes of small effect are involved, it will be much more dif®cult to ®nd them. It will also be much more dif®cult to explore the brain mechanisms that mediate genetic effects on g. The twentieth century began with the rediscovery of Mendel's laws of heredity. The word ``genetics'' was ®rst coined in 1903. Fifty years later the double helix structure of DNA was discovered. The genetic code was cracked in 1966Ðthe four-letter alphabet (G, A, T, C) of DNA is read as three-letter words that code for the 20 amino acids that are the building

56

Plomin

blocks of proteins. The crowning glory of the century and the beginning of the new millennium is the Human Genome Project, which has provided a working draft of the sequence of the three billion letters of DNA in the human genome. Some curmudgeons grumble that the four nucleotide letters that constitute the DNA alphabet are not GATC, but HYPE. They complain that the DNA sequence by itself does not tell us how we can begin life as a single fertilised cell and end up with trillions of cells containing the same DNA but expressing different genes. Knowing the DNA sequence does not by itself cure any diseases. Although it is true that sequencing the human genome is just a ®rst step towards understanding how genes work, it is a giant step forward that ought to be celebrated. The Human Genome Project and the new techniques that are emerging from it will greatly speed up the pace of discovery in genetics even though the current pace is amazing. We are at the threshold of a post-genomic world in which all genes and all DNA variation will be known. The major bene®ciary of this new world will be research on complex traits that are in¯uenced by multiple genes as well as multiple environmental in¯uences. This is where behaviour comes in. Behavioural dimensions and disorders are the most complex traits of all and are already becoming a focus for DNA analysis. The genetic analysis of behaviour includes both quantitative genetic and molecular genetic strategies, the two worlds of genetics that are rapidly coming together in the study of complex traits such as behavioural dimensions and disorders. As compared to research on psychiatric disorders, molecular genetic studies in the cognitive domain were begun relatively recently and have used QTL approaches from the start, which may contribute to the quicker successes in this domain. An example of a behavioural QTL is the association between apolipoprotein E and late-onset dementia (Corder et al., 1993), an association that has been replicated in scores of studies and remains the only known predictor of this common disorder in later life. Although a particular allele in this gene leads to a ®ve-fold increased risk for dementia, it is a QTL in the sense that many people with this allele do not have dementia and most people with dementia do not have this genetic risk factor. This gene is not associated with g in the normal range (Deary et al., in press; Plomin et al., 1995; Turic, Fisher, Plomin & Owen, 2001), although another study did not ®nd a signi®cant association when comparing high g children and normal controls (Plomin et al., 1995; Turic, Fisher, Plomin, & Owen, 2001). Another cognitive trait, reading disability, is the best example of a QTL linkage with a complex trait (Willcutt et al., in press). The advantage of linkage approaches, including QTL linkage approaches, is that they can systematically scan the genome for linkages using just a few hundred DNA markers. QTL linkage designs use many small families (usually siblings) rather than a few large families as in traditional linkage designs. The disadvantage is that they can only detect QTLs that account for a substantial amount (at best about 10%) of the genetic variance given

3. Behavioural genetics in the 21st century

57

reasonable sample sizes. In contrast, allelic association can detect QTLs that account for 1% of the variance but thousands of DNA markers are needed to screen the genome systematically for association because association can only be detected if a DNA marker is very close to a QTL. In other words, linkage is systematic but not powerful and allelic association is powerful but not systematic (Risch & Merikangas, 1996). For this reason, allelic association approaches have largely been limited to studies of ``candidate'' genes. If, as is usually the case, a DNA marker in or near a candidate gene is not itself functional (that is, it does not produce a coding difference), the marker may be close enough to a functional QTL to be in linkage disequilibrium with it and thus yield an indirect association with the disorder. One problem with a candidate gene approach is that, for behavioural disorders, any of the thousands of genes expressed in the brain could be viewed as candidate genes. The way out of this conundrum is to conduct a systematic scan of the genome for allelic association using many thousands of DNA markers, although tens of thousands or even hundreds of thousands of DNA markers would be needed to identify or exclude all QTL associations (Kruglyak, 1999). Such systematic large-scale genome scans or scans of all known candidate gene polymorphisms are becoming possible with new technologies that can quickly genotype thousands of DNA markers. Progress in identifying genes associated with complex traits has been slower than expected, in part because research to date has been underpowered for ®nding QTLs of small effect size, especially in linkage studies. Very large studies are needed in order to identify QTLs of small effect size (Cardon & Bell, 2001). A daunting target for molecular genetic research on complex traits such as g is to design research powerful enough to detect QTLs that account for 1% of the variance while providing protection against false positive results in genome scans of thousands of genes. In order to break the 1% QTL barrier, samples of many thousands of individuals are needed for research on disorders (comparing cases and controls) and on dimensions (assessing individual differences in an unselected and hopefully representative sample). Another factor in the slow progress to date is that only a few candidate gene markers have been examined rather than systematic scans of gene systems or of the entire genome. The Human Genome Project will accelerate progress towards identifying all functional DNA variants expressed in the brain, especially those for entire neurotransmitter pathways. Despite the slow progress to date in ®nding genes associated with cognitive abilities and disabilities, their substantial heritability means that DNA polymorphisms exist that affect these traits. I am con®dent that the pace of QTL discovery will pick up as the Human Genome Project continues to shower the ®eld with new information and new technologies and as studies are designed to break the 1% QTL barrier. Although attention is now focused on ®nding speci®c genes associated with complex traits, the greatest impact for the behavioural sciences will come after genes have been

58

Plomin

identi®ed. Few behavioural scientists are likely to join the hunt for genes because it is dif®cult and expensive, but once genes are found, it is relatively easy and inexpensive to use them (Plomin & Rutter, 1998). DNA can be obtained painlessly and inexpensively from cheek swabsÐblood is not necessary. Cheek swabs yield enough DNA to genotype thousands of genes, and the cost of genotyping can be surprisingly inexpensive. It is critical for the future of the behavioural sciences that we be prepared to use DNA in our research and eventually in our clinics. What has happened in the area of dementia in the elderly will be played out in many areas of the behavioural sciences. As mentioned earlier, the only known risk factor for late-onset Alzheimer's dementia (LOAD) is a gene, apolipoprotein E, involved in cholesterol transport. A form of the gene called allele 4 quadruples the risk for LOAD but is neither necessary nor suf®cient to produce the disorder; hence, it is a QTL. Although the association between allele 4 and LOAD was reported less than a decade ago (Corder et al., 1993), it has already become de rigueur to genotype subjects for apolipoprotein E in any research on dementia in order to ascertain whether the results differ for individuals with and without this genetic risk factor. Genotyping apolipoprotein E will become clinically routine if a genetic risk factor is found to predict differential response to interventions or treatments. In terms of clinical work, DNA may eventually lead to gene-based diagnoses and treatment programmes. The most exciting potential for DNA research is to be able to predict genetic risk for an individual and to intervene to prevent problems before full-blown disorders emerge and create cascades of complications that are dif®cult to counteract. Interventions for behavioural disorders, and even for single-gene disorders, are likely to involve environmental rather than genetic engineering. For example, phenylketonuria (PKU), a metabolic disorder that can cause severe mental retardation, is caused by a single gene on chromosome 12. A particular allele of the gene, found in the recessive condition (two copies of the allele, one from each parent) in 1 per 10,000 babies, damages the developing brain postnatally. This form of mental retardation has been largely prevented, not by high-tech solutions such as correcting the mutant DNA or by eugenic programmes or by drugs, but rather by a change in diet that prevents the mutant DNA from having its damaging effects. For this reason, newborns have for decades been screened for PKU in order to identify those with the disorder so their diet can be changed. The example of PKU serves as an antidote to the mistaken notion that genetics implies therapeutic nihilism, even for a single-gene disorder. This point is even more important in relation to complex disorders that are in¯uenced by many genes and by many environmental factors as well. The search for genes involved in behaviour has led to a number of ethical concerns (Plomin, 1999). For example, there are fears that the results will be used to justify social inequality, to select individuals for education or

3. Behavioural genetics in the 21st century

59

employment, or to enable parents to pick and choose among their foetuses. These concerns are largely based on misunderstandings about how genes affect complex traits (Rutter & Plomin, 1997), but it is important that behavioural scientists knowledgeable about DNA continue to be involved in this debate. Students in the behavioural sciences must be taught about genetics in order to prepare them for this future. As the recent advances from the Human Genome Project begin to be absorbed in behavioural genetic research, optimism is warranted about ®nding more QTLs associated with behavioural dimensions and disorders. The future of genetic research lies in moving from ®nding genes (genomics) to ®nding out how genes work (functional genomics). Functional genomics is usually considered in terms of bottom-up molecular biology at the cellular level of analysis. As an antidote to the tendency to de®ne functional genomics at the cellular level of analysis, the phrase ``behavioural genomics'' has been proposed (Plomin & Crabbe, 2000). Indeed, behavioural genomics may pay off more quickly than other levels of analysis in terms of prediction, diagnosis, therapy, and intervention in relation to behavioural disorders and normal behavioural variation. In addition to producing indisputable evidence of genetic in¯uence, the identi®cation of speci®c genes associated with learning and language delays and disorders will revolutionise genetic research by providing measured genotypes for investigating genetic links between behaviours, for tracking the developmental course of genetic effects, and for identifying interactions and correlations between genotype and environment (Plomin & Rutter, 1998). Identi®cation of speci®c genes might also provide greater precision in diagnoses and individually tailored educational programmes and therapies. For example, gene-based classi®cation of disorders may bear little resemblance to our current symptom-based diagnostic systems. Indeed, from a QTL perspective, common disorders may be the quantitative extreme of normal genetic variation rather than qualitatively different, as mentioned earlier. Most importantly, gene-based prediction of early-onset disorders offers the hope for preventive intervention rather than waiting to intervene when language and learning problems begin to cast a long and wide shadow over children's development. Bottom-up and top-down levels of analysis of gene±behaviour pathways will eventually meet in the brain (Andreasen, 2001). The grandest implication for science is that DNA will serve as an integrating force across diverse disciplines. CONCLUSIONS Few scientists now seriously dispute the conclusion that most behavioural traits show signi®cant genetic in¯uence. The magnitude of genetic in¯uence is still not universally appreciated, however. Taken together, this extensive

60

Plomin

body of research suggests that as much as half of the total variance of behavioural traits can be accounted for by genetic factors. The case for heritability is so strong that quantitative genetic research has moved beyond estimating heritability to address developmental, multivariate, and environmental issues. Using g as an example because more data are available for g than any other trait, its heritability increases during the life span, reaching levels in adulthood comparable to the heritability of height. Longitudinal genetic analyses of g suggest that genetic factors primarily contribute to continuity, although some evidence for genetic change has been found, for example, in the transition from early to middle childhood. Multivariate genetic analysis shows that genetic effects on diverse cognitive abilities are surprisingly generalÐgenetic g accounts for nearly all of the genetic variance on cognitive abilities. Research at the nature±nurture interface reveals that the in¯uence of shared environment diminishes sharply after adolescence and that g-related family environments are substantially mediated genetically, which suggests a more active model of the environment in which genetic propensities affect our experiences. An exciting new direction for behavioural genetic research on g is to identify some of the QTLs responsible for the heritability of g. Although progress to date has been slow, advances from the Human Genome Project, new techniques, and appreciation of the need for suf®cient power to break the 1% QTL barrier makes me optimistic that replicable QTLs will be identi®ed and the ®eld will move on to behavioural genomics.

REFERENCES Andreasen, N.C. (2001). Brave new brain: Conquering mental illness in the era of the genome. Oxford: Oxford University Press. Bouchard, T.J., Jr, Lykken, D.T., McGue, M., & Tellegen, A. (1997). Genes, drives, environment, and experience: EPD theory revised. In C.P. Benbow & D. Lubinski (Eds.), Intellectual talent: Psychometric and social issues (pp. 5±43). Baltimore, MD: Johns Hopkins University Press. Braungart, J.M., Fulker, D.W., & Plomin, R. (1992). Genetic mediation of the home environment during infancy: A sibling adoption study of the HOME. Developmental Psychology, 28, 1048±1055. Caldwell, B.M., & Bradley, R.H. (1978). Home observation for measurement of the environment. Little Rock, AR: University of Arkansas. Cardon, L.R., & Bell, J. (2001). Association study designs for complex diseases. Nature Genetics, 2, 91±99. Carroll, J.B. (1993). Human cognitive abilities. New York: Cambridge University Press. Chipuer, H.M., Rovine, M.J., & Plomin, R. (1990). LISREL modeling: Genetic and environmental in¯uences on IQ revisited. Intelligence, 14, 11±29. Corder, E.H., Saunders, A.M., Strittmatter, W.J., Schmechel, D.E., Gaskell, P.C., Small, G.W., Roses, A.D., Haines, J.L., & Pericak Vance, M.A. (1993). Gene

3. Behavioural genetics in the 21st century

61

dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science, 261, 921±923. Dale, P.S., Simonoff, E., Bishop, D.V.M., Eley, T.C., Oliver, B., Price, T.S., Purcell, S., Stevenson, J., & Plomin, R. (1998). Genetic in¯uence on language delay in 2year-old children. Nature Neuroscience, 1, 324±328. Deary, I.J., Whalley, L.J., St Clair, D., Breen, G., Leaper, S., Lemmon, H., & Starr, J.M. (in press). The in¯uence of the e4 allele of the apolipoprotein E gene on childhood IQ, intelligence in old age, and lifetime cognitive change. Intelligence. Deater-Deckard, K., Reiss, D., Hetherington, E.M., & Plomin, R. (1997). Dimensions and disorders of adolescent adjustment: A quantitative genetic analysis of unselected samples and selected extremes. Journal of Child Psychology and Psychiatry, 38, 515±525. DeFries, J.C., & Fulker, D.W. (1985). Multiple regression analysis of twin data. Behavior Genetics, 15, 467±473. DeFries, J.C., Gillis, J.J., & Wadsworth, S.J. (1998). Genes and genders: A twin study of reading disability. In A.M. Galaburda (Ed.), Dyslexia and development: Neurobiological aspects of extraordinary brains (pp. 186±344). Cambridge, MA: Harvard University Press. Dunn, J.F., & Plomin, R. (1990). Separate lives: Why siblings are so different. New York: Basic Books. Fodor, J.A. (1983). The modularity of mind. Cambridge, MA: MIT Press. Fulker, D.W., Cherny, S.S., & Cardon, L.R. (1993). Continuity and change in cognitive development. In R. Plomin & G.E. McClearn (Eds.), Nature, nurture, and psychology (pp. 77±97). Washington, DC: American Psychological Association. Ge, X., Conger, R.D., Cadoret, R.J., Neiderhiser, J.M., Yates, W., Troughton, E., & Stewart, M.A. (1996). The developmental interface between nature and nurture: A mutual in¯uence model of child antisocial behaviour and parenting. Developmental Psychology, 32, 574±589. Harris, J.R. (1998). The nurture assumption: Why children turn out the way they do. New York: The Free Press. Jensen, A.R. (1998). The g factor: The science of mental ability. Westport, CT: Praeger. Kendler, K.S., Neale, M.C., Kessler, R.C., Heath, A.C., & Eaves, L.J. (1992). Major depression and generalized anxiety disorder. Same genes, (partly) different environments? Archives of General Psychiatry, 49, 716±722. Kosslyn, S., & Plomin, R. (2001). Towards a neuro-cognitive genetics: Goals and issues. In D. Dougherty, S.L. Rauch, & J.F. Rosenbaum (Eds.), Psychiatric neuroimaging research: Contemporary strategies (pp. 491±515). Washington, DC: American Psychiatric Press. Kruglyak, L. (1999). Prospects for whole-genome linkage disequilibrium mapping of common disease genes. Nature Genetics, 22, 139±144. Loehlin, J.C. (1989). Partitioning environmental and genetic contributions to behavioral development. American Psychologist, 44, 1285±1292. Loehlin, J.C. (1992). Genes and environment in personality development. Newbury Park, CA: Sage. Loehlin, J.C., Horn, J.M., & Willerman, L. (1989). Modeling IQ change: Evidence from the Texas Adoption Project. Child Development, 60, 993±1004. McClearn, G.E., Johansson, B., Berg, S., Pedersen, N.L., Ahern, F., Petrill, S.A., &

62

Plomin

Plomin, R. (1997). Substantial genetic in¯uence on cognitive abilities in twins 80+ years old. Science, 276, 1560±1563. McGue, M., Bouchard, T.J., Jr, Iacono, W.G., & Lykken, D.T. (1993). Behavioral genetics of cognitive ability: A life-span perspective. In R. Plomin & G.E. McClearn (Eds.), Nature, nurture, and psychology (pp. 59±76). Washington, DC: American Psychological Association. Neubauer, A.C., Spinath, F.M., Riemann, R., Borkenau, P., & Angleitner, A. (2000). Genetic (and environmental) in¯uence on two measures of speed of information processing and their relation to psychometric intelligence: Evidence from the German Observational Study of Adult Twins. Intelligence, 28, 267±289. Nigg, J.T., & Goldsmith, H.H. (1994). Genetics of personality disorders: Perspectives from personality and psychopathology research. Psychological Bulletin, 115, 346±380. O'Connor, T.G., Deater-Deckard, K., Fulker, D.W., Rutter, M., & Plomin, R. (1998). Genotype±environment correlations in late childhood and early adolescence: Antisocial behavioural problems in the Colorado Adoption Project. Developmental Psychology, 34, 970±981. Paya-Cano, J.L., Galsworthy, M.J., & Plomin, R. (2001). Genetic effects on individual differences in general cognitive ability: Implications for brain mechanisms. Manuscript submitted for publication. Petrill, S.A. (1997). Molarity versus modularity of cognitive functioning? A behavioral genetic perspective. Current Directions in Psychological Science, 6, 96±99. Pike, A., McGuire, S., Hetherington, E.M., Reiss, D., & Plomin, R. (1996). Family environment and adolescent depressive symptoms and antisocial behavior: A multivariate genetic analysis. Developmental Psychology, 32, 590±603. Pinker, S. (1994). The language instinct. New York: William Morrow. Plomin, R. (1986). Development, genetics, and psychology. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Plomin, R. (1993). Nature and nurture: Perspective and prospective. In R. Plomin & G.E. McClearn (Eds.), Nature, nurture, and psychology (pp. 457±483). Washington, DC: American Psychological Association. Plomin, R. (1994). Genetics and experience: The interplay between nature and nurture. Thousand Oaks, CA: Sage. Plomin, R. (1999). Genetics and general cognitive ability. Nature, 402, C25±C29. Plomin, R., Asbury, K., & Dunn, J. (2001). Why are children in the same family so different? Nonshared environment a decade later. Canadian Journal of Psychiatry, 46, 225±233. Plomin, R., Chipuer, H.M., & Neiderhiser, J.M. (1994). Behavioral genetic evidence for the importance of nonshared environment. In E.M. Hetherington, D. Reiss, & R. Plomin (Eds.), Separate social worlds of siblings: The impact of non-shared environment on development (pp. 1±31). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Plomin, R., & Crabbe, J.C. (2000). DNA. Psychological Bulletin, 126, 806±828. Plomin, R., McClearn, G.E., Smith, D.L., Skuder, P., Vignetti, S., Chorney, M.J., Chorney, K., Kasarda, S., Thompson, L.A., Detterman, D.K., Petrill, S.A., Daniels, J., Owen, M.J., & McGuf®n, P. (1995). Allelic associations between 100 DNA markers and high versus low IQ. Intelligence, 21, 31±48.

3. Behavioural genetics in the 21st century

63

Plomin, R., & Petrill, S.A. (1997). Genetics and intelligence: What's new? Intelligence, 24, 53±77. Plomin, R., & Rutter, M. (1998). Child development, molecular genetics, and what to do with genes once they are found. Child Development, 69, 1223±1242. Reiss, D., Neiderhiser, J.M., Hetherington, E.M., & Plomin, R. (2000). The relationship code: Deciphering genetic and social patterns in adolescent development. Cambridge, MA: Harvard University Press. Riemann, R., Angleitner, A., & Strelau, J. (1997). Genetic and environmental in¯uences on personality: A study of twins reared together using the self- and peer report NEO-FFI scales. Journal of Personality, 65, 449±476. Risch, N., & Merikangas, K.R. (1996). The future of genetic studies of complex human diseases. Science, 273, 1516±1517. Rowe, D.C. (1994). The limits of family in¯uence: Genes experience, and behaviour, Vol. 5. New York: Guilford Press. Rutter, M., & Plomin, R. (1997). Opportunities for psychiatry from genetic ®ndings. British Journal of Psychiatry, 171, 209±219. Rutter, M., Silberg, J., O'Connor, T.G., & Simonoff, E. (1999). Genetics and child psychiatry: I. Advances in quantitative and molecular genetics. Journal of Child Psychology and Psychiatry, 40, 3±18. Scarr, S. (1992). Developmental theories for the 1990s: Development and individual differences. Child Development, 63, 1±19. Scarr, S., & Weinberg, R.A. (1978). The in¯uence of ``family background'' on intellectual attainment. American Sociological Review, 43, 674±692. Spearman, C. (1927). The abilities of man: Their nature and measurement. New York: Macmillan. Spinath, F.M., Riemann, R., Angleitner, A., & Strelau, J. (1999). A day in the life: Description of the German Observational Study on Adult Twins (GOSAT) assessing twin similarity in controlled laboratory settings. In I. Mervielde, I. Deary, F. DeFruyt, & F. Ostendorf (Eds.), Personality psychology in Europe, Vol. 7. Tilburg, Netherlands: Tilburg University Press. Thompson, L.A., Detterman, D.K., & Plomin, R. (1991). Associations between cognitive abilities and scholastic achievement: Genetic overlap but environmental differences. Psychological Science, 2, 158±165. Turic, D., Fisher, P.J., Plomin, R., & Owen, M.J. (2001). No association between apolipoprotein E polymorphisms and general cognitive ability in children. Neuroscience Letters, 299, 97±100. Wadsworth, S.J. (1994). School achievement. In R. Plomin & D. W. Fulker (Eds.), Nature and nurture during middle childhood (pp. 86±101). Oxford: Blackwell. Willcutt, E., DeFries, J.C., Pennington, B.F., Smith, S.D., Cardon, L.R., & Olson, R.K. (in press). In R. Plomin, J.C. DeFries, I.W. Craig, & P. McGuf®n (Eds.), Behavioral genetics in the postgenomic era. Washington, DC: American Psychological Association.

4

Nutritional de®ciencies as a biological context for development Theodore D. Wachs

Purdue University, West Lafayette, USA

It has been estimated that in developing countries approximately one third of the population of children under 5 years of age show signs indicative of signi®cant malnutrition (e.g., over 184 million children have weight for height 2 standard deviations below the mean; United Nations Subcommittee on Nutrition, 1992). In certain areas of the world such as South Asia the proportion of severely malnourished children may be even higher (UNICEF, 1998). Although the percentage of young children who are severely malnourished has declined somewhat, population gains have had the effect of increasing the actual number of such children (GranthamMcGregor, 1995). These ®gures do not take into account children who have clinical de®ciencies in speci®c micronutrients such as iodine, vitamin A, or iron de®ciency anaemia (e.g., 100 million young children may have vitamin A de®ciency, while 51% of children under the age of 4 have iron de®ciency anaemia; UNICEF, 1998). Nor do these ®gures take into account children who, while not severely malnourished, are chronically undernourished, with their level of intake of energy or speci®c micronutrients consistently below recommended allowances. Whether viewed in terms of energy de®cits (Simeon & Grantham-McGregor, 1990), or speci®c nutrient de®cits such as iron de®ciency that has not reached the anaemia stage (Stoltzfus, 2001), children who are chronically undernourished are even more prevalent than children who are severely malnourished. In striking contrast to the number of children with nutritional de®ciencies is the relative paucity of evidence on the behavioural and developmental consequences of such de®ciencies. For example, a literature search on PsychInfo for the past quarter century using the single key word ``autism'' turned up over 5900 references. A similar search using the key word ``malnutrition'' turned up only 667 references. The large number of psychological studies on a relatively rare disorder such as autism stands in striking contrast to the relatively small number of studies on one of the most common developmental risk factors of childhoodÐmalnutrition. This discrepancy is particularly surprising given the fact that systematic research on the developmental consequences of malnutrition has been going on for well over half a century (Kaplan, 1972).

4. Nutritional de®ciencies

65

Given that severe malnutrition and vitamin and trace mineral de®cits are far more common in Third World countries it is not surprising that we know far more about the behavioural consequences of nutritional variability in developing countries than in developed countries. This does not mean that we can safely assume that nutritional in¯uences upon behavioural development do not occur or operate in developed countries (Karp, 1993). Particularly for low-income groups in developed countries such as the United States, up to 10% of young children may show stunted physical growth, between 20% and 30% of such children may have iron de®ciency anaemia, while signi®cant numbers of children living in poverty have dietary intakes that are below recommended allowances for folate (13%), magnesium (11%), vitamins A (11%), B6 (16%), and C (18%), and zinc (51%) (Cook & Martin, 1995). Even in more developed countries the prevalence of iron de®ciency anaemia has been related to an increased risk of mental retardation (Hurtado, Claussen, & Scott, 1999) and problems in school achievement (Palti, Meijer, & Adler, 1985). Clinical research in developed countries also indicates a signi®cant increase in developmental risk for breast-fed offspring of vegan mothers, as a result of maternal vitamin B12 de®ciency (Sklar, 1986). In addition, in developed countries there is a moderate increment in long-term cognitive performance associated with children being breast-fed as opposed to being formula-fed, particularly for infants who have early biomedical problems (GranthamMcGregor, Fernald, & Sethuraman, 1999a). While there are multiple reasons why breast-feeding could in¯uence later cognitive performance, one critical nutritional parameter is the greater suf®ciency of essential fatty acids found in human breast milk (de Andraca, Peirano, & Uauy, 1998). To what extent are developmental scientists limiting themselves theoretically, methodologically, or in terms of practical applications if they ignore the role played by nutrition? In the following section of this chapter, I review evidence on the role played by nutritional de®cits in regard to individual behavioural developmental variability in childhood. Because of space limitations a number of potentially relevant topics have been omitted. Relevant references on omitted topics are provided below. Omitted topics include studies on behavioural outcomes associated with brief changes in dietary intake such as adding or missing breakfast (see Pollitt & Mathews, 1998) or short-term changes in intake of carbohydrates or speci®c amino acids (see Christensen, 1996), infrahuman studies on nutrition and behaviour (see Strupp & Levitsky, 1995), studies on the relation of breast-feeding to development (see Grantham-McGregor et al., 1999a) and research on the relation of nutritional de®cits to indicators of functional competence in adult and ageing populations (see Alpert & Fava, 1997; Silver, 1992; Spurr, 1987). Similarly, although obesity in the general population is increasing at an alarming rate (Popkin & Doak, 1998), and although the underlying mechanisms can involve lifestyle factors such as the nature of parental feeding patterns and level of physical activity (Popkin, 1994; Stunkard,

66

Wachs

2000), space limitations preclude further discussion on the potential developmental consequences associated with obesity. Finally, given inconsistent evidence and consistent methodological problems, I will not be considering the consequences of food sensitivities, re®ned foods, or mega-vitamin treatments for understanding behavioural disorders (Christensen, 1996; Pfeiffer, Norton, Nelson, & Shott, 1995; Toller, 1992). NUTRITION AND HUMAN BEHAVIOURAL DEVELOPMENT A summary of established research ®ndings on the relation between behavioural development and different types of nutritional de®cits is shown in Table 4.1. Although Table 4.1 seems to suggest a distinction between malnutrition (often referred to as ``protein-energy malnutrition'') and de®cits in speci®c nutrients, in reality de®cits in protein-energy intake often covary with intake de®cits in speci®c vitamins and minerals. As a result, increasing consideration is being given to the hypothesis that some of the developmental consequences attributed to protein-energy malnutrition may be due to more speci®c vitamin or mineral de®cits (Pollitt, 1995).

Severe malnutrition Prenatal malnutrition Although there is provocative epidemiological evidence indicating long-term biomedical consequences associated with indices of intrauterine growth retardation (Barker et al., 1993), evidence on the behavioural-developmental consequences of being small for gestational age is equivocal. In good part this may re¯ect methodological problems inherent in many studies in this area (Grantham-McGregor et al., 1999a), such as not taking into account the nature of postnatal psychosocial in¯uences (Grantham-McGregor, Chang, Walker, & Powell, 1998). In addition, intrauterine growth retardation can be caused by a variety of in¯uences in addition to maternal malnutrition (Kleinman, 1992), with the likely possibility that the nature of developmental outcomes may well vary depending upon the etiology of an infant being small for gestational age. To the extent that there are cognitive consequences associated with intrauterine growth retardation, de®cits tend to be less obvious early in development and are more likely to appear in the middle childhood period (Grantham-McGregor et al., 1999a). Postnatal malnutrition As shown in Table 4.1, a major consequence of severe postnatal malnutrition is an increased risk of cognitive impairment. The cognitive de®cits associated with severe malnutrition can be partially ameliorated by

4. Nutritional de®ciencies

67

Table 4.1 Summary of ®ndings on nutritional de®cits and human behavioural development Nutrient de®cit

Summary of research

References

Prenatal intrauterine growth retardation

Findings inconsistent

Grantham-McGregor et al. (1999a); Stein, Susser, Saenger, & Marolla (1975)

Severe postnatal Increased risk of general and speci®c Grantham-McGregor malnutrition cognitive de®cits as well as (1995) problems in school achievement Behaviour characterised by apathy, Geber, & Dean (1956); irritability and lower activity level Grantham-McGregor (1984) Severe mineral and vitamin de®cits: Prenatal iodine de®ciency

Increased risk of cognitive de®cits including mental retardation

Grantham-McGregor, Fernald, & Sethuraman (1996b) Iron-de®ciency Increased risk of cognitive de®cits Lozoff (1998); Watkins, anaemia & Pollitt (1998) Lower activity level, lower reactivity, Lozoff, & Wachs (2001) higher inhibition, greater negative affect, and irritability Chronic Increased risk of poorer cognitive Gorman (1995); Sigman, undernutrition and academic performance Espinosa, & Whaley (1998); Wachs (1995)

Essential fatty acid de®cits

Lower activity and reactivity in infancy and lower activity and less sociability in school-age children Improved visual acuity, sleep regulation, and problem solving in infants whose formulas were supplemented with essential fatty acids

Mild mineral and vitamin de®cits

Sigman et al. (1998); Wachs (1995, 2000a) Uauy et al. (1996); Willatts, Forsyth, DiModugno, Varma, & Colvin (1998)

Higher intake of animal source iron Sigman et al. (1998); related to higher cognitive function Wachs (1995) in infants and school-age children Some studies report lower cognitive Bruner, Joffe, Duggau, performance in iron de®cient nonCasella, & Brandt anaemic children (1996); Halterman, Kaczorowski, Aligne, Auinger, & Szilaggi (2001) Zinc supplementation related to Black (1998); Merialdi, higher activity levels Caul®eld, Zavaleta, Figueroa, & DiPietro (1998)

68

Wachs

providing supplementary food (Grantham-McGregor et al., 1999a). For example, a long-term study carried out in Guatemala showed small but signi®cant differences in adult cognitive performance favouring malnourished children who were nutritionally supplemented during the ®rst several years of life, as compared to malnourished children who received a nutritional placebo (Pollitt, Gorman, Engle, Martorell, & Rivera, 1993). As also shown in Table 4.1, there are characteristic behavioural patterns that result from severe malnutrition. Many of these symptoms vanish as the child is nutritionally rehabilitated (Grantham-McGregor, Stewart, & Powell, 1991). However, even for nutritionally rehabilitated children, longterm follow-up indicates that such children are still characterised as being more distractible and lower in reactivity, emotional control, and activity level than matched nonmalnourished controls (Grantham-McGregor, 1995; Simeon & Grantham-McGregor, 1990).

Mineral and vitamin deficits As shown in Table 4.1, cognitive de®cits can occur as a result of both prenatal iodine de®ciency and postnatal iron de®ciency that results in iron de®ciency anaemia (IDA). Maternal iodine supplementation during the ®rst and second trimesters can sharply minimise the cognitive de®cits associated with prenatal iodine de®cit (Grantham-McGregor, Fernald, & Sethuraman, 1999b). The consequences associated with postnatal iodine de®ciency are less clear. In most studies it is dif®cult to separate out the consequences associated with prenatal versus postnatal iodine de®ciency. Studies of the effects of iodine supplementation in older children generally have yielded inconsistent results (Grantham-McGregor et al., 1999b). In regard to IDA, iron supplementation has been shown to have consistent cognitive bene®ts primarily for older, iron-de®cient, anaemic children, with the evidence on supplementation effects for anaemic infants being less consistent (Grantham-McGregor et al., 1999b; Lozoff & Wachs, 2001). Outside of iron de®ciency anaemia and prenatal iodine de®ciency, little evidence is available at the human level on the relation to cognitive performance of severe de®cits in other micronutrients. Although infrahuman research has documented the role played by vitamin B6 in neurotransmitter processes underlying learning and memory, little evidence is available at the human level on relations between severe vitamin B6 de®ciencies and cognitive performance (Guilarte, 1993). Similarly, although infrahuman research has implicated zinc-de®ciency in certain types of cognitive impairments, for the most part human evidence on cognitive performance in zinc-de®cient children has not shown a consistent pattern of results (Black, 1998). In part this may re¯ect the possibility that other micronutrient de®cits may interact with zinc, thus reducing the impact of zinc supplementation (Sandstead et al., 1998).

4. Nutritional de®ciencies

69

Chronic undernutrition Protein±calorie±macronutrient deficits As shown in Table 4.1, there is a consistent pattern of relations between chronic undernutrition and cognitive, academic, and behavioural de®cits. Of particular relevance is evidence indicating that cognitive, academic, or behavioural performance may be particularly sensitive to poor-quality dietary intakeÐdiets low in animal source foods, diets that are low in critical micronutrients, or diets where the critical micronutrients are not easily bioavailable (Sigman et al., 1998; Wachs, 1995). In addition a further consequence of early chronic undernutrition may be a greater susceptibility to later dietary de®cits, given recent evidence that children with a history of chronic undernutrition show poorer cognitive performance after an experimentally induced, brief fasting period (e.g., missing breakfast) than do children with more adequate nutritional histories (Grantham-McGregor, Chang, & Walker, 1998; Pollitt, Cueto, & Jacoby, 1998). Relations between intake of essential fatty acids and early developmental competence, as shown in Table 4.1, may re¯ect inadequate levels of critical long chain polyunsaturated fatty acids in infant formula (Willatts et al., 1998). Mild mineral and vitamin deficits Results are mixed in terms of iron de®ciency. The overwhelming majority of studies using biochemical markers generally indicate no cognitive de®cit for individuals who are iron-de®cient but not anaemic (Watkins & Pollitt, 1998). However, from a methodological standpoint the overall validity of this conclusion has been questioned (Parks & Wharton, 1990). For example, if the cognitive effects of iron de®ciency without anaemia are relatively speci®c in nature, such effects would be less likely to be detected by the global assessment instruments that are typically used in studies on this question (Wauben & Wainright, 1999). In addition, the evidence supporting a conclusion of no cognitive consequences is not totally consistent. Some studies do report cognitive de®cits in children who are iron-de®cient but not anaemic (see Table 4.1), which may be the result of iron de®ciency-driven de®cits in neural function (Otero, Aguirre, Porcayo, & Fernandez, 1999). Results from some infrahuman studies (Felt & Lozoff, 1996) are consistent with this conclusion. As also shown in Table 4.1, there is a consistent body of research linking greater zinc intake to higher activity level, even when infants are still in utero (Merialdi et al., 1998). Interestingly some studies have reported that zinc±activity linkages may be stronger for boys than for girls (Sazawal et al., 1996). Although far less evidence is available, a few studies have related higher B vitamin intake to better cognitive performance and to more optimal temperament in infants and toddlers (Wachs, 1995).

70

Wachs

MECHANISMS UNDERLYING LINKAGES BETWEEN NUTRITION AND BEHAVIOURAL DEVELOPMENT

Biological mechanisms Whether viewed in terms of the development of central nervous system (CNS) macrostructure (Levitsky & Strupp, 1995; Morgane, Kemper, Stern, Hawrylewicz, & Resnick, 1978), CNS microstructure (Bedi, 1987; Tacconi, Calzi, & Salmona, 1997), or the level and operation of neurotransmitters (Christensen, 1996; Guilarte, 1993), there is consistent evidence showing how level and quality of dietary intake in¯uences both the development and function of the CNS. For example, early iron-de®ciency anaemia may have relatively permanent effects on brain myelination, thus restricting the level of gains that can result from early iron supplementation (Lozoff, 1998). Alternatively, iron-de®ciency anaemia may also act to reduce the number of D2 dopamine receptors, and thus disrupt functions mediated by the dopamine neurotransmitter system (Lozoff, 1998). Similarly, there is an increasing amount of evidence delineating the role played by essential fatty acids for both CNS structural development and neurotransmitter function (Wauben & Wainright, 1999; Yehuda, Rabinovitz, Carasso, & Mostofsky, 1998). The role played by nutrition in CNS development and function has clear implications for understanding individual variability in behavioural development. For example, both CNS structures (e.g., the hippocampus) and neurotransmitters (e.g., serotonin) have been linked to individual differences in temperament domains such as negative emotionality. These are the same structures and neurotransmitters that have also been shown to be sensitive to both quantitative (e.g., protein±calorie intake) and qualitative (e.g, tryptophan, zinc intake) nutritional de®cits (Wachs, 2000a). In addition to direct effects on CNS development and function, nutritional de®ciencies may also have indirect effects upon development, manifested through an increased likelihood of other biological risk factors. For example, maternal iron de®ciency anaemia has been associated with an increased risk of offspring low birth weight and preterm birth (Rasmussen, 2001).

Psychosocial mechanisms Although there is more evidence available on nutrition±biological pathways, evidence also supports the hypothesis that nutritional de®cits can impact upon individual behavioural development via psychosocial pathways. One such pathway involves the role of nutrition on patterns of parent± child transactions. A variety of disturbances in parent±child relations in undernourished populations have been reported, including lower parental sensitivity (Valenzuela, 1997), reduced positive affect by parents (Lozoff et al., 1998), and lower-quality vocalisation patterns by parents toward their

4. Nutritional de®ciencies

71

infants (Meeks-Gardner, Grantham-McGregor, Himes, & Chang, 1999). One explanation for this pattern of ®ndings is that inadequately nourished parents will have less energy available to allow them to provide appropriate levels of stimulation to their children (Grantham-McGregor, 1984). For example, evidence indicates that mothers with poor vitamin B6 status are less responsive to their infants' vocalisations and are less effective at soothing when their infants are distressed (McCullough et al., 1990). Alternatively, there is also evidence indicating that poorly nourished children are less able to elicit developmentally facilitative interactions from their caregivers, or are less able to respond to the stimulation provided by their caregivers (Chavez, Martinez, & Yaschine, 1971; Super et al., 1981). Obviously these two explanations are not inconsistent with each other, and there may well be a synergistic interaction effect between malnourished infants and malnourished caregivers. This issue will be discussed further in the following section.

FUTURE DIRECTIONS IN NUTRITION±DEVELOPMENT RESEARCH

Nutrient±nutrient interactions In detailing the speci®c steps de®ning nutrition!CNS!behavioural development pathways, a particularly critical area for future research will involve the study of the neural and developmental consequences of interactions among multiple nutrients. Although most current nutrition±behaviour studies involve single nutrients, humans typically consume foods containing multiple nutrients. The bioavailability of a given nutrient will often depend on the presence or absence of other nutrients in the diet (Sandstead, 1995). For example, the degree of absorption of non-heme iron is dependent upon the level of intake of a variety of other nutrients including vitamin C, phytate, calcium, and copper (Lynch, 1997). At present there is all too little evidence on the consequences of nutrient±nutrient interactions for either human neural or behavioural development. In addition to nutrient±nutrient interactions, another potentially important direction for future research involves the question of whether adequacy of nutritional intake can act to moderate the impact of other biological risk factors. For example, trace mineral de®cits in iron, calcium, or zinc have been shown to increase the absorption of lead (Goyer, 1995), which may act to increase the negative developmental consequences associated with exposure to lead in infancy and childhood. Similarly, a negative developmental spiral may occur as the result of inadequate nutrition acting to impair immune system ef®ciency, leading to an increased rate of childhood illness, which in turn may accentuate existing nutritional de®ciencies (Keusch, 1990).

72

Wachs

An expanded conceptualisation of context Although developmental scientists have increasingly emphasised the importance of contextual variability, for the most part their focus has been on the individual's psychosocial context (e.g., family characteristics, peer groups, the neighbourhood, and the larger culture; Moen, Elder, & Luscher, 1995). Far less attention is given by developmentalists to the individual's biological context, which can include disease-producing factors like sanitation or parasites (e.g., Connolly, 1998), environmental pollution (e.g., Evans, 1994), and nutritional factors like availability of food or the quality of food that is available to the individual, as discussed earlier. These two aspects of context are not independent of each other. Populations at risk for malnutrition are also far more likely to be exposed to inadequate housing, developmentally inhibiting parental rearing styles, and inadequate schooling (Grantham-McGregor, 1995; Pollitt, 1988; Ricciuti, 1993). Although higher levels of prediction have been documented using developmental risk indices involving both psychosocial and biological parameters (e.g., Sameroff, 1994), for the most part, multidimensional developmental risk indices have not included nutritional variables. The need for systematic inclusion of nutritional status variables in developmental risk indices is seen in evidence indicating signi®cant increments in prediction when both nutritional and psychosocial risk or protective factors are combined rather than being treated in isolation (Grantham-McGregor, 1995; Wachs, 2000b). Similarly, studies also indicate that the degree of gain shown for disadvantaged children adopted into advantaged environments depends upon their early level of malnutrition (My Lien, Meyer, & Winick, 1977; Winick, Meyer, & Harris, 1975), and that the degree to which disadvantaged children were able to bene®t from advanced schooling was found to vary, depending upon whether or not the child had received nutritional supplementation early in life (Pollitt et al., 1993). The reverse also holds. Except when considered as a ``nuisance'' variable to be covaried out, nutritional researchers studying nutrition±behaviour links have tended not to include measures of psychosocial context in their studies. This neglect of the psychosocial context occurs in spite of consistent evidence showing how psychosocial contextual factors can in¯uence both the extent and pattern of relations between nutrition and development (Wachs, 2000b). For example, there are a number of studies documenting how the detrimental in¯uences of early malnutrition can be attenuated when previously malnourished children are subsequently reared in more adequate psychosocial environments (Colombo, de la Parra, & Lopez, 1992; Paine, Dorea, Pasquali, & Monteir, 1992; Winick et al., 1975). Similarly, a number of intervention research studies have documented how both stronger and longer-lasting effects occur when nutritional supplementation is combined with psychosocial stimulation in populations of malnourished children (Powell, Walker, & Grantham-McGregor, 1998).

4. Nutritional de®ciencies

73

Psychosocial context can also act to alter the pattern of relations found between nutrition and development. For example, the impact upon classroom attention patterns of supplementary breakfast feeding to undernourished children varied dramatically, depending on school context. In orderly school environments breakfast feeding facilitated children's on-task behaviour in the classroom, whereas in chaotic school environments supplementary breakfast feeding led to reduced on-task classroom behaviour (Grantham-McGregor et al., 1998). Similarly, distinctly different patterns of nutrition±behaviour relations have been found to occur as a function of the harshness of ecological conditions in different villages (Barrett, Radke-Yarrow, & Klein, 1982) and as a function of contextually based expectancies about the types of behaviour patterns that are appropriate for male versus female school children (Wachs et al., 1995). These patterns of results all converge on a conclusion that neither the nutritional (biological) nor the psychosocial contexts within which children live function in isolation from each other (Wachs, 2000b). A full understanding of the role in development played by context will require research that integrates measures of both the biological and psychosocial context. Two potential approaches toward such integration are discussed next.

The study of functional isolation In its expanded version described later, functional isolation is a theoretical framework which integrates the contributions of three known in¯uences upon development: nutrition, environment, and CNS function. The concept of functional isolation arose from observational studies detailing higher levels of apathy and lower levels of activity shown by severly malnourished young children (Grantham-McGregor, 1984), as well as from infrahuman research documenting changes in exploration and learning patterns in malnourished animals (Strupp & Levitsky, 1995). Based on these ®ndings it was postulated that the developmental impact of malnutrition involves not only direct in¯uences of malnutrition upon the developing central nervous system (CNS), but also indirect in¯uences of reduced exploration and involvement with the environment (Levitsky & Barnes, 1972; Pollitt et al., 1993). Although originally formulated to explain the developmental consequences of moderate±severe protein±calorie malnutrition, the concept of functional isolation has been elaborated to include the developmental impact of trace mineral de®ciencies, such as iron-de®ciency anaemia (Lozoff, 1998) and zinc de®ciency (Black, 1998). The original functional isolation hypothesis can be further expanded by integrating in known bidirectional links between environment, nutrition, and CNS function in a way that is consistent with the original intent of this hypothesis. This expanded functional isolation hypothesis is shown in Figure 4.1. Within the functional isolation framework shown in Figure 4.1, developmental de®cits are considered to be the result of both biological and

74

Wachs Developmentally inappropriate parenting

Child nutritional deficits

CNS development and function

Developmental outcomes

Reduced level of child involvement with the environment

Figure 4.1 An expanded conceptual framework encompassing functional isolation processes.

psychosocial mechanisms. Biologically, nutritional de®ciencies are hypothesised to lead to an altered CNS, which in turn is associated with de®cits in speci®c aspects of development, depending upon what CNS functions are most affected. Ample evidence is available supporting the hypothesis that nutritional de®ciencies can result in structural and functional impairments in CNS development (Levitsky & Strupp, 1995; Lozoff, 1998; Morgane et al., 1993; Rao & Georgieff, 2000; Yehuda, Rabinovitz, & Mostofsky, 1997). In addition, there is an increasing body of evidence illustrating how variability in environmental rearing conditions can in¯uence CNS development and function (Greenough & Black, 1992; Nelson & Bloom, 1997; Schore, 1994). In terms of the psychosocial paths shown in Figure 4.1, nutritional de®ciencies are hypothesised to be associated with reduced environmental involvement by the child, as well as with developmentally inappropriate caregiver±child transactions (e.g., treating the child is if it were younger then it is). In turn, reduced environmental involvement and developmentally inappropriate parent±child transactions are also hypothesised to be associated with reduced developmental competence. Unfortunately, all too few studies have directly tested the function isolation hypothesis and those that have are not always supportive of this hypothesis (Gardner, Grantham-

4. Nutritional de®ciencies

75

McGregor, Chang, Himes, & Powell, 1995). In part this may re¯ect an overemphasis on activity level as an proxy for functional isolation, rather than on behaviours that are more likely to re¯ect functional isolation directly, such as the extent and quality of exploration by the child, or the speci®c nature of interactions between caregivers and malnourished children. When outcome measures that are more salient to functional isolation are utilised, studies have reported evidence supporting the validity of the functional isolation path between nutritional de®ciencies and reduced environmental involvement. Young moderately to severely malnourished children have been shown to demonstrate reduced attention to the environment (Lester, 1975), fewer interactions with objects in their environment (Chavez & Martinez, 1984), and reductions in the variety of exploratory behaviours utilised (Meeks-Gardner, Grantham-McGregor, Himes, & Chang, 1999), as compared to more adequately nourished children. A similar pattern of ®ndings are shown for young children who are marginally undernourished, as seen in reduced alertness (Rahmanifar et al., 1993), and reductions in the amount (Sigman, Neumann, Baksh, Bwibo, & McDonald, 1989) and variation of play behaviours shown (Wachs et al., 1993). Signi®cant increases in duration of play behaviour have been shown for undernourished children receiving either an energy or a micronutrient (vitamin and mineral) supplement, as compared to unsupplemented children (Pollitt, Saco-Pollitt, Jahari, Husaini, & Huang, 2000). In addition, several authors have noted that the increased levels of negative affect and wariness seen in young children with protein±calorie malnutrition (Meeks-Gardner et al., 1999) or iron de®ciency anaemia (Lozoff et al., 1998) can also serve to reduce the child's level of interaction with the environment. Overall, while the evidence generally supports the validity of the hypothesis that nutritional de®ciencies act to reduce the child's level of involvement with its environment, results may vary depending on the measures of environmental involvement used or the child's gender (Walka, Triana, Jahari, Husaini, & Pollitt, 2000). As a result, questions still remain in regard to what aspects of environmental involvement are most sensitive to speci®c nutritional de®ciencies and why child gender can act to moderate observed results. In terms of the ®nal nutritionally driven pathway shown in Figure 4.1, links between nutritional de®ciencies and disturbances in parent±child relation patterns were discussed in a previous section of this chapter. Of particular relevance to this pathway are the studies by Lozoff and colleagues indicating that mothers of infants with iron-de®ciency anaemia, although initiating interaction more often with their infants, were less likely to encourage their infant in task situations (Lozoff et al., 1998) and were more likely to attempt to retain close physical contact when their infant moved away from them (Lozoff, Klein, & Prabucki, 1986). It could be argued that these results may re¯ect the in¯uence of less adequate nutrition

76

Wachs

of the caregiver rather than the infant. However, results from a crosscountry replication study indicated that the infant's level of nutritional intake predicted patterns of caregiver behaviour, even after statistically controlling for caregiver's level of nutrition (Wachs et al., 1992). Similarly, signi®cant differences occurred in maternal behaviour patterns toward iron-de®cient anaemic infants versus nonanaemic infants, even though virtually all of the mothers in the sample were nonanaemic (Lozoff et al., 1998). At present there are relatively few studies that have systematically investigated the various paths underlying functional isolation processes. More detailed investigations of nutritionally driven functional isolation mechanisms are an essential area for future research. Systematic research on functional isolation mechanisms will not only allow an increased understanding of how nutritional de®ciencies translate into behavioural de®cits, but will also have the added bene®t of encouraging active scienti®c collaboration among nutritional, environmental, and neuroscience researchers.

Understanding variability in individual food intake For the most part, when we think of the causes of nutritional de®cits in infancy and childhood, we tend to focus on ecological or economic factors, such as low food availability or a lack of economic resources to purchase what foods are available. However, individual food intake is not solely determined by food availability. Even when food is available, cultural, contextual, and individual characteristics can all act to determine which individuals get a more or less adequate share of nutritional resources. As shown in Figure 4.1, culturally driven patterns of family feeding practices and the degree of effort parents believe they need to expend in order to induce the child to eat have all been associated with differences in children's nutritional status (Engle, Zeitlin, Medrano, & Garcia, 1996; Zeitlin, 1991). Individual child characteristics are also important. Depending upon cultural preferences, evidence is available indicating that males (Zeitlin, 1996) or children with dif®cult temperaments (DeVries, 1984; ScheperHughes, 1987) are more likely to get their fair share of adequate resources. In addition, individual caregiver characteristics may also play a role. The age of the primary nonmaternal caregiver has been found to be positively related to the adequacy of young children's nutrition (Engle, 1991). Evidence also indicates better nutrition for children of more intelligent or more educated mothers, even after controlling for family demographic factors (Sandiford et al., 1997; Wachs & McCabe, 2001). These studies go beyond just ecological and economic causes and point to the importance of identifying which contextual and individual factors are particularly salient in determining who is likely to be adequately or inadequately nourished, as well as the processes governing such relations.

4. Nutritional de®ciencies

77

CONCLUSIONS In spite of a high incidence level, the study of relations between nutritional de®ciencies and children's behavioural development is a topic that has been relatively neglected by developmental scientists. Evidence documenting the number of children at risk for nutritional de®ciencies in both developed and developing countries, and the role played by nutritional de®cits in behavioural±developmental variability, emphasise the need for greater involvement by human developmental scientists in this area of research. In addition, such neglect has implications for the generalisability of theories that attempt to explain individual differences in developmental outcomes, especially in regard to generalisation of such theories to less developed countries where the majority of the world's children live and where nutritional de®cits are more likely to occur. Theories that attempt to explain how nutritional de®ciencies translate into developmental de®cits emphasise the importance of collaborative research between nutritional, contextual, and brain scientists. Such collaborative research is essential both for the generation of more comprehensive and generalisable developmental theories and for a greater applicability of principles derived from such theories to the optimisation of functional competence of children across the world. ACKNOWLEDGEMENTS Thanks are due to Professor Ernesto Pollitt for his comments on an initial draft of this chapter. REFERENCES Alpert, J., & Fava, M. (1997). Nutrition and depression: The role of folate. Nutrition Reviews, 55, 145±149. Barker, D., Gluckman, P., Godfrey, K., Harding, J., Owens, J., & Robinson, J. (1993). Fetal nutrition and cardiovascular disease in adult life. The Lancet, 341, 938±941. Barrett, D., Radke-Yarrow, M., & Klein, R. (1982). Chronic malnutrition and child behavior. Developmental Psychology, 18, 541±556. Bedi, K. (1987). Lasting anatomical changes following undernutrition during early life. In J. Dobbing (Ed.), Early nutrition and later achievement (pp. 1±36). Orlando, FL: Academic Press. Black, M. (1998). Zinc de®ciency and child development. American Journal of Clinical Nutrition, 68, 464s±469s. Bruner, A., Joffe, A., Duggan, A., Casella, J., & Brandt, J. (1996). Randomised study of cognitive effects of iron supplementation in non-anaemic iron de®cient adolescent girls. The Lancet, 348, 992±996.

78

Wachs

Chavez, A., & Martinez, C. (1984). Behavioral measurement of activity in children and their relation to food intake in a poor community. In E. Pollitt & P. Amante (Eds.), Energy intake and activity (pp. 303±321). New York: Liss. Chavez, A., Martinez, C., & Yaschine, J. (1971). Nutrition, behavioral development and mother±child interaction in young rural children. Federation Proceedings, 34, 1574±1582. Christensen, L. (1996). Diet±behavior relationships: Focus on depression. Washington, DC: American Psychological Association. Colombo, M., de la Parra, A., & Lopez, I. (1992). Intellectual and physical outcome of children undernourished in early life is in¯uenced by later environmental conditions. Developmental Medicine and Child Neurology, 34, 611±622. Connolly, K. (1998). Mental and behavioral effects of parasitic infection. In Nutrition, health and child development. Pan American Health Organization Scienti®c Monographs, 566 (pp. 138±161). Washington, DC: PAHO. Cook, J., & Martin, K. (1995). Differences in nutrient adequacy among poor and nonpoor children. Center on Hunger, Poverty and Nutrition Policy Monograph. Medford, MA: Tufts University School of Nutrition. de Andraca, I., Periano, P., & Uauy, R. (1998). Nutrition and care in the pre-term and neonatal periods and later development. In Nutrition, health and child development. Pan American Health Organization Scienti®c Monographs, 566 (pp. 43±68). Washington, DC: PAHO. DeVries, M. (1984). Temperament and infant mortality among the Masai of East Africa. American Journal of Psychiatry, 141, 1189±1194. Engle, P. (1991). Maternal work and child care strategies in peri-urban Guatemala: Nutritional effects. Child Development, 62, 954±965. Engle, P., Zeitlin, M., Medrano, Y., & Garcia, L. (1996). Growth consequences of low income Nicaraguan mothers theories about feeding one year olds. In S. Harkness & C. Super (Eds.), Parents' cultural belief systems (pp. 428±446). New York: Guilford Press. Evans, G. (1994). The psychological costs of chronic exposure to ambient air pollution. In R. Isaacson & K. Jensen (Eds.), The vulnerable brain and environmental risks. Vol. 3: Toxins in air and water (pp. 167±182). New York: Plenum. Felt, B., & Lozoff, B. (1996). Brain iron and behavior of rats are not normalized by treatment of iron de®ciency anemia during early development. Journal of Nutrition, 126, 693±701. Gardner, J., Grantham-McGregor, S., Chang, S., Himes, J., & Powell, C (1995). Activity and behavioral development in stunted and non-stunted children in response to nutritional supplementation. Child Development, 66, 1785±1797. Geber, M., & Dean, R. (1956). The psychological changes accompanying kwashiorkor. Courrier, 6, 3±15. Gorman, K. (1995). Malnutrition and cognitive development. Journal of Nutrition, 125, s2239±s2244. Goyer, R. (1995). Nutrition and metal toxicity. American Journal of Clinical Nutrition, 61, 646s±650s. Grantham-McGregor, S. (1984). Rehabilitation following clinical malnutrition. In J. Brozek & B. Schurch (Eds.), Malnutrition and behavior: Critical assessment of key issues (pp. 531±554). Lausanne, Switzerland: Nestle Foundation.

4. Nutritional de®ciencies

79

Grantham-McGregor, S. (1995). A review of studies of the effect of severe malnutrition on mental development. Journal of Nutrition, 125, 2233s±2238s. Grantham-McGregor, S., Chang, S., & Walker, S. (1998). Evaluation of school feeding programs: Some Jamaican examples. American Journal of Clinical Nutrition, 67, 785s±789s. Grantham-McGregor, S., Chang, S., Walker, S., & Powell, C. (1998). School feeding studies in Jamaica. In Nutrition, health and child development. Pan American Health Organization Scienti®c Monographs, 566 (pp. 104±118). Washington, DC: PAHO. Grantham-McGregor, S., Fernald, L., & Sethuraman, K. (1999a). The effects of health and nutrition on cognitive and behavioral development in children in the ®rst three years of life: Part 1. Low birth weight, breast feeding and protein energy malnutrition. Food and Nutrition Bulletin, 20, 53±75. Grantham-McGregor, S., Fernald, L., & Sethuraman, K. (1999b). Effects of health and nutrition on cognitive and behavioral development in children in the ®rst three years of life: Part 2. Infections and micronutrient de®ciencies: Iodine, iron and zinc. Food and Nutrition Bulletin, 20, 76±99. Grantham-McGregor, S., Lira, P., Ashworth, A., Morris, S., & Assuncao, A. (1998). The development of low birth weight term infants and the effects of the environment in Northeast Brazil. Journal of Pediatrics, 132, 661±666. Grantham-McGregor, S., Stewart, M., & Powell, C. (1991). Behavior of severely malnourished children in a Jamaican Hospital. Developmental Medicine and Child Neurology, 33, 706±714. Greenough, W., & Black, J. (1992). Induction of brain structure by experience. In M. Gunnar & C. Nelson (Eds.), Developmental behavioral neuroscience (pp. 155± 192). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Guilarte, T. (1993). Vitamin B-6 and cognitive development. Nutrition Reviews, 51, 193±198. Halterman, J., Kaczorowski, J., Aligne, A., Auinger, P., & Szilagyi, P. (2001). Iron de®ciency and cognitive achievement among school-aged children and adolescents in the United States. Pediatrics, 107, 1381±1386. Hurtado, E., Claussen, A., & Scott, K. (1999). Early childhood anemia and mild or moderate mental retardation. American Journal of Clinical Nutrition, 69, 115±119. Kaplan, B. (1972). Malnutrition and mental development. Psychological Bulletin, 78, 321±334. Karp, R. (1993). Malnourished children in the United States. New York: Springer. Keusch, G. (1990). Malnutrition, infection and immune function. In R. Suskind & L. Lewinter-Suskind (Eds.), The malnourished child (pp. 37±55). New York: Raven Press. Kleinman, J. (1992). The epidemiology of low birthweight. In S. Friedman & M. Sigman (Eds.), The psychological development of low birthweight children. Norwood, NJ: Ablex. Lester, B. (1975). Cardiac habituation of the orienting response to an auditory signal in infants of varying nutritional status. Developmental Psychology, 11, 432± 442. Levitsky, D., & Barnes, R. (1972). Nutritional and environmental interactions in behavioral development of the rat. Science, 176, 68±72. Levitsky, D., & Strupp, B. (1995). Malnutrition and the brain. Journal of Nutrition, 125, 2212s±2220s.

80

Wachs

Lozoff, B. (1998). Exploratory mechanism for poorer development in iron-de®cient anemic infants. In Nutrition, health and child development. Pan American Health Organization Scienti®c Monographs, 566 (pp. 162±178). Washington, DC: PAHO. Lozoff, B., Klein, N., Nelson, E., McClish, D., Manuel, M., & Chacon, M. (1998). Behavior of infants with iron de®ciency anemia. Child Development, 69, 24±36. Lozoff, B., Klein, N., & Prabucki, K. (1986). Iron-de®cient anemic infants at play. Developmental and Behavioral Pediatrics, 7, 152±158. Lozoff, B., & Wachs, T.D. (2001). Functional consequences of nutritional anemias in infancy and early childhood: Child development and behavior. In U. Ramakrishnan (Ed.), Nutritional anemias (pp. 69±88). Boca Raton, FL: CRC Press. Lynch, S. (1997). Interaction of iron with other nutrients. Nutrition Reviews, 55, 102±110. McCullough, A., Kirksey, A., Wachs, T.D., McCabe, G., Bassily, N., Bishry, Z., Galal, O., Harrison, G., & Jerome, N. (1990). Vitamin B-6 status of Egyptian mothers: Relation to infant behavior and maternal infant interactions. American Journal of Clinical Nutrition, 51, 1067±1074. Meeks-Gardner, J., Grantham-McGregor, S., Himes, J., & Chang, S. (1999). Behaviour and development of stunted and nonstunted Jamaica children. Journal of Child Psychology and Psychiatry, 40, 1±9. Merialdi, M., Caul®eld, L., Zavaleta, N., Figueroa, A., & DiPietro, J. (1998). Adding zinc to prenatal iron and folate tablets improves fetal neurobehavioral development. American Journal of Obstetrics and Gynecology, 180, 483±490. Moen, P., Elder, G., & Luscher, K. (1995). Examining lives in context. Washington, DC: American Psychological Association. Morgane, P., Austin-La France, R., Bronzino, J., Tonkiss, J., Diaz-Cintra, S., Cintra, L., Kemper, T., & Galler, J. (1993). Prenatal malnutrition and development of the brain. Neuroscience and Biobehavioral Reviews, 17, 91±128. Morgane, P., Kemper, T., Stern, W., Hawrylewicz, E., & Resnick, O. (1978). The effect of protein malnutrition on the developing central nervous system in the rat. Neuroscience and Biobehavioral Review, 2, 137±230. My Lien, N., Meyer, K., & Winick, N. (1977). Early malnutrition and later adoption. American Journal of Clinical Nutrition, 30, 1734±1735. Nelson, C., & Bloom, F. (1997). Child development and neuroscience. Child Development, 68, 970±987. Otero, G.A., Aguirre, D.M., Porcayo, R., & Fernandez, T. (1999). Psychological and electroencephalographic study in school children with iron de®ciency. International Journal of Neuroscience, 99, 113±121. Paine, P., Dorea, J., Pasquali, J., & Monteior, A. (1992). Growth and cognition in Brazilian school children. International Journal of Behavioral Development, 15, 160±183. Palti, H., Meijer, A., & Adler, B. (1985). Learning achievement and behavior at school of anemic and non-anemic infants. Early Human Development, 10, 217± 223. Parks, Y., & Wharton, B. (1990). Iron de®ciency and the brain. In J. Dobbing (Ed.), Brain, behavior and iron in the infant's diet. London: Springer. Pfeiffer, S., Norton, J., Nelson, L., & Shott, S. (1995). Ef®cacy of vitamin B-6 and

4. Nutritional de®ciencies

81

magnesium in the treatment of autism. Journal of Autism and Developmental Disorders, 25, 481±493. Pollitt, E. (1988). A critical view of three decades of research on the effects of chronic energy malnutrition on behavioral development. In B. Schurch & M. Scrimshaw (Eds.), Chronic energy de®ciency (pp. 77±93). Lausanne, Switzerland: IDECG. Pollitt, E. (1995). Functional signi®cance of the covariance between protein energy malnutrition and iron de®ciency anemia. Journal of Nutrition, 125, 2272s±2278s. Pollitt, E., Cueto, S., & Jacoby, E. (1998). Fasting and cognition in well and undernourished school children. American Journal of Clinical Nutrition, 67, 779s±785s. Pollitt, E., Gorman, K., Engle, P., Martorell, R., & Rivera, J. (1993). Early supplementary feeding and cognition. Monographs of the Society for Research in Child Development, 58 (17), 1±99. Pollitt, E., & Mathews, R. (1998). Breakfast and cognition: An integrative summary. American Journal of Clinical Nutrition, 67, 804s±813s. Pollitt, E., Saco-Pollitt, C., Jahari, A., Husaini, M., & Huang, J (2000). Effects of an energy and micronutrient supplement on play behavior in undernourished children in Indonesia. European Journal of Clinical Nutrition, 54, s80±s91. Popkin, B. (1994). The nutrition transition in low income countries: An emerging crisis. Nutrition Review, 52, 285±298. Popkin, B., & Doak, C. (1998). The obesity epidemic is a worldwide phenomenon. Nutrition Review, 56, 106±114. Powell, C., Walker, S., & Grantham-McGregor, S. (1998). Early childhood stimulation programs in Jamaica. In Nutrition, health and child development. Pan American Health Organization Scienti®c Monograph Number, 566 (pp. 198±208). Washington, DC: PAHO. Rahmanifar, A., Kirksey, A., Wachs, T.D., McCabe, G., Bishry, Z., Galal, O., Harrison, G., & Jerome, N. (1993). Diet during lactation associated with infant behavior and caregiver infant interaction in a semirural Egyptian village. Journal of Nutrition, 123, 164±175. Rao, R., & Georgieff, M. (2000). Early nutrition and brain development. In C. Nelson (Ed.), The effects of early adversity on neurobehavioral development (pp. 1±38). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Rasmussen, K. (2001). Is there a causal relationship between iron de®ciency or ironde®ciency anemia and weight at birth, length of gestation and perinatal mortality? Journal of Nutrition, 131, 590s±603s. Ricciuti, H. (1993). Nutrition and mental development. Current Directions in Psychological Science, 2, 43±46. Sameroff, A. (1994). Ecological perspectives on longitudinal follow-up studies. In S. Friedman & C. Haywood (Eds.), Developmental follow-up (pp. 45±66). San Diego, CA: Academic Press. Sandiford, P., Cassel, J., Sanchez, G., & Coldham, C. (1997). Does intelligence account for the link between national literacy and child survival. Social Science and Medicine, 45, 1231±1239. Sandstead, H. (1995). Requirements and toxicity of essential trace elements illustrated by zinc and copper. American Journal of Clinical Nutrition, 61, 621s± 624s. Sandstead, H., Penland, J., Alcock, N., Dayal, H., Chen, X., Li, J., Zhao, F., &

82

Wachs

Yang, J. (1998). Effects of repletion with zinc and other micronutrients on neuropsychologic performance and growth of Chinese children. American Journal of Clinical Nutrition, 68, 470s±475s. Sazawal, S., Bently, M., Black, R., Dhingra, P., George, S., & Bhan, M. (1996). Effect of zinc supplementation on observed activity in low socioeconomic Indian preschool children. Pediatrics, 98, 1132±1137. Scheper-Hughes, N. (1987). Culture, scarcity and maternal thinking. In N. ScheperHughes (Ed.), Child survival (pp. 187±208). Dordrecht, The Netherlands: Reidel. Schore, A. (1994). Affect regulation and the origin of the self. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Sigman, M., Espinosa, M., & Whaley, S. (1998). Mild malnutrition and the cognitive development of Kenyan school children. In Nutrition, health and child development. Pan American Health Organization Scienti®c Monographs, 566 (pp. 91±103). Washington, DC: PAHO. Sigman, M., Neumann, C., Baksh, M., Bwibo, N., & McDonald, M. (1989). Relation between nutrition and development in Kenyan toddlers. Journal of Pediatrics, 115, 357±364. Silver, A. (1992). Nutritional aspects of memory dysfunction. In J. Morley, R. Coe, R. Strong, & J. Grossberg (Eds.), Memory function in aging related disorders (pp. 202±212). New York: Springer. Simeon, D., & Grantham-McGregor, S. (1990). Nutritional de®ciencies and children's behavior and mental development. Nutrition Research Review, 3, 1±24. Sklar, R. (1986). Nutritional vitamin B12 de®ciency in a breast-fed infant of a vegan-diet mother. Clinical Pediatrics, 25, 219±221. Spurr, G. (1987). Effects of chronic energy de®ciency on stature, work capacity and productivity. In B. Schurch & N. Scrimshaw (Eds.), Chronic energy de®ciency: Consequences and related issues (pp. 95±134). Lausanne, Switzerland: IDECG. Stein, Z., Susser, M., Saenger, G., & Marolla, F. (1975). Famine and human development. London: Oxford University Press. Stoltzfus, R. (2001). De®ning iron-de®ciency anemia in public health terms: A time for re¯ection. Journal of Nutrition, 131, 565s±567s. Strupp, B., & Levitsky, D. (1995). Enduring cognitive effects of early malnutrition. Journal of Nutrition, 125, 2221s±2232s. Stunkard, A.J. (2000). Factores determinantes de la obesidad: OpinioÂn actual. In M. PenÄa & J. Bacallao (Eds.), La obesidad en la pobreza, OrganizacioÂn Panamericana de la Salud PubbicacioÂn Cientõ®ca No. 576 (pp. 27±32). Super, C., Clement, J., Vuori, L., Christiansen, N., Mora, J., & Herrera, M. (1981). Infant and caretaker behavior as mediators of nutritional and social intervention in the barrios of Bogota. In T. Field (Ed.), Culture and early interaction (pp. 171± 188). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Tacconi, M., Calzi, F., & Salmona, M. (1997). Brain lipids and diet. In M. Hilbrand & R. Spitz (Eds.), Lipids, health and behavior (pp. 197±226). Washington, DC: American Psychological Association. Toller, S. (1992). Biochemistry individual differences and psychonutrition. In A. Gale & M. Eysenck (Eds.), Handbook of individual differences: Biological perspectives. New York: John Wiley. Uauy, R., Peirano, P., Hoffman, D., Mena, P., Birch, D., & Birch, E. (1996). Role of essential fatty acids in the function of the developing nervous system. Lipids, 31, 167s±176s.

4. Nutritional de®ciencies

83

UNICEF. (1998). The state of the world's children 1998. Malnutrition: causes, consequences and solutions. Nutrition Reviews, 56, 115±123. United Nations Subcommittee on Nutrition. (1992). Second report on the world nutrition situation, Vol. 1. Geneva: World Health Organization. Valenzuela, M. (1997). Maternal sensitivity in a developing society: The context of urban poverty and infant chronic undernutrition. Developmental Psychology, 33, 845±855. Wachs, T.D. (1995). Relation of mild to moderate malnutrition to human development: Correlational studies. Journal of Nutrition, 125, 2245s±2254s. Wachs, T.D. (1999). The nature and nurture of child development. Food and Nutrition Bulletin, 20, 7±22. Wachs, T.D. (2000a). Linking nutrition and temperament. In V. Molfese & D. Molfese (Eds.), Temperament and personality development across the life span (pp. 57±84). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Wachs, T.D. (2000b). Necessary but not suf®cient: The respective roles of single and multiple in¯uences on human development. Washington, DC: American Psychological Association. Wachs, T.D., Bishry, Z., Yunis, F., McCabe, G., Swei®, E., Harrison, G., Shaheen, F., Galal, O., Jerome, N., & Kirksey, A. (1995). Nutritional intake and context as predictors of cognition and adaptive behavior of Egyptian school age children. International Journal of Behavioral Development, 18, 425±450. Wachs, T.D., & McCabe, G. (2001). Relation of maternal intelligence and schooling to offspring nutritional intake. International Journal of Behavioral Development, 25, 444±449. Wachs, T.D., Moussa, W., Bishry, Z., Yunis, F., Sobhy, A., McCabe, G., Jerome, N., Galal, O., Harrison, G., & Kirksey, A. (1993). Relations between nutrition and cognitive performance in Egyptian toddlers. Intelligence, 17, 151±172. Wachs, T.D., Sigman, M., Bishry, Z., Moussa, W., Jerome, N., Neumann, C., Bwibo, N., & McDonald, M. (1992). Caregiver child interaction patterns in two cultures in relation to nutrition. International Journal of Behavioral Development, 15, 1±18. Walka, H., Triana, N., Jahari, A., Husaini, M., & Pollitt, E. (2000). Effects of an energy and micronutrient supplement on play behavior in undernourished children in Indonesia. European Journal of Clinical Nutrition, 54, s91±s106. Watkins, W., & Pollitt, E. (1998). Iron de®ciency and cognition among school aged children. In Nutrition, health and child development. Pan American Health Organization Scienti®c Monographs, 566 (pp. 179±197). Washington, DC: PAHO. Wauben, P.M., & Wainright, P.E. (1999). The in¯uence of neonatal nutrition on behavioral development: A critical appraisal. Nutrition Reviews, 57, 35±44. Willatts, P., Forsyth, J., DiModugno, M., Varma, S., & Colvin, M (1998). Effect of long-chain polyunsaturated fatty acids in infant formula on problem solving at 10 months of age. The Lancet, 352, 688±691. Winick, M., Meyer, K., & Harris, R. (1975). Malnutrition and environmental enrichment by early adoption. Science, 190, 1173±1175. Yehuda, S., Rabinovitz, S., Casasso, R., & Mostofsky, D. (1998). Fatty acids and brain peptides. Peptides, 19, 416±419. Yehuda, S., Rabinovitz, S., & Mostofsky, D (1997). Effects of essential fatty acid preparation (SR-3) on brain lipids, biochemistry and behavioral and cognitive

84

Wachs

functions. In S. Yehuda & D. Mostofsky (Eds.), Handbook of essential fatty acid biology (pp. 427±452). Totowa, NJ: Humana Press. Zeitlin, M. (1991). Nutritional resilience in a hostile environment. Nutrition Reviews, 49, 259±268. Zeitlin, M. (1996). My child is my crown: Yoruba parental theories and practices in early childhood. In S. Harkness & C. Super (Eds.), Parents' cultural belief systems (pp. 407±429). New York: Guilford Press.

5

Prospects for research on the quality of pre-school experience Edward C. Melhuish

Institute for the Study of Children, Families & Social Issues, Birkbeck College, University of London, UK

INTRODUCTION In the later part of the second millennium, research into child care and preschool provision attracted considerable attention for theoretical, social, and political reasons (e.g., Melhuish & Moss, 1990; Fox & Fein, 1990, ClarkeStewart, 1989). A strong theme of much research on pre-school provision has been the importance of quality of care. Differences in children's experiences in different pre-school settings may have developmental consequences, and be part of the explanation of variations in the bene®ts of pre-school experience. There are alternative perspectives on the quality of care and the emphasis placed on these different perspectives largely re¯ects political ideology. Katz (1995) considered quality in terms of the perspective taken, i.e., whether the perspective is that of parents, children, caregivers, or community. Similarly Woodhead (1996) developed a conceptual framework with three dimensions, namely the stakeholders, bene®ciaries, or indicators of quality. By considering speci®c stakeholders, speci®c bene®ciaries, and different indicators, it is possible to arrive at a multitude of different perspectives on quality. Munton, Mooney, and Rowland (1995) constructed a two-dimensional framework within which different de®nitions might be deconstructed as a multidimensional concept. All these frameworks for quality are responses to the fact that quality is a value-laden term and they allow for the interrelationship of different perspectives to be taken into account. The existence of multiple perspectives on quality does not invalidate any one approach to quality, as long as the values underlying the approach are recognised. Most research on quality in pre-school care and education adopts the perspective that quality re¯ects what is bene®cial for children's development. The judgement of what is bene®cial might be based upon expert judgement or based upon empirical research of the relationship between indicators of quality and developmental outcomes. The measurement of quality needs to become more substantially based upon indicators with a demonstrated relationship with developmental outcomes and less based on ``expert'' opinion of what should be best for children. This approach enables the concept of quality to advance beyond conjecture.

86

Melhuish

This chapter takes a selective review of recent research and formulates a view of the dilemmas and possible solutions for research to progress. First it examines the developmental effects of pre-school experience and why quality of care is an important issue. Quality of care measurement is a major issue, not only for psychometric reasons, but also because the nature of measurement objecti®es theoretical assumptions about what quality is. Hence this chapter considers how quality is measured and how it might be measured in the future. DEVELOPMENTAL EFFECTS Some pre-school provision has explicit educational aims and is usually targeted on children from 3 years upwards (e.g., nursery schools, kindergarten). Other forms of provision are more care-orientated and, while operating for under-threes, also cater for older children (e.g., day care centres). There is an overlap between the care- and education-orientated sections with the distinction becoming increasingly blurred. Moreover, as the importance of the under-threes environment for fostering learning becomes more apparent, the blurring of the distinction between care and education will accelerate. Within research on pre-school care and education, the importance of quality has become pre-eminent (Melhuish, 1993). This is partly the result of sociopolitical fashions of what is appropriate to emphasise and partly because of research evidence. A comprehensive review of studies of quality of child care within the USA is given by Vandell and Wolfe (2000). High-quality care and education (based on various measures) has been associated with short-term and long-term cognitive, social, and emotional bene®ts for children's development. The strongest effects appear to occur for children from disadvantaged backgrounds (Bruner, 1980; Clark, 1988; Meisels & Shonkoff, 1990; Peisner-Feinberg et al., 2001; Pugh, 1992; Reifel, 1993; Sylva, 1990; Zigler & Styfco, 1993). The studies involved have largely been based in the USA but some studies (e.g., Andersson, 1992; Beller, 1993; Hadeed, 1994; Johansson, 1993; Kagitcibasi, Sunar, & Bekman, 1988; Petrogiannis & Melhuish, 1996) have occurred in other countries, indicating that the results are not culture-speci®c; it is clear, though, that cultural issues need to be given greater weight in this area (Rosenthal, 1999). Cost-effectiveness of pre-school programmes has also been studied in terms of long-term gains related to crime, welfare dependence, unemployment, teenage pregnancy, school dropout, and family instability. The long-term bene®ts for society apparently can easily justify additional costs of pre-school provision according to the limited evidence available (Barnett & Escobar, 1990; Schweinhart, Barnes & Weikhart, 1993). It should be noted that this evidence is highly dependent on the costs associated with later criminal activity for small numbers of individuals. Future research is likely to

5. The quality of pre-school experience

87

increasingly focus on cost±bene®t issues as measurement in terms of quality, developmental outcomes, and costs improve. The High/Scope Perry Pre-school Project (Berrueta-Clement, Schweinhart, Barnett, Epstein, & Weikart, 1984; Schweinhart et al., 1993) has been the most extensively evaluated study of pre-school effects and has a strong experimental design. The pre-school programme is noted for its high quality and includes speci®c adult±child interactions to guide children's experience towards achieving their potential. A follow-up of children in this project at 27 years of age found evidence of continuing developmental bene®ts, and a cost±bene®t analysis indicated the potentially high ®nancial payoff for society of pre-school education: ``$7.16 returned for every one dollar invested'' (Schweinhart et al., 1993). The Head Start initiative was a source of many studies that examined the effects of pre-school education. McKey, Condelli, Barrett, McConkey, and Plantz (1985) developed a meta-analysis of 210 studies evaluating Head Start programmes. They concluded that Head Start programmes have an immediate positive effect on child development, but these effects ``wash out'' after 2 years. However, many studies were poorly controlled. In one of the more robust studies, Lee, Brooks-Gunn, and Schnur (1988) considered data on 969 children and 19 pre-schools. Some children had attended a Head Start pre-school, some had attended other pre-schools, and some children had not attended pre-school. They found evidence of bene®cial pre-school effects for Head Start upon cognitive measures, with the greatest effects occurring for the most disadvantaged children. The Consortium of Longitudinal Studies (Lazar, Darlington, Murray, Royce, & Snipper, 1982) considered 11 studies, which included centrebased, home-based, and centre/home-based programmes, with children from disadvantaged backgrounds. Quasi-experimental and approximately randomised designs were used to assess a variety of cognitive, motivational, and social development outcomes. The meta-analysis across the 11 studies provided strong evidence of lasting bene®cial effects for pre-school education. Gains in cognitive development were evident for several years, but not at 19 years of age. However, differences in motivational aspects of social behaviour were still present. A number of other studies have found similar effects with children from disadvantaged backgrounds (e.g., Fuerst & Fuerst, 1993; Garber, 1988; Ramey, Bryant, Campbell, Sparling, & Wasik, 1988). Sylva (1994) concluded in her review that strong experimental studies support the claims that pre-school experiences actually cause relatively lasting bene®ts for children's development. Rutter (1985) concluded in reviewing such research: ``the long-term educational bene®ts stem not from what children are speci®cally taught but from effects on children's attitudes to learning, their self-esteem and on their task orientation''. The pre-school programmes that were most consistently associated with positive developmental effects were regarded as high-quality programmes. For instance, in terms of curriculum, Jowett and Sylva (1986) found

88

Melhuish

nursery education graduates did better in primary school than playgroup graduates, suggesting the value of an educationally orientated pre-school for fostering educationally relevant outcomes. Schweinhart (1987) has documented what he sees as quality indicators for pre-school education validated by research. These are:

· · · · · ·

Developmentally appropriate curriculum that features child-initiated learning activities within a supportive environment. Careful selection of staff and effective training. Staff±child ratios. Partnership with parents and community. Strong support services, administration, health and nutrition. Procedures for evaluating children's progress.

Lazar et al. (1982) suggest that there may be mutual reinforcement between pre-school participation by parents and parents' attitudes. A virtuous circle is seen to operate in promoting better motivation for the children to learn, resulting in long-term gains. Although the results from research on pre-school education have been fairly consistent, the research evidence on the effects of day care upon development has been equivocal, with some studies ®nding negative effects, some no effects, and some positive effects. Partly discrepant results relate to age of starting. Andersson (1989), in a Swedish longitudinal study of development related to day care and pre-school education, found that children who had started early in high-quality centre-based care had better academic and social development than children who had less pre-school experience in these high-quality centres. Field (1991) reported similar evidence for the USA. Besides age, the discrepancy between results from different studies probably relates at least partly to differences in the quality of day care received by children. Some studies have indicated that the quality of care may be related to language development (e.g., Goelman & Pence, 1987; Melhuish, Lloyd, Martin, & Mooney, 1990a). With regard to social development, Howes and Olencik (1986) compared children attending high-quality day care, low-quality day care, and no day care, controlling for family background. Interestingly, the children from the high-quality day care were found to be the most socially compliant. Similar links between quality of day care and social adjustment have been reported elsewhere (Campbell, Lamb, & Hwang, 2000; Hutt, Tyler, Hutt, & Foy, 1989; Lamb, Sternberg, Hwang, & Broberg, 1992; Vandell, Henderson, & Wilson, 1988). The effects of pre-school care and education are mediated by family background. The greatest effects for high-quality day care occur for those children from the most disadvantaged homes (Schlieker, White, & Jacobs, 1991). Melhuish (1991) produced a model of the interrelationship of quality of care at home and in pre-school care. This work predicted that negative,

5. The quality of pre-school experience

89

neutral, and positive effects will occur depending on the relative balance of quality of care at home and in day care. Where the quality of care in both settings is equivalent there is likely to be little effect from attending preschool care. However, where quality of care in pre-school care is superior to the home, then children receiving day care are likely to show bene®cial effects, as seen in several studies of children from disadvantaged families receiving good-quality day care (e.g., Golden et al., 1978; O'Connell & Farran, 1982; Ramey & Mills, 1977). However, where the quality of out-ofhome care is inferior to that at home, then negative or detrimental effects may occur, as reported in at least one study in both the USA and the UK (e.g. Desai, Chase-Lansdale & Michael, 1989; Melhuish et al., 1990a). The most signi®cant study concerning the effects of pre-school care has been the NICHD study, which has been following over 1100 children from nine states of the USA from birth onwards (NICHD, 2000). In analyses of the effects of child care experience upon children's development at 4Ý years of age, even after controlling for multiple child and family characteristics, distinct effects were found related to early child care experience. These effects were related to the quality of care, the quantity of care, and also the type of care. There were developmental risks and developmental bene®ts associated with child care experience. Quality of care and experience in group settings (type of care) was found to be related to better educationaloriented outcomes and language performance. Greater quantity, i.e. more hours of care, was associated with higher levels of behaviour problems as rated by caregivers. The issue of whether infant day care was bad for children re-emerged in the 1980s with the proposition that infant day care may be a risk factor for insecure attachment to the mother (Belsky, 1986, 1988, in press). This proposition is in an ideologically and politically sensitive ®eld, and the resulting political and scienti®c debate led to the funding of one of the largest studies of day care, the NICHD study. The NICHD study could ®nd no direct or main effect of amount, quality, or type of day care on attachment security. However, low maternal sensitivity combined with poor-quality day care was associated with an increased risk of insecure attachment (NICHD, 1997). This suggested that poor-quality day care can be a risk factor, as proposed by Belsky (1988), and conversely that highquality day care may be a protective factor. Later reports from this study indicated that the quantity of care was related to caregiver-reported problems at 2 years and 4Ý years of age (NICHD, in press). Quality of care was the child care variable related to the widest range of child outcomes at 2 and 3 years of age and was predictive of cognitive development, social competence, compliance, and behaviour problems (NICHD, 1998). There is also a ®nding that whereas quantity of child care is associated with less maternal engagement with children, higher-quality child care is associated with greater maternal sensitivity (NICHD, 1999c). Presumably these maternal characteristics in¯uence the development of child competencies.

90

Melhuish

This ®nding of associations of child care and maternal characteristics illustrates the problem of dissociating the effects of child care variables from those of family variables when there are subtle and complex interactions between these classes of variable. The NICHD study has reinforced the evidence that quality of day care is important (NICHD, 1999a) despite indications that the sampling procedures of this study may under-represent poor-quality care and hence lead to an underestimation of the effects of quality of care (NICHD, 1998). The quality of care, particularly language stimulation, was related to cognitive and language developments at 15, 24, 36, and 54 months of age (NICHD, 2000, in press). This replicates earlier ®ndings on the impact of quality of care on language development (Goelman & Pence, 1987; McCartney, 1984) and speci®cally the quality of the language environment (Melhuish et al., 1990a). The NICHD study did some interesting work on the comparative importance of quality and quantity of care by making comparisons with the effect sizes on child development outcomes at 4Ý years, for those other major environmental factors of parenting and poverty. Quantity effects were primarily in the socio-emotional domain, e.g., behaviour problems. Quality effects were primarily in the cognitive domain, e.g., language development. The effect size of quantity of child care for behaviour problems was slightly less than that of poverty and slightly more than the effects of parenting. The effect size of quality of child care for pre-academic (educationally oriented) skills and language development was a lot less than that of parenting. It was around a third to a half as large as the effect size for parenting, which includes both shared genes and environmental in¯uences. However, the effect sizes for quantity (for behaviour problems) and quality of child care (for cognitive outcomes) were of a similar order of magnitude (NICHD, in press). However, it is noteworthy that the effects of quantity primarily related to the socio-emotional domain of development (e.g., behaviour problems) whereas the effects of quality primarily related to the cognitive domain (pre-academic and language skills). This pattern of results reinforces the general ®nding that quality effects are more common for cognitive factors even though some evidence of effects upon social development does exist. In considering the relative importance of preschool and family factors it must be remembered that family effects incorporate genetic factors. Hence, family and child care effects are likely to be much more equivalent in terms of environmental in¯uence than a straightforward comparison implies. Also family factors and child care quality covary, low-income families tending to have the lowest-quality care (Whitebook, Howes, & Phillips, 1989). The analysis strategy of the NICHD study, in common with many studies, attributes variance to child care factors only after family factor variance has been extracted. Where the two covary this tends to produce underestimates or conservative estimates of child care effects. One of the lessons of the NICHD study is that focusing on one facet of child care, e.g., quality or quantity or type, may be

5. The quality of pre-school experience

91

misleading in that different facets may all have in¯uences, but on different aspects of development. A surprising result of the NICHD study is the lack of any signi®cant child care by family interactions, i.e. the effect of quality of care varying for different types of family. There are main effects for quantity, quality, and type of child care. These effects are independent of family factors and show no interactions. Also there is no signi®cant interaction between quantity and quality of child care. This lack of signi®cant interactions may be a consequence of the range of quality within the study. This aspect of the NICHD results is not easily compatible with earlier ®ndings of greater effects for high-quality child care for disadvantaged children (e.g., Burchinal, Peisner-Feinberg, Bryant, & Clifford, 2000; Peisner-Feinberg et al., 2001) and conceptual approaches predicting such interactions (e.g., Melhuish, 1991). MEASURES OF QUALITY OF CARE Most approaches to quality focus on the setting, i.e. a centre, a day care home. At this level the measure of quality indicates what is typical for children in that setting. A common distinction made in considering quality of pre-school provision is between process and structural characteristics. Process characteristics refer to actual experiences of the child (e.g., interactions, language stimulation) and structural characteristics refer to aspects such as adult:child ratio, quali®cations of caregivers, group size. Often process and structural characteristics are correlated but this is not necessarily so. Some measures of quality focus on process measures and some on structural measures, but commonly a measure will combine both sorts of characteristics. A thorough review of observational measures of child care is provided by Barnes (2001). The most widely used method is the Early Childhood Environment Rating Scale (Harms & Clifford, 1980). This instrument has 37 items combined in seven subscale scores, which in turn produce a total score. Most studies analyse results in terms of total score and subscales scores. Following ECERS, a family of similar instruments has been developed. These include the Infant/Toddler Environmental Rating Scale (ITERS; Harms, Cryer, & Clifford, 1990), and the Family Day Care Rating Scales (FDCRS; Harms & Clifford, 1989). Recently the ECERS has been revised to produce ECERS-R (Harms, Clifford, & Cryer, 1998) with less ambiguous scoring. The revision improves an already good instrument. Although ECERS-R covers many aspects of the child's environment, it does not cover some important curricular aspects. Hence ECERS-E was produced (Sylva, Sammons, Melhuish, Siraj-Blatchford, & Taggart, 1999a), which uses a similar strategy to ECERS-R to produce subscale scores for literacy, mathematics (numeracy), science and the environment, and diversity. This extension to ECERS-R is particularly applicable to settings for 3± 6-year-old children, which have an educational purpose.

92

Melhuish

Other observational methods of assessing the quality of care within a setting include the Quality of Day Care Environment Scale (QDCE; Bradley, Caldwell, Fitzgerald, Morgan, & Rock, 1996), the Daycare Quality Assessment Inventory (DQAI; Peterson & Peterson, 1986), the Early Childhood Classroom Observation Scale (ECCOS; Bredekamp, 1986), the Assessment Pro®le for Early Childhood Programs (Abbot-Shim & Sibley, 1987), and the Child Care Facility Schedule (WHO, 1990). These observational techniques were preceded by an interview-based assessment, the Day Care Environmental Inventory (DCEI; Prescott, Kritchevsky, & Jones, 1972). Whereas these methods focus on the settings, some other methods provide an assessment of speci®c caregiver functioning. Arnett (1989) has produced a 26-item rating scale comprising four sub-scales of positive interaction, punitiveness, permissiveness, and detachment. This instrument is sometimes used as a summary measure of several caregivers, often in conjunction with the ECERS. Other observational measures of speci®c caregivers include the Adult Involvement Scale (AIS; Howes & Stewart, 1987), a measure of caregiver sensitivity devised by Goossens and Melhuish (1996) and a measure of interaction style devised by McGuire and Richman (1987). Although the observation methods that focus on settings are the most commonly used, they have the disadvantage that the experience of individual children within one setting may vary substantially. Methods that focus on individual children overcome this disadvantage but are usually more expensive in terms of researchers' time. One example of such focal child observational methods is described by Melhuish, Mooney, Martin, and Lloyd (1990b). The behaviours of a target child and others interacting with the target child are coded sequentially. Observations have a sequence and time base and patterns of interaction, particularly the responsiveness of one participant to another, can be measured. Similar methods have been developed by Pierrehumbert et al. (1996), called Observation du Lieu de Vie de l'Enfant (OLIVE), and NICHD (1996), called Observation of Speci®c Children and Caregivers (ORCE). Another focal child observation method much used in pre-school research is the Target Child Observation (Sylva, 1997). These focal child methods are particularly adapted to following the experience of speci®c children. Hence, they give more accurate information for the speci®c child, but the results may not generalise to other children in that setting. McCartney and Jordan (1990) have noted the similarities between the aims of research into child care and school effects. Both ®elds address the issue of the relationship between child development and nonfamily environments. Both areas have gone through three waves of research. The ®rst wave considered the question of whether the nonfamily environment (child care or school) matters. The second wave recognised the diversity in the non-family environments, and issues of quality of environment arise. The third wave considers the interrelationship between family and non-

5. The quality of pre-school experience

93

family environments and how the interaction may affect child development. In arriving at the most recent paradigms for research, both ®elds have come to adopt hierarchical models of children nested within families, families within settings, settings within cultures (communities), etc. The hierarchical nature of the theoretical models required particular methods of statistical analysis such as multilevel modelling (Goldstein, 1987, 1995) or hierarchical linear modelling (Bryk & Raudenbush, 1992). These statistical techniques allow research to address the question: ``Does a particular institution attended by a child make a difference?'' This question can be considered if suf®cient children from each institution are studied. In order for the statistical analysis to identify speci®c institution effects, there should be suf®cient children for each institution (around 20). Having suf®cient children within any one setting enables the multilevel modelling (or hierarchical linear modelling) to assign a level of effect to that institution. Hence most studies, even one as large as the NICHD study, do not have an adequate sample for utilising such an approach. A limitation of this approach is that many children experience care in settings with fewer than 20 children. Such situations may need other approaches. A major advantage is the identi®cation of speci®c institutions having a marked positive or negative effect upon child development outcomes, i.e., the effectiveness of speci®c centres or schools. The effectiveness of an institution can be taken as a measure of the quality, where quality is de®ned in terms of bene®cial effects upon child development. Measures of quality commonly used in child care and pre-school research are measures of characteristics presumed by ``experts'' to be representative of quality. To move to a de®nition of quality that is rooted in demonstrable bene®cial effects on child development is an advance that is now required if this research ®eld is to make advances in theory and provide evidence of suf®cient quality to guide the formation of social policy. Once a speci®c effective (high-quality) or ineffective (low-quality) institution has been identi®ed, the characteristics and processes involved in producing that level of effectiveness or quality can be investigated. Qualitative research methods may be particularly appropriate. Such research methods upgrade current methodologies by: 1 2

Providing an external validation of the measurement of quality. Allowing the identi®cation of speci®c high- or low-quality centres, which could then be studied intensively in order to understand the processes underlying quality of provision.

A study using such an approach in the area of child care and pre-school provision has recently started (Sylva, Sammons, Melhuish, Siraj-Blatchford, & Taggart, 1999b). The Effective Provision of Pre-school Education (EPPE) project and the Effective Pre-school Provision in Northern Ireland (EPPNI) project are linked projects sharing a common methodology.

94

Melhuish

Together these longitudinal studies are following over 4000 children from age 3 to age 7 years. Retrospective information is gathered by parental interview on child care before age 3 years. Extensive data on current child care and pre-school experience for each child from age 3 years is gathered by interviews and observations within pre-school and child care settings. Assessments of cognitive and social/behavioural development occur for all children at entry to the study (age 3), at the start of primary school (age 5), and after 1 and 2 years of primary school (6 and 7 years). The research design allows the identi®cation of the effectiveness of speci®c centres. Centres identi®ed as varying in effectiveness are currently being studied by intensive case study methods. Because these case studies are focused on centres with a known level of effect upon child development variables, they can examine the crucial differences associated with effects more thoroughly, in the hope of determining the processes leading to such differences between centres. This type of information is currently missing from the research ®ndings in this area. Reports on the study sample at the start of the study are available. Sammons, Sylva, Melhuish, SirajBlatchford, and Taggart (1999) have shown how multilevel modelling techniques can be applied to evaluate the relative contribution of speci®c centres, and types of centre, when child and family factors are controlled. Melhuish, Sylva, Sammons, Siraj-Blatchford, and Taggart (1999) have shown how child and family characteristics vary with type of pre-school and socio-economic status. For example, the variation in the educational environment of the child's home and in the range of child care experience before age 3 is particularly striking, and related to type of pre-school and socio-economic variation. Melhuish, Sylva, Sammons, Siraj-Blatchford, and Taggart (2001) show how these aspects of child, family, and child care experience, up to age 3, relate to cognitive and social/behavioural development. There are now reports from this study showing how speci®c centres have distinctly different effects upon child development variables, after allowing for a wide range of background child and family characteristics (Sylva, Melhuish, Sammons, Siraj-Blatchford & Taggart, 2001). The ``centre effects'' related to pre-literacy, language, numeracy, and nonverbal cognitive skills, with the effects being strongest for pre-literacy skills. Further analyses are exploring the evidence regarding ``centre effects'' for social/ behavioural aspects of development. Such ``centre effects'' are amongst the strongest available evidence of the in¯uence of pre-school experience upon children's development. Where a study considers the quality of a setting in terms of both observational ratings of child experience (e.g., ECERS) and effects on developmental outcomes, different settings may appear to be of high quality according to the different criteria. A centre rated highly on ECERS may not be associated with improved developmental progress, whereas a centre rated mediocre on ECERS may be associated with developmental progress signi®cantly above that to be expected from the characteristics of the

5. The quality of pre-school experience

95

intake. If developmental progress is measured in terms closely linked to educational progress, then it is likely that those settings where experiences relevant to educational outcomes, e.g., literacy and numeracy, are emphasised will have children showing superior developmental progress. In terms of developmental progress such a setting would be of high quality, yet it may not have much emphasis on the items in a scale such as ECERS. Hence in terms of ``expert opinion'' that setting may be of low quality. The results of the EPPE study reveal this dilemma. Although some aspects of ECERS and related measures show some relationship with developmental outcomes, the effects are small in relation to the total effects of child care environments (Sylva et al., 2001). This raises the issue of values, i.e. what values are placed on different development outcomes, and what is the relative value of children's experience and their developmental progress. Such ®ndings will require a reconceptualisation of the issue of quality in children's environments. In summary, the studies of the second millennium con®rm the importance of quality of care and education in the ®elds of early years day care and pre-school education. Research strategies are moving toward an integration of approaches to quality that incorporate validation of the construct through direct links with developmental outcomes. Large-scale longitudinal studies are starting to provide detailed evidence of the links between developmental outcomes and alternative measures of quality of child care. It is becoming clear that some aspects of these measures, based on ``expert opinion'', are not predictive of child outcomes. Are these measures to be abandoned or kept as measures of ``good'' environments for children, with no consequence for developmental outcomes? It is possible that future research may reveal links with developmental outcomes, but the emerging evidence will produce a new synthesis of the relationship between environmental characteristics, child experience, and developmental outcomes. Also, it is now the norm for children in most industrialised societies to experience some form of child care/education in the early years. Hence, it is appropriate for all longitudinal studies to incorporate measurement of this experience. It is likely, therefore, that much future research in child care will be part of larger longitudinal studies. This is already happening. The NICHD study sampling frame was essentially that of a longitudinal study, which incorporated extensive child care measures, with no selection made on the basis of child care. Within longitudinal studies advances in genetic research may enable the incorporation of genetic information into the analysis of developmental patterns and in this way the analysis of child care experience will contribute to the perennial issue of gene/environment interaction. Overall the improvement of measures of children's experience or quality in child care facilitate the incorporation of such measures into mainstream developmental studies, integrating the study of child care more ®rmly within developmental psychology.

96

Melhuish

REFERENCES Abbot-Shim, M., & Sibley, A. (1987). Assessment pro®le for early childhood programs: Pre-school, infant and school age. Atlanta, GA: Quality Assist. Andersson, B. (1989). Effects of public day care: A longitudinal study. Child Development, 60, 857±866. Andersson, B. (1992). Effects of day care on cognitive and socio-emotional competence of thirteen-year-old Swedish schoolchildren. Child Development, 63, 20±36. Arnett, J. (1989). Caregivers in day care centres: Does training matter? Journal of Applied Developmental Psychology, 10, 541±552. Barnes, J. (2001). Using observations to evaluate paid child care settings. In K. Petrogiannis & E.C. Melhuish (Eds.), International perspectives on pre-school research [in Greek]. Athens: Kastaniotis. Barnett, W.S., & Escobar, C.M. (1990). Economic cost and bene®ts of early intervention. In S. Meisels & J. Shonkoff (Eds.), Handbook of early childhood intervention. Cambridge: Cambridge University Press. Beller, E. (1993). What should early childhood education aim to do for children? What is the effect on young children's lives? Paper presented at Third European Conference on the quality of early childhood education. Kriopigi, Greece: European Early Childhood Education Research. Belsky, J. (1986). Infant day care: A cause for concern? Zero to Three, 6, 1±9. Belsky, J. (1988). The ``effects'' of infant day care reconsidered. Early Childhood Research Quarterly, 3, 235±272. Belsky, J. (in press). Developmental risks (still) associated with early child care. Journal of Child Psychology and Psychiatry. Berrueta-Clement, J.R., Schweinhart, L.J., Barnett, W.S., Epstein, A.S., & Weikart, D.P. (1984). Changed lives: The effects of the Perry Pre-school Programme on youths through age 19. Ypsilanti, MI: High/Scope Press. Bradley, R.H., Caldwell, B.M., Fitzgerald, J.A., Morgan, A.G., & Rock, S.L. (1996). Experience in day care and social competence among maltreated children. Child Abuse and Neglect, 10, 181±189. Bredekamp, S. (1986). The reliability and validity of the Early Childhood Classroom Observation Scale for accrediting early childhood programs. Early Childhood Research Quarterly, 1, 103±118. Bruner, J. (1980). Under ®ve in Britain: The Oxford Pre-school Research Project. Oxford: Grant McIntyre. Bryk, A.S., & Raudenbush, S.W. (1992). Hierarchical linear models: Applications and data analysis methods. Newbury Park, CA: Sage. Burchinal, M.R., Peisner-Feinberg, E., Bryant, D.M., & Clifford, R. (2000). Children's social and cognitive development and child care quality: Testing for differential associations related to poverty, gender or ethnicity. Applied Developmental Science, 4, 149±165. Campbell, J., Lamb, M.E., & Hwang, C.P. (2000). Early child care experiences and children's social competence between 1.5 and 15 years of age. Applied Developmental Science, 4, 166±175. Clark, M.M. (1988). Children under ®ve: Educational research and evidence. London: Gordon & Breach.

5. The quality of pre-school experience

97

Clarke-Stewart, K.A. (1989). Infant day care: Maligned or malignant? American Psychologist, 44, 266±273. Desai, S., Chase-Lansdale, P.L., & Michael, R.T. (1989). Mother or market? Effects of maternal employment on the intellectual ability of four-year-old children. Demography, 26, 545±561. Field, T. (1991). Quality infant day care and grade school behaviour and performance. Child Development, 62, 863±870. Fox, N., & Fein, G.G. (Eds.). (1990). Infant day care: The current debate. Norwood, NJ: Ablex. Fuerst, J.S., & Fuerst, D. (1993). Chicago experience with early childhood programs: The special case of the child±parent center programs. In L.J. Schweinhart et al. (Eds.), Signi®cant bene®ts: The High/Scope Perry Pre-school Study through age 27. Ypsilanti, MI: High/Scope Press. Garber, H.L. (1988). The Milwaukee Project: Preventing mental retardation in children at risk. Washington DC: American Associates on Mental Retardation. In L.J. Schweinhart et al. (Eds.), Signi®cant bene®ts: The High/Scope Perry Preschool Study through age 27. Ypsilanti, MI: High/Scope Press. Goelman, H., & Pence, A.R. (1987). Effects of child care, family and individual characteristics on children's language development. In D. Phillips (Ed.), Quality in child care: What does the research tell us? (pp. 89±104). Washington, DC: National Association for the Education of Young Children. Golden, M., Rosenbluth, L., Grossi, M.T., Policare, H.J., Freeman, H., & Brownlee, E.M. (1978). The New York City Infant Day Care Study. New York: Medical and Health Research Association of New York City. Goldstein, H. (1987). Multilevel models in education and social research. London: Charles Grif®n. Goldstein, H. (1995). Multilevel statistical models (2nd ed.). London: Edward Arnold. Goossens, C., & Melhuish, E.C. (1996). On the ecological validity of measuring the sensitivity of professional caregivers: The laboratory versus the nursery. European Journal of Psychology of Education, 11, 169±176. Hadeed, J.A. (1994). The effects of preschool experience on some aspects of child development in Bahrain. PhD thesis, University of London Institute of Education. Harms, T., & Clifford, R.M. (Ed.). (1980). Early Childhood Environmental Rating Scale. New York: Teacher's College Press. Harms, T., & Clifford, R.M. (1989). Family Day Care Rating Scale. New York: Teacher's College Press. Harms, T., Clifford, R.M., & Cryer, D. (1998). Early Childhood Environmental Rating ScaleÐRevised. New York: Teacher's College Press. Harms, T., Cryer, D., & Clifford, R.M. (1990). Infant/Toddler Environmental Rating Scale. New York: Teacher's College Press. Howes, C., & Olencik, M. (1986). Family and child care in¯uences on toddler's compliance. Child Development, 57, 202±216. Howes, C., & Stewart, P. (1987). Child's play with adults, toys and peers: An examination of family and child care in¯uences. Developmental Psychology, 23, 423±430. Hutt, S.J., Tyler, S., Hutt, C., & Foy, H. (1989). Play, exploration and learning. London: Routledge. Johansson, I. (1993). Quality in early childhood servicesÐWhat is that? Paper

98

Melhuish

presented at Third European Conference on the quality of early childhood education, Kriopigi, Greece: European Early Childhood Education Research. Jowett, S., & Sylva, K. (1986). Does kind of research matter? Educational Research, 28, 21±31. Kagitcibasi, C., Sunar, D., & Bekman, S. (1988). Comprehensive pre-school education projects. Ottowa: International Development Research Center (IDRC). Katz, L.G. (Ed.). (1995). Talks with teachers of young children: A collection. Norwood, NJ: Ablex. Lamb, M., Sternberg, K., Hwang, C., & Broberg, A. (Eds.). (1992). Child care in context. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Lazar, I., Darlington, R., Murray, H., Royce, J., & Snipper, A. (1982). Lasting effects of early education: A report from the consortium for longitudinal studies. Monographs of the Society for Research in Child Development, 47, 2±3. Lee, V.E., Brooks-Gunn, J., & Schnur, E. (1988). Does Head-Start work? Developmental Psychology, 24, 210±222. McCartney, K. (1984). Effect of quality of day care environment upon children's language development. Developmental Psychology, 20, 244±260. McCartney, K., & Jordan, E. (1990). Parallels between research on child care and research on school effects. Educational Researcher, 19, 24±27. McGuire, J., & Richman, N. (1987). Management of behaviour problems in day nurseries. Early Child Development and Care, 45, special issue. McKey, H.R., Condelli, L., Barrett, B.J., McConkey, C., & Plantz, M. (1985). The impact of Head Start on children, families and communities: Final report of the Head Start evaluation, synthesis and utilisation project. The Head Start Bureau, Administration for Children, Youth and Families, Of®ce of Human Development Services. Washington, DC: CSR Incorporated. Meisels, S.J., & Shonkoff, J.P. (Eds.). (1990). Handbook of early childhood intervention. Cambridge: Cambridge University Press. Melhuish, E.C. (1991). Research issues in day care? In P. Moss & E.C. Melhuish (Eds.), Current issues in day care for young children: Research and policy implications. London: HMSO. Melhuish E.C. (1993). Predskolni pece a vychova. Pedagogika, 4, 449±458. Melhuish, E.C., & Moss, P. (Eds). (1990) Day care for young children: International perspectives. London: Routledge. Melhuish, E.C., Lloyd, E., Martin, S., & Mooney, A. (1990a) Type of childcare at 18 months: II. Relations with cognitive and language development. Journal of Child Psychology and Psychiatry, 31, 861±870. Melhuish, E.C., Mooney, A., Martin, S., & Lloyd, E. (1990b). Type of childcare at 18 months: I. Differences in interactional experiences. Journal of Child Psychology and Psychiatry, 31, 849±859. Melhuish, E.C., Sylva, K., Sammons, P., Siraj-Blatchford, I., & Taggart, B. (1999). The Effective Provision of Pre-school Education Project, Technical Paper 4: Parent, family and child characteristics in relation to type of pre-school and socioeconomic differences. London: Institute of Education/DfEE. Melhuish, E.C., Sylva, K., Sammons, P., Siraj-Blatchford, I., & Taggart, B. (2001). The Effective Provision of Pre-school Education Project, Technical Paper 7: Social/ behavioural and cognitive development at 3±4 years in relation to family background. London: Institute of Education/DfEE. Munton, A.G., Mooney, A., & Rowland, L. (1995). Deconstructing quality: A

5. The quality of pre-school experience

99

conceptual framework for the new paradigm in day care provision for the under eights. Early Child Development and Care, 114, 11±23. NICHD Early Child Care Research Network. (1996). Characteristics of infant child care: Factors contributing to positive caregiving. Early Research Childhood Quarterly, 11, 269±306. NICHD Early Child Care Research Network. (1997). The effects of infant child care on infant-mother attachment security: Results of the NICHD Study of Early Child Care. Child Development, 68, 860±879. NICHD Early Child Care Research Network. (1998). Early child care and selfcontrol, compliance, and problem behaviors at twenty-four and thirty-six months. Child Development, 69, 1145±1170. NICHD Early Child Care Research Network. (1999a). Child outcomes when child care centre classes meet recommended standards for quality. American Journal of Public Health, 89, 1072±1077. NICHD Early Child Care Research Network. (1999b). Does child care quality matter? Effect sizes from the NICHD Study of Early Child Care. Paper presented at the biennial meetings of the Society for Research in Child Development, Albuquerque, NM. NICHD Early Child Care Research Network. (1999c). Child care and mother±child interaction in the ®rst three years of life. Developmental Psychology, 35, 1399±1413. NICHD Early Child Care Research Network. (2000). The relation of child care to cognitive and language development. Child Development, 71, 958±978. NICHD Early Child Care Research Network. (in press). Early child care and children's development prior to school entry: Results from the NICHD study of early child care. American Educational Research Journal. O' Connell, J.C., & Farran, D.C. (1982). Effects of day care experience on the use of Intentional communicative behaviours in a sample of socio-economically depressed infants. Developmental Psychology, 18, 22±29. Peisner-Feinberg, E.S., Burchinal, M.R., Clifford, R.M., Culkin, M.L., Howes, C., Kagan, S.L., & Yazejian, N. (2001). The relation of preschool child-care quality to children's cognitive and social developmental trajectories through second grade. Child Development, 72, 1534±1553. Peterson, C., & Peterson, R. (1986). Parent±child interaction and day care: Does quality of day care matter? Journal of Applied Developmental Psychology, 7, 1±15. Petrogiannis, K., & Melhuish, E.C. (1996). Aspects of quality in Greek day care centres. European Journal of Psychology of Education, 11, 177±192. Pierrehumbert, B., Ramstein, T., Krucher, R., El-Najar, S., Lamb, M.E., & Halfon, O. (1996). L'evaluation du lieu de vie du jeune enfant: Developpement d'un instrument. Bulletin de Psychologie, XLIX, 426, 565±584. Prescott, E., Kritchevsky, S., & Jones, K. (1972). The Day Care Environment Inventory. Washington, DC: Department of Health, Education and Welfare. Pugh, G. (1992). Contemporary issues in the early years: Working collaborative with children. London: National Children's Bureau and Paul Chapmann. Ramey, C.T., Bryant, D.M., Campbell, F.A., Sparling, J.J., & Wasik, B.H. (1988). In L.J. Schweinhart, H. Barnes, & D. Weikart (Eds.), Signi®cant bene®ts: The High/Scope Perry Pre-school Study through age 27. Ypsilanti, MI: High/Scope Press. Ramey, C.T., & Mills, P. (1977). Social and intellectual consequences of day care for

100

Melhuish

high-risk infants. In R. Webb (Ed.), Social development in childhood: Daycare programs and research. Baltimore, MD: Johns Hopkins University Press. Reifel, S. (Ed.). (1993). Advances in early education and day care: Perspectives on developmentally appropriate practice. Greenwich, CT: JAI Press. Rosenthal, M.K. (1999). Out of home child care research: A cultural perspective. International Journal of Behavioural Development, 23, 477±518. Rutter, M. (1985). Family and school in¯uences on cognitive development. Journal of Child Psychology and Psychiatry, 26, 683±704. Sammons, P., Sylva, K., Melhuish, E.C., Siraj-Blatchford, I., & Taggart, B. (1999). The Effective Provision of Pre-school Education Project: Technical Paper 2. Characteristics of the EPPE project sample at entry to the study. London: Institute of Education/DfEE. Schlieker, E., White, D.R., & Jacobs, E. (1991). The role of day care quality in the prediction of children's vocabulary. Canadian Journal of Behavioural Science, 23, 12±24. Schweinhart, L.J. (1987). When the buck stops here: What it takes to run good early childhood programmes. Ypsilanti, MI: High/Scope Press. Schweinhart, L.J., Barnes, H., & Weikhart, D. (Eds.). (1993). Signi®cant bene®ts: The High/Scope Perry Pre-school Study through age 27. Ypsilanti, MI: High/ Scope Press. Sylva, K. (1990). Evaluating early education programmes. Early Child Development and Care, 58, 97±107. Sylva, K. (1994). The impact of early learning on children's later development. In C. Ball (Ed.), Start right: The importance of early learning. London: The Royal Society for the Encouragement of Arts, Manufacturers and Commerce. Sylva, K. (1997). Target child observation. In K. Sylva & J. Stevenson (Eds.), Social behaviour and competence in childhood. Windsor, UK: NFER-Nelson. Sylva, K., Melhuish, E.C., Sammons, P., Siraj-Blatchford, I., & Taggart, B. (2001). The Effective Provision of Pre-school Education Project. Symposium to the British Educational Research Association Conference, Leeds, UK. Sylva, K., Sammons, P., Melhuish, E.C., Siraj-Blatchford, I., & Taggart, B. (1999a). The Effective Provision of Pre-school Education Project: Technical Paper 6. Characteristics of the centres in the EPPE sample: Observational pro®les. London: Institute of Education/DfEE. Sylva, K., Sammons, P., Melhuish, E.C., Siraj-Blatchford, I., & Taggart, B. (1999b). The Effective Provision of Pre-school Education Project, Technical Paper 1: An introduction to the EPPE project. London: Institute of Education/DfEE. Vandell, D.L., Henderson, V.K., & Wilson, K.S. (1988). A longitudinal study of children with day-care experiences of varying quality. Child Development, 59, 1286±1292. Vandell, D.L., & Wolfe, B. (2000). Child care quality: Does it matter and does it need to be improved? Institute for the Study of Poverty, SR#78. http://www.ssc.wisc. edu/irp/ Whitebook, M., Howes, C., & Phillips, D. (1989). Who cares? Child care teacher and the quality of care in America: Final report of the National Child Care Staf®ng Study. Oakland, CA: Child Care Employee Project. WHO (World Health Organisation). (1990). WHO Child Care Facility Schedule with User's Manual. Geneva: WHO Division of Mental Health.

5. The quality of pre-school experience

101

Woodhead, M. (1996). In search of the rainbow: Pathways to quality in large-scale programmes for young disadvantaged children. The Hague: Bernard Van Leer Foundation. Zigler, E., & Styfco, S. (Eds.). (1993). Head Start and beyond: A national plan for extended childhood intervention. New York: Yale University Press.

6

Development of children's knowledge about the mental world John H. Flavell

Stanford University, USA

A developmental psychologist shows a 5-year-old a cookies box with a picture of cookies on it and asks her what is in it. ``Cookies'', is the ready answer. The child then looks inside the box and to her surprise sees that it actually contains crayons, not cookies. What would another child who had not yet opened the box think was in it? the experimenter now asks. ``Cookies!'' says the child, amused at the trick. The experimenter then tries the same procedure with a 3-year-old. The answer to the ®rst question is the expected ``cookies'', but the response to the second is unexpected: ``crayons''. Even more surprising, the child also maintains that he himself had initially thought that the box would contain crayons. Unlike the 5year-old, the 3-year-old shows no evidence of understanding that either he or other people could hold a belief that is false. Results such as this are found in a currently ¯ourishing research area concerning the development of our knowledge and beliefs about the mental worldÐour folk psychology or naive theory of mind. More than did earlier metacognitive and social-cognitive investigations in the same domain, this approach probes children's conceptions of the most fundamental components of the mind, such as beliefs and desires, and children's knowledge of how these components affect and are affected by perceptual inputs and behavioural outputs. In less than 20 years, this fast-growing area has spawned hundreds of research articles and scores of book- and monograph-length treatments. Indeed, the spate of papers and posters on this topic at recent meetings of the Society for Research in Child Development reminded several older participants of the way Piagetian research dominated the programme in years past. To illustrate, a recent meta-analysis of false belief studies aloneÐ just one topic in this areaÐincluded 178 separate studies (Wellman, Cross, & Watson, 2001). Developmental ®ndings in this area have also become of interest recently to philosophers of mind, who believe that these ®ndings may help clarify philosophical disputes about the nature of folk psychology. Why this intense research interest in the development of knowledge about the mental world? One answer is that human social and cognitive life bereft of such knowledge seems virtually unimaginable, and that the development of something that important is surely worth learning about. In her

6. Development of children's knowledge

103

lectures on this topic, Alison Gopnik likes to make the point in the following way: Imagine what it would be like for you to give a lecture to an audience if you had no conception of mental states. The audience might appear to you as bags of meat with two small holes at the top. You would see these bags and the shiny things in their holes shift around unpredictably in a way that perplexes and terri®es you, although of course you do not realise that you are perplexed and terri®ed. Gopnik's scenario may not be as imaginary as it seems. Autistic individuals are known to be de®cient in knowledge about the mind, and sometimes act as if they view other people as unpredictable and scary in much this way. The plan of this chapter is as follows. The chapter begins with a history of this research area. Next comes a summary of its main theories and research ®ndings. The chapter then concludes with some speculations about the future of the area. HISTORY As with so many other areas of cognitive development, the history of this one mainly begins with Piaget (Flavell & Miller, 1998; Shantz, 1983). A central Piagetian claim was that children begin development by being cognitively egocentric (Flavell, 1992). By this, Piaget meant that they initially do not know that there are such things as conceptual, perceptual, and emotional perspectives or points of view. As a result, they naturally cannot be aware that they themselves have such perspectives vis-aÁ-vis external objects and events, or that others do, or that their own perspective may not be the same as those of others, or that they may be unwittingly reporting their own perspectives when asked to report another person's. Piaget also considered as egocentric children who have some awareness that perspectives exist but who are not skilled at discriminating their own from another person's. Piaget and his colleagues used egocentrism and other concepts to interpret their developmental studies of a wide variety of socialcognitive topics: perceptual perspective-taking; egocentric communication; the misattribution of mental characteristics to physical objects (animism) and physical characteristics to mental events (realism); and understanding of thoughts, dreams, intentions, and morality. Research on these and related topics still continues, although usually not from a Piagetian theoretical perspective (e.g., Fein®eld, Lee, Flavell, Green, & Flavell, 1999; Flavell, 1992; Flavell, Green, & Flavell, 1995b; Woolley & Wellman, 1992). Shantz (1983) also describes more recent stage theories of various aspects of social-cognitive development in the Piagetian tradition by Damon, Selman, and Turiel. She also summarises numerous studies of perspective-taking and related Piagetian topics by Borke, Chandler, Feffer, Flavell, Selman, and many other researchers. There is widespread agreement today that young children are not as totally egocentric as Piaget believed them to be,

104

Flavell

but also that perspective-taking abilities and related psychological knowledge do show marked increases with age, much as he said they did. Those of us trying to peer into the ontogenesis of knowledge about the mind are clearly standing on Piaget's shoulders. A second wave of theory and research in this general area was the extensive work on metacognitive development that began in the early 1970s. Useful reviews of this large literature include Brown, Bransford, Ferrara, and Campione (1983), Flavell, Miller, and Miller (2002), Kuhn (1999), Moshman (1998), and Schneider and Bjorklund (1998). Metacognition (cognition about cognitionÐhence the ``meta'') has been de®ned as any knowledge or cognitive activity that takes as its cognitive object, or that regulates, any aspect of any cognitive activity (Flavell et al., 2002, p. 164). It is a broad concept that encompasses people's knowledge about the nature of people as cognisers, about the nature of different cognitive tasks, and about possible strategies for coping with different tasks. It also includes executive skills for monitoring and regulating one's own cognitive activities. The majority of developmental studies classi®ed as metacognitive have investigated children's metamemory, that is, their knowledge about variables affecting memory performance and, especially, their knowledge and use of memory strategies. The term has also been applied to numerous studies of children's cognition concerning comprehension, communication, language, perception, and attention, and problem solving. Research in the metacognitive development tradition is still being done, although it is not the hot topic it used to be. The third wave in our history is still very much in motion, and is the primary concern of this article. It is commonly referred to as theory-of-mind research. Prior to about 1983, most investigators of children's knowledge about the mental world would probably classify their work as either metacognitive or in the general Piagetian tradition. Today most would say they are doing one or another kind of theory-of-mind research. In fact, they would probably use that label as a shorthand, easily recognisable characterisation of the general line of work they are in, even if they were not convinced, as some are not, that children actually acquire bona ®de theories of mind rather than just knowledge and skills concerning it. Since the mid1980s, theory-of-mind research has been one of the liveliest, most productive research areas in all of developmental psychology. I predict that it will continue to be so for some time to come. How did this third wave come to be? In the 1978 issue of Behavioral and Brain Sciences, Premack and Woodruff reported some research in which they attempted to test whether chimpanzees have a theory of mind, which they de®ned as follows (Premark & Woodruff, 1978, p. 515): An individual has a theory of mind if he imputes mental states to himself and others. A system of inferences of this kind is properly viewed as a theory because such states are not directly observable, and the system

6. Development of children's knowledge

105

can be used to make predictions about the behavior of others. As to the mental states the chimpanzees may infer, consider those inferred by our own species, for example, purpose or intention, as well as knowledge, belief, thinking, doubt, guessing, pretending, liking, and so forth. In their commentaries on this article, three philosophers independently suggested that one might be able to ®nd out whether an animal possessed the concept of belief in something like the following fashion (Bennett, 1978; Dennett, 1978; Harman, 1978). The subject animal sees another individual put object X in container A and then leave the scene. The subject then sees someone else move X from container A into container B while the individual is still absent. The subject animal should then be credited with some understanding of belief if it acts as if it expects that the returning individual will search for X in A rather than B: ``If the subject chimpanzee expects the second chimpanzee to reach into the pot which originally contained the banana, that would seem to show that it has a conception of mere belief'' (Dennett, 1978, p. 557). Philosophers and developmental psychologists consider false-belief tasks to be better tests of the concept of belief than are true-belief tasks because children could be correct on true-belief tasks by egocentrically assuming that others know what they themselves know and just reporting the true state of affairs. These ideas were taken up in the early 1980s by two Austrian psychologists, Josef Perner and Heinz Wimmer. In a pioneering and highly in¯uential series of studies, they used the ``unexpected transfer'' method proposed by the philosophers to test young children's understanding of false belief (Wimmer & Perner, 1983). Similarly, Bretherton and her colleagues examined infants' gestural and verbal communication for evidence that infants have ``what Premack and Woodruff (1978) have called a `theory of mind' '' (Bretherton, McNew, & Beeghly-Smith, 1981, p. 339). Around the same time, Wellman and his co-workers had independently begun to conceptualise children's developing metacognitive knowledge and understanding of mental terms as the development of a theory of mind (e.g., Johnson & Wellman, 1980; Shatz, Wellman, & Silber, 1983; Wellman, 1983, 1985). In addition, a number of other researchers who had not yet begun to conceptualise children's social-cognitive development in quite this way had been doing research that subsequently became part of the theory-of-mind movement. An example would be the work on children's knowledge about perception and about the appearance±reality distinction by Flavell and colleagues (e.g., Flavell, Flavell, & Green, 1983; Flavell, Flavell, Green, & Wilcox, 1980; see Astington, Harris, & Olsen, 1988, for other such projects). The movement was given added identity and coherence by two conferences that were held in the spring of 1986: the International Conference on Developing Theories of Mind, organised by Janet Astington, Lynd Forguson, Alison Gopnik, and David Olson at the University of Toronto, and the Workshop on Children's Early Concept of Mind, organised by

106

Flavell

Paul Harris at the University of Oxford. The presentations given at these two conferences were later published in a book entitled Developing theories of mind (Astington, Harris, & Olson, 1988), and the movement was of®cially launched. A look through this milestone publication provides an immediate sense of the broad and diverse array of acquisitions judged to be instances of theory-of-mind development, and more have been added since. What it does not convey is the high excitement felt by the conference participants as they sensed the birth of a new approach to the development of children's knowledge about the mental world. The rest, as they say, is history. Publications concerned with theory-ofmind development must now number well up in the hundreds and the ¯ow continues unabated. Reviews of work on this topic include Astington (1993), Baron-Cohen, Tager-Flusberg, and Cohen (2000), Bartsch and Wellman (1995), Carpenter, Nagell, and Tomasello (1998), Flavell (1999), Flavell and Miller (1998), Hala (1997), Mitchell (1997), Mitchell and Riggs (2000), Moore (1996), Taylor (1996), Wellman (1993), and Wellman and Gelman (1998). It may be helpful in understanding the metacognition and theory-of-mind approaches to compare and contrast them. In a general sense, researchers in both traditions share the same overall objective, that is, to investigate the development of children's knowledge and cognition about mental phenomena. In fact, most psychologists would probably consider the terms ``metacognition'' and ``theory of mind' as being more or less synonymousÐ as just alternative ways of designating the same general set of cognitive phenomena. This sense of synonymity is further heightened by the extensive use of the expression ``metarepresentational'' in the theory-of-mind development literature: ``Metarepresentational'' and ``metacognitive'' sound like they mean pretty much the same thing. Despite these commonalities, the research literatures in these two areas have been surprisingly distinct and unconnected. Most theory-of-mind articles do not cite research in the metacognitive development traditionÐ for example, research on metamemory developmentÐand conversely, most metacognitive development articles do not refer to work in the theory-inmind tradition (for a clear exception, see a recent chapter by Kuhn, 1999). One gets a further sense of discontinuity when one looks for the adult counterparts of these two research traditions. There is currently a fair amount of psychological research being done with adults under the rubric ``metacognition''Ðon feelings of knowing, for example (Jost, Kruglanski, & Nelson, 1998; Metcalfe & Shimamura, 1994)Ðbut very little that is construed as research on adult theory of mind1. Why this lack of connection? I do not think it is due to provincialism or lack of vision on the part of researchers concerned. More likely, it is 1 There is relevant work here in the area of adult social cognition, to be sure, but it is scarcely ever presented as research on adult theory of mindÐthat is, as research on the adult outcome of a process of theory-of-mind development.

6. Development of children's knowledge

107

because the two traditions have tended to focus on different developments within the broad cognition-about-cognition umbrella. Most theory-of-mind studies have investigated children's initial knowledge about our most basic mental statesÐdesires, percepts, beliefs, knowledge, thoughts, intentions, feelings, and so on. Researchers in this tradition attempt to determine what children of different ages know about the existence and behaviour of these various states, and also what they know about how mental states are causally linked to perceptual inputs, to behaviour, and to other mental states. For example, do young children understand what it means to know something, or do they realise that unsatis®ed desires typically cause negative feelings and renewed behavioural efforts to satisfy these desires? In contrast, students of metacognitive development have usually focused more on task-related mental activitiesÐoften, on what one should do with one's mind in trying to solve some problem or task. These metacognitive activities include strategies for making cognitive progress on various tasks and problemsÐon memory or comprehension tasks, for exampleÐand also attempts to monitor that progress. Much of the metacognition studied is therefore problem-centred and goal-oriented; one could think of it as a kind of ``applied theory of mind''. Because most theory-of-mind researchers have been looking for the origins and earliest expressions of knowledge about the most elementary types of mental states (desires, beliefs, etc.), they have tended to study infants and young children predominantly. Conversely, because the knowledge and skills metacognition researchers investigate usually presuppose the prior acquisition of some understanding of these states, they have mainly tested older children and adolescents. One cannot test for a child's understanding of memory strategies (metacognition) if the child is too young even to know what remembering something means (theory of mind). Because it has this applied-theory-of-mind focus, most metacognitive development research is concerned with what the subject knows about how to use his or her own mind rather than somebody else's. How often other people or people in general use their minds in task situations may be useful as models for how the subject should use his or hers, but it is the subject's own use or nonuse that is usually of primary interest. In contrast, it is the subject's understanding of some other person's mind, or of minds in general, that is usually of concern in theory-of-mind studies. For example, in a false-belief task the child subject is typically asked what a naive other child would think is in the cookies boxÐthe deceptive box which the subject has just learned really contains crayons rather than cookies. MAIN THEORIES AND FINDINGS Figure 6.1 illustrates the main directions that theory-of-mind development research has taken since it began in the early 1980s. Much of the earliest

108

Flavell Theories: domain-specific theory development, innate or early-maturing modules, simulation, information-processing, etc.

Later developments

Earlier developments

Consequents Tasks: FB, AR, PT Ages: 3–5 years

Intracultural Antecedents

Differences

Intercultural Interspecies

Other mental states: desires, intentions, emotions, percepts, fictional representations, attention, thinking, consciousness, etc.

Figure 6.1 Overview of research directions in the area of theory-of-mind development.

work was focused on documenting a striking improvement between 3 and 5 years of age in children's performance on various false-belief (FB), appearance-reality (AR), and Level 2 visual perspective-taking (PT) tasks. Thus, for example, older but not younger preschoolers were usually found to show an understanding that the naive other child cited in the ®rst paragraph of this article would falsely believe that the cookie box contains cookies (false belief ), that a fake rock looks like a rock but is really a sponge (appearance-reality), and that a picture book oriented correctly for them on the table will look upside-down to a person seated opposite (Level 2 visual perspective taking). From those beginnings work has progressed more or less concurrently in a variety of directions, as shown by the arrows in Figure 6.1. Researchers have charted the development of children's understanding of many additional mental states. They have elaborated several classes of theories to explain theory-of-mind development. They have extended their inquiries both downward into infancy and upward into middle childhood and adolescence. A few attempts have been made to identify child-rearing or

6. Development of children's knowledge

109

other variables that predict facility with theory-of-mind tasks (antecedents) and to identify behaviours that such facility might mediate (consequents). Finally, investigators have examined differences in theory-of-mind understandingsÐdifferences within the same culture, differences between cultures, and differences between humans and other primates. These points are brie¯y elaborated in the sections that follow. Fuller discussion of them can be found in the substantive reviews cited earlier.

Theories Several types of theories have been offered as explanations for the development of children's mentalistic understanding. One is the so-called theory theory (Gopnik & Meltzoff, 1997; Gopnik & Wellman, 1994; Perner, 1991; Wellman & Gelman, 1998). Theory theorists argue that our knowledge about the mind comprises not a formal scienti®c theory but an informal, everyday ``framework'' or ``foundational'' theory. A number of steps in children's progression toward the adult theory of mind have been described. For instance, Bartsch and Wellman (1995) have argued that children begin with a desire psychology, then progress to a desire-belief psychology, and ®nally attain our adult belief-desire psychology, in which one recognises that what people believe, as well as what they desire, crucially affects how they behave. Theory theorists argue that experience plays a major formative role in children's theory-of-mind development. In contrast, modularity theorists such as Leslie (1994) argue that young children are not acquiring a theory about mental representations at all. Rather, Leslie postulates the acquisition through neurological maturation of a succession of three domain-speci®c and modular mechanisms for dealing with agents versus nonagent objects. Although experience may be necessary to trigger the operation of these mechanisms, it does not determine their nature. Baron-Cohen (1995) also has a developmental theory involving innate or early-maturing modular mechanisms dedicated to mental state computations. Harris (1992) and others have proposed yet a third approach. According to their simulation theory, children become able to compute the mental states of other people through a kind of role-taking or simulation process. What develops is the ability to make increasingly accurate simulations of this kind. Like theory theorists, simulation theorists also assume that experience plays a crucial formative role, in that it is through practice in role taking that children improve their simulation abilities. Still other developmentalists believe that young children's failures on false-belief and other theory-of-mind tasks may be caused by more domaingeneral limitations in executive functioning (e.g., Carlson, Moses, & Hix, 1998; Hughes, 1998). For example, an inability to inhibit a dominant, ready-to-go response could cause the child subject to blurt out the cognitively salient real contents of the cookies box when asked for what the

110

Flavell

naive other child thought it contained. Other investigators have argued that the tasks may be misunderstood by young children or may not be engaging enough to elicit their best thinking (Flavell & Miller, 1998). How shall we evaluate all these different theories? It seems likely to me that an adequate theory here will ®nally have to include elements from all of them (Flavell, 1999, p. 27): That is, the following seem likely: (a) that development in this area builds from innate or early maturing people-reading capacities; (b) that we have some introspective ability that we can and do exploit when trying to infer the mental states of other creatures who are like ourselves but in a different psychological situation (e.g. ignorant of the facts, differently motivated); (c) that much of our knowledge of the mind can be characterized as an informal theory; (d) that improved information-processing and other abilities (e.g. linguistic skills) enable and facilitate theory of mind development (and certainly help children show what they know on theory-of-mind tasks); and (e) that a variety of experiences serve to engender and change children's conceptions of the mental world and their ability to use these conceptions in predicting and explaining their own and other people's behavior.

Developments during infancy Infants are born with or develop early a number of abilities and dispositions that will help them learn about people. They ®nd human faces, voices, and movements highly interesting. They have impressive abilities to perceptually analyse and discriminate human stimuli. They seem impelled to attend to and interact with other people and they certainly impel other people to attend to and interact with them. Babies respond differently to people from the way they do to objects and seem to expect people to behave differently to objects (Poulin-Dubois, 1999). More speci®cally, we could say that they come to construe people as ``compliant agents'', that is, as objects that are self-propelled and capable of independent movement (agents) but also can be in¯uenced at a distance by communicative signals (compliant). One can hardly imagine a more felicitous initial design for a creature destined for theory-of-mind development. Late in the ®rst year infants start to learn that people's behaviour possesses ``aboutness'' or ``intentionality'' (intentionality in a broad senseÐ not just the narrow sense of ``on purpose''). An individual's behaviour is ``about'' an object in this sense if the individual perceptually attends to it, labels it, thinks about it, wants it, fears it, intends or tries to get it, or relates to it in any other psychological way. Infants do a variety of things that re¯ect a beginning awareness of intentionality. They try to engender new ``aboutnesses'' in others through various communicative gestures, such as pointing to or vocalising about an object and checking to see if the other person

6. Development of children's knowledge

111

attends to it. They also develop skill at reading the aboutnesses the other person already has going, as when they follow the person's gaze. Carpenter, Nagell, and Tomasello (1998) have recently documented a three-step developmental sequence in which infants progress from sharing to following to directing others' attention and behaviour. Studies by Meltzoff (1995) have also demonstrated that 18-month-olds can infer what action another person is trying to perform (e.g., attempting to pull one object away from another object to which it is attached), even though the person is unsuccessful in the attempt (does not succeed in pulling it away) and therefore never actually demonstrates the intended action. This suggests that infants of this age have some sense that people's actions are intentional and goal-directed. Precursors of such understanding can even be observed in early infancy (Woodward, 1998). By age 18 months infants understand that they should give an experimenter a food that she reacts to with pleasure rather than one towards which she acts disgusted, even when they themselves prefer the latter food (Repacholi & Gopnik, 1997); this suggests at least some limited ability to reason nonegocentrically about people's desires. Infants also recognise that it is the adult's attentional focus rather than their own that gives clues as to the adult's referential intent when the adult labels an object (Woodward & Markman, 1998). Similarly, they develop the ability to learn what an object is like by reading the adult's attentional focus when the adult is expressing a positive or negative emotional reaction to it (a process called social referencing). For instance, they may selectively avoid an object towards which their parent shows negative affect. Thus, by 12 months or so they can recognise that the adult's emotional display refers to or is ``about'' a particular object much as they can recognise that the adult's spoken label refers to or is ``about'' a particular object (Moses, Baldwin, Rosicky, & Tidball, 2001). By the end of infancy children may also do other things suggestive of a beginning understanding of human psychology, such as trying to comfort people in distress and correctly using mental state terms such as ``want'' and ``see''.

Later developments A very large literature has accrued since the early 1980s on theory-of-mind acquisitions that occur subsequently to the infancy period. What follows is a brief synopsis of some of the major ®ndings, organised by type of mental state. Visual perception During the early pre-school period children already understand that a person will see an object if and only if the person's eyes are aimed in the general direction of the object, and if there are no vision-blocking obstacles interposed between the person and the object (Flavell, 1992). With this

112

Flavell

understanding, they are able to do simple, nonegocentric visual perspectivetaking; for example, they can infer that you may see something that they do not and vice versa (referred to as Level 1 knowledge about visual perception). Later in the pre-school period they go on to recognise that the same thing may present different visual appearances to two people if they view it from different positions (called Level 2 knowledge). Attention As already mentioned, even infants pay attention to other people's attending and seem to have some understanding of its implications. In subsequent years they come to appreciate that attention is selective and limited, and that different people may mentally represent the same attended-to input differently (Fabricius & Schwanen¯ugel, 1994; Flavell, Green, & Flavell, 1995a; Pillow, 1995). Desires By the age of 3 children are not only using some desire terms correctly, they also seem to grasp simple causal relations among desires, outcomes, emotions, and actionsÐsuggestive evidence that they are developing something like an implicit theory. For example, they understand that people will feel good if they get what they want and feel bad if they do not, and that people will quit searching if they ®nd a sought-after desired object but will keep searching if they do not (Bartsch & Wellman, 1995). Emotions Although we do not know yet whether infants actually impute inner feelings to people who display emotions, there is little doubt but that young preschoolers ``evidence an understanding of emotions as experiential states of persons, as distinguished from the actions (e.g., hitting) and expressions (e.g., smiling) that emotions cause, and they distinguish between the subjective emotional experiences of different individuals'' (Wellman, Harris, Banerjee, & Sinclair, 1995, p. 118). In later years children gradually learn more advanced truths about emotions, for example, that people do not always really feel what they appear to feel, that their affective reactions to an event may be in¯uenced by earlier emotional experiences with similar events or by their current mood, and that people can experience two con¯icting emotions more or less simultaneously (Flavell & Miller, 1998). Beliefs and related mental representations There have been a great many studies of children's developing understanding of ``serious'' mental representations, that is, nonpretence mental

6. Development of children's knowledge

113

states such as beliefs that are meant to represent reality accurately (Flavell & Miller, 1998). The majority of these have dealt with children's comprehension of representations that differ from person to person or differ from reality: the appearance±reality distinction, Level 2 knowledge of visual perception, interpretation and constructive processing, deception, and, most studied of all, false belief. The distinction between perceptual appearance and reality is conceptually similar both to the distinction between false belief and reality and to the Level 2 distinction between two different perceptual perspectives. Consistent with this fact, there is some correlational evidence suggesting that these three distinctions tend to develop synchronously, within subjects, during the pre-school years. As noted in the Theories section, however, exactly what false-belief and appearance-reality tasks measure remains uncertain; some researchers (including me) believe they mainly assess the child's nascent understanding of mental representation, but others disagree. The evidence indicates that children's knowledge about mental representations continues to increase after the pre-school period. In particular, it is not until middle childhood and later that children appear to gain any substantial understanding of the mind as an active, interpretive, constructive processor (e.g., Carpendale & Chandler, 1996). For instance, understanding that people's interpretation of an ambiguous event may be in¯uenced by their pre-existing biases or expectations seems to be a middle-childhood insight (Pillow & Henrichon, 1996). Knowledge Young preschoolers appear to be unclear about just what it means for someone to know something and about how knowledge is acquired (Flavell & Miller, 1998). Even 4- and 5-year-olds may claim that they have always known information that they have just learned during the experimental session (Taylor, Esbensen, & Bennett, 1994). An important early±middlechildhood discovery is that perceptual information has to be adequate as well as merely present to engender knowledge. For example, children come to realise that a person often cannot be certain of an object's identity when only a little bit of it is visible; this realisation is another example of their burgeoning conception of the mind as an interpretive device. Pretence The acquisition of pretend-play skills during early childhood is currently viewed as part of the development of children's knowledge about the mind, thanks largely to an important analysis by Leslie (1987, 1994). Leslie argues that the ability to understand pretence and the ability to understand false belief and other mental states are mediated by a common, early-maturing theory-of-mind module. This argument has some plausibility: ``Pretending

114

Flavell

that'' and ``believing that'' are both prepositional attitudes. Moreover, adults regard both as mental representations or construals of something as being a certain wayÐeither for real (belief ) or just temporarily, for play purposes (pretence). Nevertheless, Leslie's claim is currently controversial, (Lillard, 1998b). The related topic of children's understanding of imagination is also receiving considerable study (Woolley, 1995). Thinking Children achieve some important elementary knowledge and skills concerning thinking during the early pre-school years. For example, they come to construe it as an internal human activity that refers to or represents real or imaginary things. However, there are also important knowledge and skills concerning thinking that preschoolers clearly lack (Flavell, Green, & Flavell, 1995b; Flavell & O'Donnell, 1999). They are not aware that people are continually experiencing mental content spontaneously in an ever¯owing stream of consciousness. For example, unlike older children, preschoolers do not consistently attribute any mental activity at all to a person who just sits quietly, ``waiting''. These same dif®culties are equally evident when preschoolers are asked to report their own mental activity rather than another person's (Flavell, Green, & Flavell, 2000). That is, they tend to be very poor at reporting their own recent or present thinking, even in situations especially designed to make such introspection extremely easy (but see Estes, 1998, for an exception). They also do not differentiate very clearly between the mental contents of conscious and unconscious states (Flavell, Green, Flavell, & Lin, 1999). In particular, just as they tend to attribute too little ongoing ideation to a conscious person (they are unaware of the stream of consciousness), they also attribute too much to an unconscious one (they attribute conscious thought and self-awareness to people who are unconscious).

Differences in development Intracultural differences Investigators have recently been examining three kinds of differences in development: intracultural, intercultural, and interspecies (Flavell & Miller, 1998). Regarding intracultural differences, researchers have identi®ed some social experiences that appear to facilitate theory-of-mind development (Bartsch & Estes, 1996). For example, Ruffman, Perner, Naito, Parkin, and Clements (1998) have shown that young children who have more older siblings to interact with perform better on false-belief tasks than those who have fewer or none. Likewise, deaf children whose hearing parents are not ¯uent in sign language (as most are not) perform more poorly on a

6. Development of children's knowledge

115

false-belief test than deaf children of ¯uent-signing deaf parents (e.g., Peterson & Siegal, 1997). Such ®ndings suggest the importance of socialcommunicative experiences for the development of mentalistic understanding. This understanding in its turn undoubtedly facilitates the development of social skills (Watson, Nixon, Wilson, & Capage, 1999). The most striking intracultural differences, however, are manifest in the severe de®cits in theory-of-mind development exhibited by autistic children and adults (Baron-Cohen, Tager-Flusberg, & Cohen, 2000). How well these tragic de®cits can be reduced or compensated for by training is currently under study (Swettenham, 2000). There are also signi®cant intracultural differences among unimpaired individuals. Dweck and her co-workers have documented important individual differences in people's implicit theories about intelligence and other human attributes (Dweck, Chiu, & Hong, 1995). Textbooks in the ®elds of personality, social psychology, and social cognition also describe many other ways that normal adults differ from one another in their naive theories and knowledge regarding themselves and other people; great works of literature are an even richer source. And of course psychologists and other scientists have espoused widely different conceptions of human cognition and personality over the years: Just think of the differences between B.F. Skinner's and Freud's views of the mind. Intercultural differences The question of between-culture similarities and differences in theory-ofmind development is a fascinating one. How universal are the developments described in this article? An important review of the existing evidenceÐ mostly from ethnographic studiesÐsuggests that there are important differences between cultures in adult theories of mind (Lillard, 1998a). However, there is also reason to think that there are some deep universals as well (Avis & Harris, 1991; Wellman, 1998). Interspecies differences Researchers have also tested for theory-of-mind knowledge and abilities in other primates. Some recent experimental work with chimpanzees suggests that they may actually be less knowledgeable in this area than originally thought (Call & Tomasello, 1999; Reaux, Theall, & Povinelli, 1999). For instance, Reaux et al. (1999) present evidence suggesting that chimps may possess a behaviouristic rather than a mentalistic conception of seeing. That is, although they follow a person's gaze, they appear not to understand that the person sees and knows about things as a result of directing his or her gaze at them. However, the issue of chimp theory-of-mind understanding is far from settled (Hare, Call, Agnetta, & Tomasello, 2000; Povinelli & Giambrone, 2001).

116

Flavell

FUTURE RESEARCH Researchers have learned a considerable amount about the development of children's mentalistic understanding, especially from the theory-of-mind investigations of the past two decades. What will the next few decades bring? The following are some guesses (see also Flavell & Miller, 1998, pp. 882±887). As indicated in the Theories section, there is no shortage of theories intended to explain how children acquire an understanding of the mental world. It is a safe prediction that the future will see further theoretical and experimental work in this area. In particular, how the development of executive functioning may be related to theory-of-mind development will probably continue to be the subject of considerable inquiry (Hughes, 1998). Infant development is the hottest research area in the theory-of-mind ®eld currently and will probably continue to be for some time to come. It seems unlikely that researchers will discover new theory-of-mind acquisitions in this age period (i.e., things we did not know developed). Rather, they will be better able to date, describe, and explain developments already under study. The nonverbal research measures presently used are ingenious (e.g., looking time), but one can hope for even better ones in the future. Nonverbal measures will also continue to be used to compare the performance of other primates with that of human children. The recent work cited in the previous section demonstrates how fruitful this research strategy can be. My bet is that such comparisons will continue to suggest a lack of signi®cant understanding of mental states in nonhuman subjects. Towards the other end of ontogenesis, we can expect to see more research on middle-childhood, adolescent, and adult understanding. What do adolescents and adults know or believe about the mental world that elementary school children do not, and what understanding do the latter have that pre-schoolers do not? We still lack adequate answers to these questions. My guess is that there exist advanced competencies in this area that we have not yet identi®ed. How similar these more complex and subtle forms of understanding are from adult to adult, within and between cultures, is an important and clearly researchable question. One possible difference between children and adults that has not been explored much may lie not in what mental states they are capable of having conscious thoughts about, but rather in how easily and how often they are spontaneously conscious of them. It is one thing to know what thoughts and feelings are and that people have them, but quite another to spontaneously detect or predict their occurrence when they happen. Such increased sensitivity may be one of the products of life's advanced courses on mindreading. In future research investigators will also try to tell longer and richer developmental stories about each acquisition they study. Consider, for example, the acquisition of an understanding of belief (Flavell & Miller, 1998, pp. 873±875). Even if one accepts (still controversial) that 3-year-olds

6. Development of children's knowledge

117

tend not to understand beliefs but 4-year-olds do, it is becoming clear that this developmental story is incomplete. For the later part of the story, there is evidence that children continue to acquire knowledge about beliefs after the age of 4 (Carpendale & Chandler, 1996; Fabricius & Imbens-Bailey, 2000). As to the early part, Clements and Perner (1994) have found that young 3-year-olds who respond incorrectly to standard false-belief task questions nonetheless show by their eye movements that they may have some sort of rudimentary, implicit understanding of false beliefs. In a recent review, Haith and Benson (1998, p. 245) called for ``a graded approach to understanding infant cognition'', one that considers the succession of ``partial accomplishments'' that is likely to occur in each domain. Future research on theory-of-mind development will probably follow their prescription and try to identify a number of different levels of understanding of each mental state concept. There is relatively little research to date on how new acquisitions in this area translate into changes in children's everyday social and cognitive behaviours (Flavell & Miller, 1998; Slaughter, Dennis, & Pritchard, 2001). We will likely see much more of this kind of research in the future. There will also be efforts to ®nd out how best to aid children who need such help to acquire socially and academically useful theory-of-mind competencies (Swettenham, 2000). Whatever else it may turn out to be, the twenty-®rst century seems certain to be the ``Century of the Brain''. My ®nal prediction is that advances in neuroscience will help us understand better the child's developing understanding of the mental world (Baron-Cohen, Tager-Flusberg, & Cohen, 2000). REFERENCES Astington, J.W. (1993). The child 's discovery of the mind. Cambridge, MA: Harvard University Press. Astington, J.W., Harris, P.L., & Olson, D. (Eds.). (1988). Developing theories of mind. Cambridge: Cambridge University Press. Avis, J., & Harris, P.L. (1991). Belief-desire reasoning among Baka children: Evidence for a universal conception of mind. Child Development, 62, 460±467. Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. Cambridge, MA: MIT Press. Baron-Cohen, S., Tager-Flusberg, H., & Cohen, D.J. (2000). Understanding other minds: Perspectives from autism. Developmental cognitive neuroscience (2nd ed.). Oxford: Oxford University Press. Bartsch, K., & Estes, D. (1996). Individual differences in children's developing theory of mind and implications for metacognition. Learning and Individual Differences, 8, 281±304. Bartsch, K., & Wellman, H.M. (1995). Children talk about the mind. New York: Oxford University Press.

118

Flavell

Bennett, J. (1978). Some remarks about concepts. Behavioral and Brain Sciences, 1, 557±560. Bretherton, I., McNew, S., & Beeghly-Smith, M. (1981). Early person knowledge as expressed in gestural and verbal communication: When do infants acquire a ``theory of mind''? In M. Lamb & L. Sherrod (Eds.), Social cognition in infancy (pp. 333±373). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Brown, A.L., Bransford, J.D., Ferrara, R.A., & Campione, J.C. (1983). Learning, remembering, and understanding. In P.H. Mussen (Series Ed.), J.H. Flavell & E.M. Markman (Eds.), Handbook of child psychology: Vol. 3. Cognitive development (pp. 77±166). New York: John Wiley. Call, J., & Tomasello, M. (1999). A nonverbal false belief task: The performance of children and great apes. Child Development, 70, 381±395. Carlson, S.M., Moses, L.J., & Hix, H.R. (1998). The role of inhibitory processes in young children's dif®culties with deception and false belief. Child Development, 69, 672±691. Carpendale, J.I., & Chandler, M.J., (1996). On the distinction between false belief understanding and subscribing to an interpretive theory of mind. Child Development, 67, 1686±1706. Carpenter, M., Nagel, K., & Tomasello, M. (1998). Social cognition, joint attention, and communative competence from 9±15 months of age. Monograph of the Society for Research in Child Development, 63 (4, Serial No. 255). Clements, W.A., & Perner, J. (1994). Implicit understanding of belief. Cognitive Development, 9, 377±395. Dennett, D.C. (1978). Beliefs about beliefs. Behavioral and Brain Sciences, 1, 568±570. Dweck, C.S., Chiu, C.Y., & Hong, Y.Y. (1995). Implicit theories and their role in judgements and reactions: A world from two perspectives. Psychological Inquiry, 6, 267±285. Estes, D. (1998). Young children's awareness of their mental activity: The case of mental rotation. Child Development, 69, 1345±1360. Fabricius, W.V., & Imbens-Bailey, A.L. (2000). False belief about false beliefs. In P. Mitchell & K. Riggs (Eds.), Children's reasoning and the mind (pp. 267±280). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Fabricius, W.V., & Schwanen¯ugel, P.J. (1994). The older child's theory of mind. In A. Demetriou & A. Efklides (Eds.), Intelligence, mind, and reasoning: Structure and development (pp. 111±132). Amsterdam: Elsevier. Fein®eld, K.A., Lee, P.P., Flavell, E.R., Green, F.L., & Flavell, J.H. (1999). Young children's understanding of intention. Cognitive Development, 14, 463±486. Flavell, J.H. (1992). Perspectives on perspective taking. In H. Beilin & P. Pufall (Eds.), Piaget's theory: Prospects and possibilities (pp. 107±139). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Flavell, J.H. (1999). Cognitive development: Children's knowledge about the mind. Annual Review of Psychology, 50, 21±45. Flavell, J.H., Flavell, E.R., & Green, F.L. (1983). Development of the appearance± reality distinction. Cognitive Psychology, 15, 95±120. Flavell, J.H., Flavell, E.R., Green, F.L., & Wilcox, S.A. (1980). Young children's knowledge about visual perception: Effect of observer's distance from target on perceptual clarity of target. Developmental Psychology, 16, 10±12.

6. Development of children's knowledge

119

Flavell, J.H., Green, F.L., & Flavell, E.R. (1995a). The development of children's knowledge about attentional focus. Developmental Psychology, 31, 706±712. Flavell, J.H., Green, F.L., & Flavell, E.R. (1995b). Young children's knowledge about thinking. Monographs of the Society for Research in Child Development, 60 (1, Serial No. 243). Flavell, J.H., Green, F.L., & Flavell, E.R. (2000). Development of children's awareness of their own thoughts. Journal of Cognition and Development, 1, 97±112. Flavell, J.H., Green, F.L., Flavell, E.R., & Lin, N.T. (1999). Development of children's knowledge about unconsciousness. Child Development, 70, 396±412. Flavell, J.H., & Miller, P.H. (1998). Social cognition. In D. Kuhn & R.S. Siegler (Eds.), W. Damon (Series Ed.), Handbook of child psychology: Vol. 2. Cognition, perception, and language (5th ed., pp. 851±898). New York: John Wiley. Flavell, J.H., Miller, P.H., & Miller, S.A. (2002). Cognitive development (4th ed.). Englewood Cliffs, NJ: Prentice Hall. Flavell, J.H., & O'Donnell, A.K. (1999). Le deÂveloppement de savoirs intuitifs aÁ propos des expeÂriences mentales. Enfance, 51, 267±276. Gopnik, A., & Meltzoff, A.N. (1997). Words, thoughts, and theories. Cambridge, MA: MIT Press. Gopnik, A., & Wellman, H.M. (1994). The ``theory'' theory. In L.A. Hirschfeld & S.A. Gelman (Eds.), Mapping the mind: Domain speci®city in cognition and culture (pp. 257±293). Cambridge: Cambridge University Press. Haith, M.M., & Benson, J.B. (1998). Infant cognition. In D. Kuhn & R.S. Siegler (Eds.), W. Damon (Series Ed.), Handbook of child psychology: Vol. 2. Cognition, perception, and language (5th ed., pp. 199±254). New York: John Wiley. Hala, S. (Ed.). (1997). The development of social cognition. Hove, UK: Psychology Press. Hare, B., Call, J., Agnetta, B., & Tomasello, M. (2000). Chimpanzees know what conspeci®cs do and do not see. Animal Behavior, 59, 771±785. Harman, G. (1978). Studying the chimpanzee's theory of mind. Behavioral and Brain Sciences, 1, 576±577. Harris, P.L. (1992). From simulation to folk psychology: The case for development. Mind and Language, 7, 120±144. Hughes, C. (1998). Finding your marbles: Does preschoolers' strategic behavior predict later understanding of mind? Developmental Psychology, 34, 1326±1339. Johnson, C.N., & Wellman, H.M. (1980). Children's developing understanding of mental verbs: Remember, know, and guess. Child Development, 51, 1095±1102. Jost, J.T., Kruglanski, A.W., & Nelson, T.O. (1998). Social metacognition: An expansionist review. Personality and Social Psychology Review, 2, 137±154. Kuhn, D., (1999). Metacognitive development. In C. Tamis Le-Monda (Ed.), Child psychology: A handbook of contemporary issues. Philadelphia: Psychology Press. Leslie, A.M. (1987). Pretense and representation: The origins of ``theory of mind.'' Psychological Review, 94, 412±426. Leslie, A.M. (1994). ToMM, ToBy, and agency: Core architecture and domain speci®city. In L.A. Hirschfeld & S.A. Gelman (Eds.), Mapping the mind: Domain speci®city in cognition and culture (pp. 119±148). Cambridge: Cambridge University Press. Lillard, A.S. (1998a). Ethnopsychologies: Cultural variations in theories of mind. Psychological Bulletin, 123, 3±33.

120

Flavell

Lillard, A.S. (1998b). Wanting to be it: Children's understanding of intentions underlying pretense. Child Development, 69, 979±991. Meltzoff, A.N. (1995). Understanding the intentions of others: Re-enactment of intended acts by 18-month-old children. Developmental Psychology, 31, 838±850. Metcalfe, J., & Shimamura, A.P. (Eds.). (1994). Metacognition. Cambridge, MA: MIT Press. Mitchell, P. (1997). Introduction to theory of mind: Children, autism and apes. London: Arnold. Mitchell, R., & Riggs, K.J. (Eds.). (2000). Children's reasoning and the mind. Hove, UK: Psychology Press. Moore, C. (1996). Theories of mind in infancy. British Journal of Developmental Psychology, 14, 19±40. Moses, L.J., Baldwin, D.A., Rosicky, J.G., & Tidball, G. (2001). Evidence for referential understanding in the emotions domain at 12 and 18 months. Child Development, 72, 718±735. Moshman, D. (1998). Cognitive development beyond childhood. In D. Kuhn & R.S. Siegler (Eds.), W. Damon (Series Ed.), Handbook of child psychology: Vol. 2. Cognition, perception, and language (5th ed., pp. 947±978). New York: John Wiley. Perner, J. (1991). Understanding the representational mind. Cambridge, MA: MIT Press. Peterson, C.C., & Siegal, M. (1997). Psychological, physical, and biological thinking in normal, autistic, and deaf children. In H.M. Wellman & K. Inagaki (Eds.), The emergence of core domains of thought: Children's reasoning about physical, psychological, and biological phenomena (New Directions for Child Development, No. 75, pp. 55±70). San Francisco: Jossey-Bass. Pillow, B.H. (1995). Two trends in the development of conceptual perspectivetaking: An elaboration of the passive±active hypothesis. International Journal of Behavioral Development, 18, 649±676. Pillow, B.H., & Henrichon, A.J. (1996). There's more to the picture than meets the eye: Young children's dif®culty understanding biased interpretation. Child Development, 67, 803±819. Poulin-Dubois, D. (1999). Infants' distinction between animate and inanimate objects: The origins of naive psychology. In P. Rochat (Ed.), Early social cognition (pp. 257±280). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Povinelli, D.J., & Giambrone, S. (2001). Reasoning about beliefs: A human specialization? Child Development, 72, 691±695. Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1, 515±526. Reaux, J.E., Theall, L.A., & Povinelli, D.J., (1999). A longitudinal investigation of chimpanzees' understanding of visual perception. Child Development, 70, 275±290. Repacholi, B.M., & Gopnik, A. (1997). Early reasoning about desires: Evidence from 14- and 18-month-olds. Developmental Psychology, 33, 12±21. Ruffman, T., Perner, J., Naito, M., Parkin, L., & Clements, W.A. (1998). Older (but not younger) siblings facilitate false belief understanding. Developmental Psychology, 34, 161±174. Schneider, W., & Bjorklund, D.F. (1998). In D. Kuhn & R.S. Siegler (Eds.), W.

6. Development of children's knowledge

121

Damon (Series Ed.), Handbook of child psychology: Vol. 2. Cognition, perception, and language (5th ed., pp. 467±521). New York: John Wiley. Shantz, C.U. (1983). Social cognition. In P.H. Mussen (Series Ed.), J.H. Flavell & E.M. Markman (Eds.), Handbook of child psychology: Vol. 3. Cognitive development (pp. 495±555). New York: John Wiley. Shatz, M., Wellman, H.M., & Silber, S. (1983). The acquisition of mental verbs: A systematic investigation of ®rst references to mental state. Cognition, 14, 301±321. Slaughter, V., Dennis, M.J., & Pritchard, M. (2001). Theory of mind and peer acceptance in preschool children. Unpublished manuscript, University of Queensland, Brisbane, Australia. Swettenham, J. (2000). Teaching theory of mind to individuals with autism. In S. Baron-Cohen, H. Tager-Flusberg, & D.J. Cohen (Eds.), Understanding other minds: Developmental cognitive neuroscience (2nd ed., pp. 442±456). Oxford: Oxford University Press. Taylor, M. (1996). A theory of mind perspective on social cognitive development. In R. Gelman & T. Au (Eds.), E.C. Carterette & M.P. Friedman (Gen. Eds.), Handbook of perception and cognition: Vol. 13. Perceptual and cognitive development (pp. 283±329). New York: Academic Press. Taylor, M., Esbensen, B.M., & Bennett, R.T. (1994) Children's understanding of knowledge acquisition: The tendency for children to report they have always known what they have just learned. Child Development, 65, 1581±1604. Watson, A.C., Nixon, C.L., Wilson, A., & Capage, L. (1999). Social interaction skills and theory of mind in young children. Developmental Psychology, 35, 386±391. Wellman, H.M. (1983). Metamemory revisited. In M.T.H. Chi (Ed.), Trends in memory development research (pp. 31±51). Basel: Karger. Wellman, H.M. (1985). The child's theory of mind: The development of conceptions of cognition. In S.R. Yussen (Ed.), The growth of re¯ection in children (pp. 169± 206). San Diego, CA: Academic Press. Wellman, H.M. (1993). Early understanding of mind: The normal case. In S. BaronCohen, H. Tager-Flusberg, & D. Cohen (Eds.), Understanding other minds: Perspectives from autism (pp. 10±39). Oxford: Oxford University Press. Wellman, H.M. (1998). Culture, variation, and levels of analysis in our folk psychologies. Psychological Bulletin, 123, 33±36. Wellman, H.M., Cross, D., & Watson, J.K. (2001). A meta-analysis of theory of mind development: The truth about false belief. Child Development, 72, 655±684. Wellman, H.M., & Gelman, S.A. (1998). Knowledge acquisition in functional domains. In D. Kuhn & R. S. Siegler (Eds.), W. Damon (Series Ed.), Handbook of child psychology: Vol 2. Cognition, perception, and language (5th ed., pp. 523± 573). New York: John Wiley. Wellman, H.M., Harris, P.L., Banerjee, M., & Sinclair, A. (1995). Early understandings of emotion: Evidence from natural language. Cognition and Emotion, 9, 117±149. Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children's understanding of deception. Cognition, 13, 103±128. Woodward, A.L. (1998). Infants selectively encode the goal object of an actor's reach. Cognition, 69, 1±34. Woodward, A.L., & Markman, E.M. (1998). Early word learning. In D. Kuhn &

122

Flavell

R.S. Siegler (Eds.), W. Damon (Series Ed.), Handbook of child psychology: Vol. 2. Cognition, perception, and language (5th ed., pp. 371±420). New York: John Wiley. Woolley, J.D. (1995). The ®ctional mind: Young children's understanding of imagination, pretense, and dreams. Developmental Review, 15, 172±211. Woolley, J.D., & Wellman, H.M. (1992). Children's conceptions of dreams. Cognitive Development, 7, 365±380.

7

Domain-speci®c constraints of conceptual development Giyoo Hatano

The University of the Air, Chiba, Japan and

Kayoko Inagaki

Chiba University, Japan In the last few decades, we have witnessed some marked changes in cognitive development research. In the 1960s and early 1970s, the Piagetian theory of cognitive development dominated the research area. As is often pointed out (e.g., Karmiloff-Smith, 1992), that theory had two separable aspects. On the one hand, it posited that knowledge is constructed through the interaction between a child and the environment, via the complementary processes of assimilation and accommodation. On the other hand, it indicated that developing individuals' competence depends on their stages, in other words, logico-mathematical structures that are applicable across domains. Recent trends in conceptual development research, which became more and more notable during the last quarter of the twentieth century, were to emphasise domain speci®city instead of domain-general stages while preserving the Piagetian constructivistic view of cognitive development (Wellman & Gelman, 1998). In the current approaches, knowledge is divided into domains, without assigning special status to logico-mathematical knowledge. The knowledge in each domain is often supposed to constitute a naive or intuitive theory, that is, a body of coherent pieces of knowledge involving characteristic causal devices. Also, a variety of domain-speci®c constraints, which facilitate the acquisition and elaboration of the theory, have been proposed. Conceptual development, which constitutes a core of cognitive development, is no longer clearly distinguished from other forms of long-term learning or the acquisition of knowledge such as expertise in knowledge-rich domains or everyday cognition, because conceptual development is likewise de®ned as the construction of domain-speci®c knowledge under constraints. HISTORICAL BACKGROUNDS How were these trends produced? One may argue that they were induced by the recognised inadequacy of the Piagetian theory. During much of the 1980s and 1990s three successor positions appeared: those that emphasise

124

Hatano and Inagaki

the competence of young (pre-operational) children in a few selected domains of thought, those that pay special attention to within-individual variations in problem-solving competence due to accumulated experience in the relevant domains, and those that stress roles of the sociocultural context that serves as cognitive scaffolding. Each of them was sought to reinforce a soft point of the Piagetian theory. We adopt another interpretation, however, for the shift from the Piagetian theory to a variety of contemporary accounts. More speci®cally, we claim that, outside as well as inside the research area, a few streams of studies had revealed that the Piagetian notion of stages, de®ned by logicomathematical structures applicable across domains, was problematic, long before these studies became in¯uential among cognitive development researchers. These streams of research drew attention only when the research programmes based on the Piagetian theory were no longer regarded highly productive. We will discuss several of these studies brie¯y in the following paragraphs. For more detailed reviews, see Hirschfeld and Gelman (1994) for a discussion of domain speci®city and Keil (1981) for constraints. Among the streams that contributed to the shift from the domain-general formulations of cognitive development to the domain-speci®c ones, studies on animal learning are worth special mention. These studies at ®rst appear almost irrelevant, but have in¯uenced cognitive developmentalists (e.g., Gallistel, Brown, Carey, Gelman, & Keil, 1991), because they questioned the applicability of general laws of learning and suggested that experiencebased changes are species-speci®c and domain-speci®c. In domains such as identifying food and recognising conspeci®cs, they clearly showed not only that there are marked differences among species but also that there are domain-speci®c learning mechanisms that privilege some associations. For example, Garcia and Koelling (1966) demonstrated that rats very quickly come to avoid a certain taste if it is followed by internal bodily disturbances, but they do not if it is followed by electric shock. We cannot ignore Chomsky's contribution (e.g., Chomsky, 1988) to the advancement of current views on cognition in general and cognitive development in particular. Although he too remains outside the research area of conceptual development, his emphasis on modularity of language and his idea of ``Universal Grammar'' as an innately given acquisition device are considered to be major sources for the notions of domain speci®city and innate constraints. A stream of research on problem solving, which seemed to affect cognitive developmentalists through the cognitive science community, demonstrated the limitations of domain-general approaches to human and arti®cial intelligence and the importance of domain-speci®c knowledge as the basis for problem-solving competence (e.g., Chi, Glaser, & Rees, 1982). Although people may use several domain-general heuristics such as means±ends analysis (Newell & Simon, 1972), it is dif®cult, time-

7. Conceptual development

125

consuming, or even impossible to solve problems solely with these heuristics. Therefore, individual competence varies considerably from domain to domain, depending upon how much experience is accumulated in each domain. Moreover, what is the most critical determinant of problemsolving competence is not general intelligence but the relevant domainspeci®c knowledge. Still another stream of studies concerned cross-cultural comparisons and cognition in everyday situations. Those studies clearly showed that the dif®culty of problems sharing the same logical structure varies greatly depending on sociocultural contexts in which they are presentedÐone can only solve problems that are translated into familiar or ethnographically valid forms (Cole, Gay, Glick, & Sharp, 1971). These studies also showed that the knowledge and skills that are acquired are related directly to the activity one engages in and that generalisations from such experience are highly limited. In addition, many careful studies conducted within the Piagetian research programme cast doubt on the ideas of broad stages of development and of major qualitative shift (Gelman & Baillargeon, 1983). In other words, many proponents of the successor positions were, in part, forced to become so by their own experimental ®ndings. It this sense, the Piagetian theory produced its own successors. It should be noted, however, that the studies outside cognitive development, mentioned earlier, may also have been in¯uential. For instance, the view that very young children possess naive theories for selected aspects of the world could not have become popular without the preceding ethological studies as well as Chomsky's formulation. How much have we progressed in understanding the growth of the mind by this shift from the Piagetian theory to the current views of domainspeci®c construction under constraints? There are a few obvious advantages of the current views, as will be discussed in detail in the next section. However, because the current views are based on the several streams of studies that have different foci and emphases, although they all favour domain-speci®c mechanisms, the views do not yet constitute a coherent theory of conceptual development that completely replaces the Piagetian theory. We still face some serious problems. Among others, the notion of ``constraints'', which play key roles in the current views of conceptual development, is not yet well articulated. In what follows, we ®rst summarise the current views of conceptual development. Next, we reconceptualise the notion of ``innate constraints'', more speci®cally as learning mechanisms, rather than as innate knowledge or representational content. We then expand the notion of ``constraints'' to include interactive, sociocultural constraints as well as internal, cognitive ones. Through these attempts, we specify what types of domain-speci®c constraints are involved in conceptual development and how they are related to one another. Finally, we offer our prospect for the future of theories about conceptual development.

126

Hatano and Inagaki

CURRENT VIEWS OF CONCEPTUAL DEVELOPMENT How does conceptual development take place? We summarise in this section the views of conceptual development as ``the domain-speci®c construction of knowledge under constraints'' in terms of four interrelated characterisations. Whereas the ®rst two characterisations are inherited from the Piagetian theory, the last two are in contrast with it. We also point out a few problems inherent in the views.

Knowledge is acquired by construction The ®rst characterisation indicates that knowledge is acquired by construction, not by transmission alone (Resnick, 1987). That knowledge is constructed was an idea inspired by Piaget, and constitutes the ``Zeitgeist'' shared by current researchers. Humans interact with the environment, ®nd regularities, and construct condition±action rules and more elaborate forms of knowledge (Hatano & Inagaki, 2000). Humans acquire knowledge richer than the knowledge they are presented with, or even invent knowledge that has never been presented, often as a by-product of their problem-solving and/or comprehension activities. Knowledge can be transmitted to some extent, but transmitted knowledge becomes applicable to a variety of situations only after it is reconstructed by being interpreted and connected to the prior knowledge of the learner.

Knowledge acquisition involves restructuring The second characterisation of conceptual development is that it involves restructuring; that is, not only does the amount of knowledge increase but also one's body of knowledge is reorganised as more and more pieces of knowledge are acquired. This process of reorganisation is now often called conceptual change or theory change (Carey, 1985), which, although domain-speci®c, may have been described as a qualitative change in structures or as a shift from one stage to another in the Piagetian terminology. Knowledge systems before and after the conceptual change may sometimes be locally incommensurable; that is, some pieces of knowledge in one system cannot properly be translated into the other, as exempli®ed by the shift from children's undifferentiated heat/temperature concept to adults' two separate concepts of heat and temperature (Wiser & Carey, 1983). However, restructuring can also include milder and more subtle forms of change: Concepts can be substantially modi®ed, but the old and new concepts are not incommensurable (e.g., the concept of ``animal'' is expanded to include birds, ®sh, and insects). Relationships between concepts can also change, and the same phenomenon may be explained differently as causal devices change.

7. Conceptual development

127

Knowledge is acquired domain by domain The third characterisation emphasises that most knowledge is acquired domain by domain in conceptual development. As mentioned earlier, it represents the greatest difference between the Piagetian and current views. Intra-individual variations in problem-solving, comprehension, and learning are now explained in terms of domain-speci®c prior knowledge. Children as well as adults reveal higher levels of competence in those domains in which they possess richer and better-structured bodies of knowledge. Horizontal decalage, which was considered exceptional and puzzling by Piagetians, is now taken for granted. The term ``domain'' in conceptual development refers to a range of knowledge that is covered by a theory: Hirschfeld and Gelman (1994) characterised the domain as ``a body of knowledge that identi®es and interprets a class of phenomena assumed to share certain properties and to be of a distinct and general type'' (p. 21). How domains are divided is in a sense a cultural product (Laboratory of Comparative Human Cognition, 1983). However, how knowledge is divided into domains is also determined at least in part on innate bases. For these reasons, the de®nition of domains differs widely among scholars. Although we do not discuss it here, specifying the notion of domains is thus an important challenge for conceptual development researchers. Recent studies have demonstrated, however, that individual competence varies considerably from domain to domain, irrespective of how domains are de®ned. To reiterate, the Piagetian stage theory, which posited that an individual's competence depends on his or her logico-mathematical structures, which are applicable across domains, has been challenged or even rejected by many current researchers (Gelman & Baillargeon, 1983; LCHC, 1983; Siegler, 1978). Many contemporary researchers believe instead that the course and process of development vary from domain to domain; in other words, knowledge is acquired, in part, in a unique fashion in each domain. However, a small number of pieces of knowledge are shared by a number of domains (e.g., knowledge about literacy or measurement), and analogical transfer or generalisation of knowledge based on recognised isomorphisms across domains may sometimes occur.

The process of knowledge acquisition is constrained The fourth characterisation indicates that the process of knowledge acquisition is constrained (Gelman, 1990; Keil, 1981). Here the term ``constraints'' refers to conditions or factors that facilitate the process of acquisition as well as restrict its possible range. Some constraints are domain-general (e.g., architectural constraints as well as logico-mathematical knowledge aÁ la Piaget), and others are domain-speci®c (Keil, 1990). Here we will be concerned only with the latter, because domain-speci®c constraints have attracted the attention of recent researchers, who are particularly interested

128

Hatano and Inagaki

in early competence in a few selected domains. These constraints serve to carve experiences into different domains and also to enhance the accumulation and elaboration of knowledge in relevant domains. The acquisition of some forms of knowledge is easy, whereas that of others is dif®cult owing to differences in domain-speci®c constraints. The construction and successive revision of knowledge in a domain takes place under the same set of constraints so that the acquired knowledge in the domain is often similar, if not identical, in different individuals. Many investigators agree that the process of construction is constrained both by innate tendencies and by acquired prior knowledge (i.e., by cognitive constraints), although, as will be discussed in the next section, the exact nature of innate constraints is yet to be speci®ed and has been the target of heated debate. Many current researchers, although this number is fewer, also assume that the construction process is constrained interactively (i.e., by sociocultural constraints). The sociocultural constraints include usable artefacts that are shared by a majority of people in a community or its subgroup, including physical facilities and tools, social institutions and organisations, documented pieces of knowledge, and common sense and beliefs. They also include the behaviour of other people, interactions with them, and social contexts created by them. These constraints are differentiated from external, task constraints (Keil, 1990), because they do not formally restrict the search process. Instead, sociocultural constraints operate through continuous interaction between the cogniser and his or her responsive environment. These constraints were neglected until recently, and are still ignored by many researchers, because cognitive developmentalists traditionally concentrate on the study of how individuals manipulate symbols within their mind. How to conceptualise these constraints is a challenge for conceptual development research, and we will discuss this issue in the fourth section. To summarise, although current views on conceptual development can be characterised in succinct and coherent ways, these views still have shortcomings. Among others, the notion of ``constraints'' is not well speci®ed. More speci®cally, (a) researchers disagree how best to conceptualise ``innate constraints''; and (b) ``sociocultural constraints'', which uniquely characterise human development, have not been given their due. In other words, proponents of the current views all agree that conceptual development takes place under domain-speci®c constraints, but they widely differ in characterisations of the constraints. DOMAIN-SPECIFIC INNATE CONSTRAINTS IN CONCEPTUAL DEVELOPMENT What are innate constraints and how do they operate in conceptual development? As pointed out earlier, how best to conceptualise domain-speci®c innate constraints is still debatable, even among those investigators who

7. Conceptual development

129

believe in the importance of such constraints in conceptual development. In this section we propose our own moderate view. As Keil (1990) pointed out, the need for some sort of constraints on learning is undeniable because, without constraints, it would take more than a reasonable amount of time and effort to choose a promising hypothesis or interpretation for a given set of observations. The necessity of assuming constraints is even more obvious in conceptual development, in which basic knowledge systems such as naive physics, psychology, and biology are acquired early and uniformly (Wellman & Gelman, 1998). In fact, a number of studies have shown that even 3-year-olds have clearly acquired naive physics and psychology (e.g., Wellman & Inagaki, 1997). Although it is still debatable whether 3-year-olds have the biological domain of thoughtÐmore speci®cally, whether they possess a uniquely biological mode of reasoningÐit is now agreed that children possess autonomous biology before formal schooling (Carey, 1995; Hatano & Inagaki, 1994; Inagaki & Hatano, 2002). To put it differently, we have to explain (1) how a child can ®nd which aspects of the world he/she should pay attention to and how he/she should segment and encode the information obtained (attention/encoding issues), and (2) how a child can pick out a good candidate hypothesis when an almost in®nite number of hypotheses and interpretations are consistent with the sample observations (induction issues). Most developmentalists agree that the solution to these ``learnability'' issues requires us to assume some constraints that force the child to pay attention to particular aspects of the object or event and encode them in particular ways, and that force him/her to entertain a restricted set of hypotheses. However, there is no consensus about what form these constraints take, because this is exactly the heart of the debate between nativists and emergentists, the latest version of the time-honoured nativist/empiricist controversy (Keil, 1998). Nativists (e.g., Spelke, 1994) assume that there are innate constraints in the form of pieces of knowledge or representational contents that constitute the core of the theory in each domain. In contrast, emergentists (e.g., Elman et al., 1996; Munakata, McClelland, Johnson, & Siegler, 1997) claim that the domain-speci®c knowledge that serves as constraints on further learning is a product of the interactions between such domain-general constraints as architectural and temporal constraints and patterns of inputs or learning tasks in a particular domain. In other words, emergentists admit that some domain-general constraints are innate, but do not believe that there can be innate representations. How can we reconcile these opposing views of innateness and domain speci®city? We propose a set of assumptions that represent the middle ground. More speci®cally, we assume that those domain-speci®c constraints that seem to be innate (1) are acquired very early if not innate; (2) take the form of biases and preferences, not speci®c pieces of knowledge; (3) serve as learning mechanisms; and (4) are critically important at the initial phase of

130

Hatano and Inagaki

conceptual development, in other words, are indispensable for bootstrapping. They can best be described as general tendencies or biases (or general, abstract principles) that guide the acquisition of more speci®c, and immediately useful, pieces of knowledge through concrete experiences. We will elaborate on each of these assumptions in the following paragraphs.

Some domain-specific constraints are innate or acquired very early First, although we believe that humans are genetically endowed with some domain-speci®c constraints, we agree that this is just an assumption, and that it is almost impossible to show that these constraints are innate in any strict sense. These constraints operate not just at birth but appear later. In this case it is possible to explain the time lag in terms of either an individual's maturation or early experience. However, if a knowledge system deals with aspects of the world important for the survival of the human species and thus is shared by most adults both within and between cultures, if the knowledge system is acquired early and without dif®culty, and if it is hard to attribute the acquisition to experiential factors, we can reasonably hypothesise that humans possess innate domain-speci®c constraints for acquiring the knowledge system. In this sense, as Atran (1994, p. 318) put it, ``such `core' domains of human knowledge seem to operate under severe cognitive constraints'', whether they are innate or acquired very early.

Innate constraints are tendencies Second, we assume that innate or early cognitive constraints take the form of biases and preferences, because such constraints can effectively mitigate the learnability problems. We do not regard early cognitive constraints as speci®c pieces of knowledge for three reasons: (1) there is no evidence that speci®c pieces of knowledge can be genetically transmitted or neurally prewired; (2) recent studies have shown that what seem to be innately given pieces of knowledge emerge from the interaction between a cognitive system and a particular learning task (e.g., Elman et al., 1996); and (3) because the process of evolution is very slow, possessing speci®c pieces of innate knowledge may be detrimental when ecological environments change. For example, even if humans were endowed with some pieces of knowledge about a speci®c animal X, such knowledge would be useful only when X was available in their environment. If humans were born in an ecological niche containing an animal Y that was apparently similar to but behaved differently from X, having the same pieces of knowledge might be harmful.

Innate constraints are domain-specific learning mechanisms Third, our assumption of innate or early cognitive constraints implies that children need to have some experience with the target phenomena in order

7. Conceptual development

131

to acquire naive theories. We do not claim that some domain-speci®c pieces of knowledge are innate, whereas others are acquired through experience. Rather, we claim that every piece of the basic knowledge system is acquired, but this acquisition process is enhanced by relevant innate tendencies. This aspect of our assumption seems close to Baillargeon, Kotovsky, and Needham's (1995) claims regarding infant physical cognition. It is also similar to Carey (1995), who stated, ``innate domains are learning mechanisms'' and thus ``to characterize an innate domain one must characterize the precise role of experience . . .'' (p. 303). We would like to emphasise that the notion of constraints is not based on direct observation but has been introduced to solve the learnability problems. In short, such constraints are posited to enhance early differentiated understanding of important aspects of the world, because it is highly unlikely that domains can be differentiated so early in life by domain-general learning mechanisms alone.

Innate constraints are indispensable for bootstrapping Fourth, domain-speci®c innate (or very early) constraints are indispensable only during the construction of the initial naive theory of a domain. This is because, once such a theory becomes available, the theory itself serves as a set of constraints for further learning or the elaboration and revision processes. Although we are often fascinated by infants' competence, the process of conceptual development takes many years before adult versions of a naive theory are constructed. Innate or early constraints do not play dominant roles throughout the course of development. Rather, their roles become less and less important as acquired pieces of knowledge increase and sociocultural constraints are exploited more effectively, as will be discussed in the next section. Based on the foregoing assumptions, we propose the following scenario. Innate constraints operating in the domains serve to control and direct attention or coding, and restrict the range of hypothesis space to be searched. One set of constraints in each domain draws attention to relevant aspects of the target objects or phenomena so that even young children can distinguish those that should be interpreted within the domain from those that should not. In the case of naive biology, for example, there seem to be a set of constraints directing attention to those aspects of living things that serve to distinguish them from nonliving things. Even infants can differentiate animate entities (humans) that spontaneously move from inanimate things that do not move by themselves (Golinkoff, Harding, Carlson, & Sexton, 1984). By the time they reach age 3 or 4 years, children can predict whether a given object can move up and down a hill by itself, disregarding its physical appearance (Massey & Gelman, 1988). Another set of constraints serves to eliminate in advance a large number of logically possible interpretations or hypotheses. Again, an example from

132

Hatano and Inagaki

naive biology helps to clarify this point. As mentioned earlier, even rats tend to attribute illness to the consumption of a novel food item (Garcia & Koelling, 1966). It seems likely that humans are endowed with similar constraints; for example, they try to ®nd the cause of diarrhoea among a variety of foods, whereas they seek a physical cause for a cut. These innate constraints may begin to work after a few experiences are given. Their operation is functionally equivalent to the possession of an abstract principle, which precludes a range of alternatives to be considered before speci®c information is known. In other words, it is assumed that, because those tendencies and biases enable humans to search the hypothesis space highly selectively, they can reach, in most cases, a reasonable interpretation promptly, and thus can accumulate pieces of knowledge constituting a naive theory. We believe that the foregoing assumptions represent a number of conceptual development researchers' ideas in the middle ground, and are acceptable to some extent to those who are sympathetic with emergentists and those who are regarded as being nativists. We consider KarmiloffSmith and Spelke to be good examples of the two opposing camps, respectively. Karmiloff-Smith (1992) indicated that ``we are beginning to witness an increasing tendency on the part of connectionists to explore the ways in which domain-speci®c constraints might shape learning'' (p. 180). She also stated that ``there is nothing about the connectionist framework that precludes the introduction of initial biased weights'' (p. 179), that is, weights that represent innately speci®ed predispositions instead of random weights, which are usually adopted in connectionist simulations. Even in 1996, after she had joined the connectionist camp headed by Elman, she, in her singleauthored essay, distinguished representational constraints from representational contents, and indicated that we should avoid a strong nativism ``in favour of minimal representational biases'' (Karmiloff-Smith, 1996, p. 10). Thus she seems to assume that at birth babies have the tendency to pay attention to important inputs in the environment, although she may emphasise that such tendencies are subcortical. Similarly, the notion of innate constraints as learning mechanisms seems to be within an acceptable range to Spelke, although she is clearly a nativist in her sole-authored papers (e.g., Spelke, 1994). For example, Spelke and Newport (1998) offer two possibilities regarding the language acquisition device: ®rst, children may know what is universal about languages; second and alternatively, they possess ``learning mechanisms that do not `know' anything about linguistic universals . . . but rather perceive, remember, or analyze linguistic input in ways that bias certain outcomes'' (p. 306). We believe this view can also be applied to conceptual development. We also believe that our assumptions are supported empirically. As will be reviewed in the next section, it is clear that naive biology does not take a universal form across cultures. Rather, it is instantiated differently, depending on the direct and indirect experiences of individuals. As for naive psychology, its normal development requires intensive communication with

7. Conceptual development

133

signi®cant others. Even for naive physics, very few physical principles operate immediately after birth. To summarise, each domain of thought is acquired through experience, particularly experience that leads to predicting and interpreting typical ``behaviours'' of representative objects of the domainÐfor instance, in the case of naive physics, the motion of inanimate, solid objects, and in the case of naive psychology, the social activity of humans. However, it is true that conceptual development in selected domains is early, easy, and relatively uniform, probably because it is helped by domain-speci®c innate constraints, which direct children's attention to relevant behaviours or objects, and focus the search process for promising modes of causal reasoning. To put it differently, we assume that innate constraints work, but, unlike radical nativists, we do not posit that the entire course of conceptual development is strongly in¯uenced by innate constraints or that there is little room for other types of constraints to work; instead, we assume that innate constraints are critically important only at the initial phase, and both acquired knowledge and sociocultural constraints operate, together with innate constraints, even in the initial phase. SOCIOCULTURAL CONSTRAINTS IN CONCEPTUAL DEVELOPMENT Conceptual development takes place not under innate constraints alone. As children accumulate experiences in sociocultural contexts, such experiences are represented in the form of pieces of domain-speci®c knowledge, which work as acquired cognitive constraints in further development (Keil, 1990). These constraints are likely to operate to a greater extent later in conceptual development. We believe that much of the ``prior knowledge'' in human development and learning has a sociocultural origin, because the growth of the human mind is achieved by incorporating experiences accumulated in earlier generations in the form of culture (Inagaki & Hatano, 2002). Conceptual development is also enhanced by interactive, sociocultural constraints. When we solve a problem, understand a target phenomenon, or learn something new, we do not merely manipulate symbols that represent the external world in our head. Because we continuously interact with other people and artefacts, our competence in these cognitive activities is heavily dependent on them. In addition to providing developing individuals with materials by which they can construct knowledge, sociocultural contexts serve as constraints in two different ways: as cognitive constraints that have a sociocultural origin and as the human and instrumental scaffolds that are available to the cogniser. Although the former type of constraints have been considered in the dominant views of conceptual development fairly often, the latter, interactive constraints have not been given due attention. We assert in the following paragraphs that both types of sociocultural

134

Hatano and Inagaki

constraints play indispensable roles throughout the course of conceptual development.

Conceptual development is influenced by sociocultural factors It is often asserted that the sociocultural contribution can be conceptualised as merely facilitating or inhibiting the ®xed course of the initial phase of conceptual development (e.g., Carey & Spelke, 1994). We claim, on the contrary, that sociocultural factors in¯uence conceptual development from the beginning by embodying abstract or skeletal core principles in each domain of thought, because innate constraints or ®rst principles are so skeletal (Gelman, 1990) that they have to be instantiated by using socioculturally provided data. In other words, sociocultural contexts necessarily produce variations in conceptual development. To elucidate the extent to which culture or society in¯uences conceptual development, we will examine variations in the initial acquisition of naive theories, using the formation of young children's biological taxonomy as an example. If sociocultural factors contribute to the acquisition of and changes in basic conceptual devices such as biological taxonomy, we can reasonably expect that their in¯uence will be even larger for other devices that are learned later. Atran (1998) asserts that the taxonomy of living kinds is more or less universal (even in adult's naive biology) because it is a product of ``an autonomous, natural classi®cation scheme of the human mind'' (p. 567). However, as he also admits, prototypes may be different from culture to culture. A high frequency of contact, the sharing of a large number of properties with other members, and an ecological signi®cance all contribute to making the taxon most typical or truly representative of the animal or plant category to which it belongs. Likewise, conceptualising relationships between biological or ontological categories, unlike the categorisation itself, has to rely on cultural models, because it is a matter of interpreting the world. Even if children's naive-biological categories are universal, the distance between them may be socioculturally variable. In fact, engaging in activities for raising animals and plants may lead children to accept the target organism as prototypical, although it is reported that the children in normative circumstances exclusively use humans as a prototype (e.g., Carey, 1985; Inagaki & Hatano, 1987). Inagaki (1990) compared the biological knowledge of 5-year-olds who had actively raised gold®sh for an extended period at home with that of children of the same age who had never raised an animal. Although these two groups of children did not differ in factual knowledge about typical mammals, the gold®sh raisers had much richer procedural, factual, and conceptual knowledge about gold®sh. More interestingly, the gold®sh raisers used their knowledge about gold®sh as a source for analogies in predicting the reactions of an unfamiliar ``aquatic'' animal that they had

7. Conceptual development

135

never raised (i.e., a frog), and produced reasonable predictions that they could support with some explanations. For example, when asked whether we could keep a baby frog the same size forever, one of the gold®sh raisers answered, ``No, we can't, because a frog will grow bigger as gold®sh grew bigger. My gold®sh were small before, but now they are big''. Moreover, the gold®sh raisers tended to use person analogies as well as gold®sh analogies in making predictions about a frog. In other words, the gold®sh raisers possessed a different form of biology to the nonraisers, because they could use two sources for making analogical predictions. Thus, the knowledge about gold®sh acquired by raising them helped the young children's inference, and probably their future learning, in biology. That rich and well-organised knowledge enables young children to employ advanced deductive reasoning was also shown by Chi, Hutchinson, and Robin (1989). In this study, children who were dinosaur ``experts'' assigned unobservable attributes to novel dinosaurs using category-based inference (e.g., ``He's pretty dangerous. 'Cause he's a meat-eater''), or classi®ed dinosaurs based on superordinate categories (e.g., ``These dinosaurs are plant-eaters. They aren't harmful''). Another aspect of biological taxonomy that may be in¯uenced markedly by culture or society is the distance between ontological categories. For example, in some cultures plants may be regarded as similar to humans and animals, whereas in others they are considered very close to nonliving kinds; conversely, some nonliving entities may be treated as somewhat similar to living things. Hatano et al. (1993) asked 6-, 8-, and 10-year-olds from Israel, Japan, and the United States whether two instances each of four entity classes (people, nonhuman animals, plants, and nonliving things) possessed each of 16 attributes including life (being alive), unobservable animal attributes (e.g., has a heart), sensory attributes (e.g., feels pain), and attributes true of all living things (e.g., grows). They found both similarities and differences across cultures, but here we will discuss only a few marked cultural differences seen among the youngest groups. First, the young Israeli children were considerably more likely than their cross-national counterparts not to attribute life to plants and not to attribute other properties that are shared by all living things (with even a larger crossnational difference among older children). Second, the young Japanese children, whose overall accuracy was comparable to the Israelis, were considerably more likely to attribute to nonliving things properties that are unique to living things (with no cross-national difference among older children, who generally did not attribute living characteristics to nonliving things). These differences are especially interesting because they suggest that the early form of biology is in¯uenced by beliefs within the culture. Features of culture and language may account for Israeli children being less apt than American or Japanese children to attribute to plants life status and properties of living things. As suggested by Stavy and Wax (1989), within the Israeli culture, plants are regarded as very different from humans

136

Hatano and Inagaki

and other animals in their life status. This cultural attitude parallels that of a Biblical passage (Genesis, 1: 30) indicating that plants were created as food for ``living things'' including mammals, birds, and insects. Perhaps re¯ecting their cultural beliefs, the Hebrew word for ``animal'' is very close to that for ``living'' and ``alive'', whereas the word for ``plant'' has no obvious relation to such terms (Stavy & Wax, 1989). From a similar perspective, we can suggest a reason why young Japanese children might be more likely than their American or Israeli counterparts to view some nonliving things as alive and possessing attributes of living things. In Japanese folk psychology, even nonliving things are sometimes considered to have minds. It is at least not a silly idea for a Japanese person to assign life or divinity to nonliving things, especially big or old ones. In addition, linguistic factors seem to in¯uence Japanese children's attributional judgements. Do particular experiential variables in¯uence conceptual development in other domains such as naive psychology or naive physics? With regard to naive psychology, there have been several studies indicating that the answer is af®rmative. For example, Peterson and Siegal (1997) found that deaf children raised by hearing parents, and thus without ¯uent signing conversational interaction, had selective de®cits in naive-psychological reasoning, whereas deaf children raised in families with a ¯uent signer did not show such de®cits. Dunn, Brown, Slomkowski, Tesla, and Youngblade (1991) reported that not only the mother±child discourse about feelings and causality at home but also the child's cooperative interaction with his or her sibling were associated with differences in the child's performance on a theory-of-mind task 7 months later. These ®ndings suggest that the conversational experience with signi®cant others contributes to the development of naive psychological understanding. As to naive physics, there have been few studies clearly showing the in¯uence of particular experiences, probably because controlling particular physical experience of infants is very dif®cult. However, it often takes several months of interactions with the world before infants reveal some general understanding of physics, such as that objects are solid. As Baillargeon et al. (1995) put it, such understanding is more likely to be the product of domain-speci®c learning mechanisms and infants' repeated perceptual-motor experiences than the manifestation of innate knowledge.

How sociocultural constraints operate in interaction As we have seen earlier, accumulated experiences in sociocultural contexts produce variations in conceptual development. This is attributed in part to the fact that pieces of knowledge representing these experiences serve as constraints for further development. But that is not the only in¯uence. Sociocultural constraints such as other people who have extensive prior experience and shared physical, symbolic, and social tools also help devel-

7. Conceptual development

137

oping individuals acquire some knowledge through continuous interaction with them. When interacting with other people, our cognition and learning are constrained in the following ways (Inagaki & Hatano, 2002; Tomasello, Kruger, & Ratner, 1993): What we observe is not randomly selected out of almost in®nite pieces of information but directed by joint attention, what we try is not a randomly chosen chain of responses from our repertory but is often triggered by imitation, and how we interpret a set of observations is in¯uenced by guided comprehension activity. Sociocultural constraints enable us not only to be competent but also to acquire knowledge and skills readily. These constraints are domain-speci®c, because different sets of constraints work in different activities, though mechanisms through which the constraints operate are the same. Imagine a case where children raise an animal at home. In one likely scenario, children learn from their parents or older siblings who already know how to raise the animal of interest. The children are likely to learn the procedures involved promptly, because their observation and choice of procedures are directed by joint attention with or by imitation of their more mature partners. In addition, active engagement in raising animals may lead children to acquire a sort of conceptual knowledge or a mental model of the raised animals, often helped by guided comprehension activity induced for the sake of the animals. Thus, children tend to be more competent when they engage in a joint activity than in a solitary activity, and construct and elaborate on relevant pieces of knowledge more promptly (see also Rogoff, 1990). Those tools or artefacts that are used in raising animals also serve to direct children's attention and narrow the range of procedures and interpretations they consider. For example, there is a manual for raising gold®sh and other pets, which may include some pieces of information that are understandable to the children. A ®sh tank or a cage may suggest what kind of care the animal needs. Some special food may be available so that the procedure for feeding the animals can readily be standardised. Although the enabling constraints offered by tools are primarily needed for successfully performing raising tasks, they tend to enhance the acquisition of knowledge as well. To generalise, even young children acquire knowledge in a uniquely human way, in other words, through participating in practices. Their learning is helped by interactive sociocultural constraints, the operation of which is based on their ability to mentally represent other people's mental representations (Tomasello et al., 1993). However, we would also like to point out that even before their participation in practices, children may be socialised or enculturated in their modes of thought. For example, parents often talk to toddlers or even infants about aspects of the world in discourse that re¯ects cultural models shared in the community. Although very young children's comprehension ability is limited, their minds are subject to sociocultural in¯uences to some extent. As Tomasello (1992)

138

Hatano and Inagaki

observed, a child as young as 2 years of age could recognise an adult's intention when a pragmatic context was given. Taken together, sociocultural constraints operate in conceptual development from a very early age, complementing innate constraints. FUTURE POSSIBILITIES As we have seen, the various successor positions of the Piagetian theory, although similar in many ways, differed widely in their emphases concerning where they thought the theory needed to be modi®ed and how such revision had to be done. Because of these different emphases, the successor positions have not been integrated into a larger, consistent theoretical framework. Perhaps such integration can be done, which might lead to a real alternative to the Piagetian theory. It could even produce a new, more general theory of knowledge acquisition for future research efforts (see Hatano & Inagaki, 2000). We have proposed an attempt towards such a theory in this chapter, and we naturally expect conceptual development research in the near future to become more and more integrative, rather than stressing only one type of constraints while neglecting others. However, there could be other trajectories as well. We assume, as the basis for divining the future, that the development of a scienti®c theory is enhanced by attempts in two opposite directions, that is, (1) efforts to make the target theory more integrative, incorporating different mechanisms or causal devices, and (2) efforts to maximise the explanatory power of a given mechanism. We call efforts (1) ``integrative'' and efforts (2) ``exploitative''. Using this terminology, the Piagetian theory was exploitative in its emphasis on the domain-general logico-mathematical structures as constraints as well as products of cognitive development, but integrative regarding the factors affecting the development. For example, Piaget listed four factors in his discussion on psychobiological development: hereditary factors, physical environment, social environment, and equilibration that regulates their interactions (Piaget, 1956). The framework that we are proposing so far is, in contrast, exploitative in terms of its emphasis on domain-speci®c constraints, but integrative in that innate tendencies, acquired knowledge, and interactive sociocultural constraints are all taken into account. Thus, an alternative attempt that may become salient in future would be the attempt that integrates both domain-general structures and domainspeci®c constraints. As appropriately pointed out by Keil (1990), it is undeniable that some constraints (as well as some pieces of knowledge) are domain-general. Most of the current investigators of conceptual development, except for those who are called neo-Piagetians, have not paid due attention to domain-general structures that serve as constraints, only because they have focused on trying to extend the domain-speci®c explanation to its limits.

7. Conceptual development

139

Another set of alternatives might be exploitative in an emphasis on one or another type of the domain-speci®c constraints. Such efforts are needed in order to clarify or even increase conceptual power of each type of constraints, and a considerable number of researchers seem to be most concerned with demonstrating the importance of their own favourite constraints. For example, some radical proponents of the modularity/innate constraint accounts have argued, from evolutionary psychology perspectives, that innate constraints are extremely important for almost all aspects of cognitive development (e.g., Barkow, Cosmides, & Tooby, 1992). They claim that what we human beings can readily understand, solve, or learn is based on the capabilities that our species have acquired through millions of years of evolution. They sometimes assert that these capabilities take forms of ``innate knowledge'' that is written into our genetic code. Similarly, there are some radical proponents of the sociocultural account called the ``situated cognition view''; they believe that our competencies are dependent almost solely on interactive sociocultural constraints. According to these proponents, development is no more than changes in the mode of participation in socioculturally organised practices (e.g., Rogoff, 1998), and what individuals learn is how to rely on such constraints. Still another group of investigators considers development to be basically the acquisition of domain-speci®c knowledge through reliance on the prior knowledge of the domain. Such investigators can be found not only among developmental researchers who also study expertise (e.g., Chi et al., 1989) but also emergentists (e.g., Elman et al., 1996) who strive to minimise the importance of any types of constraints other than prior knowledge (and domain-general architectural constraints). It will be interesting to see, in the twenty-®rst century, who attain hegemony in conceptual development research and how long their domination lasts. ACKNOWLEDGEMENTS The preparation of this chapter was supported in part by Grant-in-Aid for Scienti®c Research on Priority Areas given to the ®rst author, The growth of the mind: Mechanisms of conceptual development (09207105). We thank Willard Hartup and Henry Wellman for their valuable comments and suggestions for revision on an earlier version of the chapter. REFERENCES Atran, S. (1994). Core domains versus scienti®c theories: Evidence from systematics and Itza-Maya folkbiology. In L.A. Hirschfeld & S.A. Gelman (Eds.), Mapping the mind: Domain speci®city in cognition and culture (pp. 316±340). New York: Cambridge University Press.

140

Hatano and Inagaki

Atran, S. (1998). Folkbiology and the anthropology of science: Cognitive universals and cultural particulars. Behavioral and Brain Sciences, 21, 547±609. Baillargeon, R., Kotovsky, L., & Needham, A. (1995). The acquisition of physical knowledge in infancy. In D. Sperber, D. Premack, & A.J. Premack (Eds.), Causal cognition: A multidisciplinary debate (pp. 79±116). New York: Oxford University Press. Barkow, J.H., Cosmides, L., & Tooby, J. (1992). The adapted mind: Evolutionary psychology and the generation of culture. New York: Oxford University Press. Carey, S. (1985). Conceptual change in childhood. Cambridge, MA: MIT Press. Carey, S. (1995). On the origin of causal understanding. In D. Sperber, D. Premack, & A.J. Premack (Eds.), Causal cognition: A multidisciplinary debate (pp. 268± 302). New York: Oxford University Press. Carey, S., & Spelke, E. (1994). Domain-speci®c knowledge and conceptual change. In L.A. Hirschfeld & S.A. Gelman (Eds.), Mapping the mind: Domain speci®city in cognition and culture (pp. 169±200). New York: Cambridge University Press. Chi, M.T.H., Glaser, R., & Rees, E. (1982). Expertise in problem solving. In R.J. Sternberg (Ed.), Advances in the psychology of human intelligence, Vol. 1. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Chi, M.T.H., Hutchinson, J.E., & Robin, A.F. (1989). How inferences about novel domain-related concepts can be constrained by structured knowledge. MerrillPalmer Quarterly, 35, 27±62. Chomsky, N. (1988). Language and problems of knowledge. Cambridge, MA: MIT Press. Cole, M., Gay, J., Glick, J., & Sharp, D.W. (1971). The cultural context of learning and thinking. New York: Basic Books. Dunn, J., Brown, J., Slomkowski, C., Tesla, C., & Youngblade, L. (1991). Young children's understanding of other people's feelings and beliefs: Individual differences and their antecedents. Child Development, 62, 1352±1366. Elman, J.L., Bates, E.A., Johnson, M.H., Karmiloff-Smith, A., Parisi, D., & Plunkett, K. (1996). Rethinking innateness: A connectionist perspective on development. Cambridge, MA: MIT Press. Gallistel, C.R., Brown, A.L., Carey, S., Gelman, R., & Keil, F. C. (1991). Lessons from animal learning for the study of cognitive development. In S. Carey & R. Gelman (Eds.), The epigenesis of mind: Essays on biology and cognition (pp. 3± 36). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Garcia, J., & Koelling, R.A. (1966). The relation of cue to consequence in avoidance learning. Psychonomic Science, 4, 123±124. Gelman, R. (1990) First principles organize attention to and learning about relevant data: Number and the animate-inanimate distinction as examples. Cognitive Science, 14, 79±106. Gelman, R., & Baillargeon, R. (1983). A review of some Piagetian concepts. In J.H. Flavell & E.M. Markman (Eds.), Handbook of child psychology: Vol. 3. Cognitive development (pp. 167±230). New York: John Wiley. Golinkoff, R.M., Harding, C.G., Carlson, V., & Saxton, M.E. (1984). The infant's perception of causal events: The distinction between animate and inanimate objects. In L.L. Lipsitt & C. Rovee-Collier (Eds.), Advances in infancy research, Vol. 3 (pp. 145±165). Norwood, NJ: Ablex. Hatano, G., & Inagaki, K. (1994). Young children's naive theory of biology. Cognition, 50, 171±188.

7. Conceptual development

141

Hatano, G., & Inagaki, K. (2000). Knowledge acquisition and use in higher-order cognition. In K. Pawlik & M.R. Rosenzweig (Eds.), International handbook of psychology (pp. 167±190). London: Sage. Hatano, G., Siegler, R.S., Richards, D.D., Inagaki, K., Stavy, R., & Wax, N. (1993). The development of biological knowledge: A multi-national study. Cognitive Development, 8, 47±62. Hirschfeld, L.A., & Gelman, S.A. (1994). Toward a typography of mind: An introduction to domain speci®city. In L.A. Hirsch®eld & S.A. Gelman (Eds.), Mapping the mind: Domain speci®city in cognition and culture (pp. 3±35). New York: Cambridge University Press. Inagaki, K. (1990). The effects of raising animals on children's biological knowledge. British Journal of Developmental Psychology, 8, 119±129. Inagaki, K., & Hatano, G. (1987). Young children's spontaneous personi®cation as analogy. Child Development, 58, 1013±1020. Inagaki, K., & Hatano, G. (2002). Young children's naive thinking about the biological world. New York: Psychology Press. Karmiloff-Smith, A. (1992). Beyond modularity: A developmental perspective on cognitive science. Cambridge, MA: MIT Press. Karmiloff-Smith, A. (1996). The connectionist infants: Would Piaget turn in his grave? SRCD Newsletter, 1±3 and 10. Keil, F.C. (1981). Constraints on knowledge and cognitive development. Psychological Review, 88, 197±227. Keil, F.C. (1990). Constraints on constraints: Surveying the epigenetic landscape. Cognitive Science, 14, 135±168. Keil, F.C. (1998). Cognitive science and the origins of thought and knowledge. In R.M. Lerner (Ed.), Handbook of child psychology (5th ed.): Vol. 1. Theoretical models of human development (pp. 341±413). New York: John Wiley. Laboratory of Comparative Human Cognition (LCHC). (1983). Culture and cognitive development. In W. Kessen (Ed.), Handbook of child psychology (4th ed.), Vol. 1, History, theory, and method (pp. 295±356). New York: John Wiley. Massey, C.M., & Gelman, R. (1988). Preschooler's ability to decide whether a photographed unfamiliar object can move itself. Developmental Psychology, 24, 307±317. Munakata, Y., McClelland, J.L., Johnson, M.J., & Siegler, R.S. (1997). Rethinking infant knowledge: Toward an adaptive process account of successes and failures in object permanence tasks. Psychological Review, 104, 686±713. Newell, A., & Simon, H.A. (1972). Human problem solving. Englewood Cliffs, NJ: Prentice Hall. Peterson, C.C., & Siegal, M. (1997). Domain speci®city and everyday biological, physical, and psychological thinking in normal, autistic, and deaf children. In H.M. Wellman & K. Inagaki (Eds.), The emergence of core domains of thought: Children's reasoning about physical, psychological, and biological phenomena (pp. 55±70). San Francisco: Jossey-Bass. Piaget, J. (1956). The general problems of the psychobiological development of the child. In J.M. Tanner & B. Inhelder (Eds.), Discussions on child development, Vol. 4 (pp. 3±27). London: Tavistock Publications. Resnick, L.B. (1987). Constructing knowledge in school. In L.S. Liben (Ed.), Development and learning: Con¯ict or congruence? (pp. 19±50). Hillsdale, NJ: Lawrence Erlbaum Associates Inc.

142

Hatano and Inagaki

Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. New York: Oxford University Press. Rogoff, B. (1998). Cognition as a collaborative process. In D. Kuhn & R.S. Siegler (Eds.), Handbook of child psychology (5th ed.): Vol. 2. Cognition, perception, and language (pp. 679±744). New York: John Wiley. Siegler, R.S. (Ed.). (1978). Children's thinking: What develops? Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Spelke, E.S. (1994). Initial knowledge: Six suggestions. Cognition, 50, 431±445. Spelke, E.S., & Newport, E.L. (1998). Nativism, empiricism and the development of knowledge. In R.M. Lerner (Ed.), Handbook of child psychology (5th ed.): Vol. 1. Theoretical models of human development (pp. 275±340). New York: John Wiley. Stavy, R., & Wax, N. (1989). Children's conceptions of plants as living things. Human Development, 32, 88±94. Tomasello, M. (1992). First verbs: A case study in early grammatical development. Cambridge: Cambridge University Press. Tomasello, M., Kruger, A.C., & Ratner, H.H. (1993). Cultural learning. Behavioral and Brain Sciences, 16, 495±552. Wellman, H.M., & Gelman, S.A. (1998). Knowledge acquisition in foundational domains. In D. Kuhn & R.S. Siegler (Eds.), Handbook of child psychology (5th ed.): Vol. 2. Cognition, perception, and language (pp. 523±573). New York: John Wiley. Wellman, H.M., & Inagaki, K. (Eds.). (1997). The emergence of core domains of thought: Children's reasoning about physical, psychological, and biological phenomena. San Francisco: Jossey-Bass. Wiser, M., & Carey, S. (1983). When heat and temperature were one. In D. Gentner & A.L. Stevens (Eds.), Mental models (pp. 53±73). Hillsdale, NJ: Lawrence Erlbaum Associates Inc.

8

Research on memory development: Past and present Wolfgang Schneider

University of WuÈrzburg, Germany

The main purpose of this chapter is to provide a comprehensive account of historical trends in the ®eld of memory development, covering more than 120 years of research on this issue conducted mainly in Europe and the United States. The chapter begins with a review of historical trends that summarises major ®ndings from 1880 to 1965 and also includes an overview of current research trends and recent developments, followed by some speculations about the future of the ®eld. HISTORY Experimental studies of memory are as old as scienti®c psychology. When Ebbinghaus (1885) started with his classic experiments on memory and forgetting in 1879, Wundt founded the ®rst psychological laboratory. Although this is widely known, it is not equally well known that research on memory development also started at about that time. Numerous studies were carried out in Europe around the turn of the century that investigated developmental and individual differences in children's memory. There were three rather independent lines of research that contributed to this early trend. First, whereas the development of children did not attract much interest before the end of the nineteenth century, carefully conducted case studies of young children's development (which also included systematic observations of memory development in early childhood) received a lot of attention. A second line of research was directly derived from memory experiments with adults. Some of these studies explored whether ®ndings obtained for adult populations could be generalised to children of different ages. Other investigations were less basic in nature and were driven by educational interests. The third line of research on children's memory was even more applied, focusing on children and adults' testimonial competence. The major ®ndings of early research on memory development are summarised. In the following pages, three different time periods are distinguished, which clearly differed regarding the focus of research.

144

Schneider

1880±1935 Evidence from child psychology studies Wilhelm Preyer's (1882) careful observations of his son during the ®rst 3 years of life represented the beginning of scienti®c child psychology. Preyer voted for an explicit, systematic, and methodologically controlled observational approach, which allows for the construction of an age-graded behavioural inventory. His own inventory included numerous cognitive features such as sensorimotor skills, perception, speech acquisition, learning, memory, and the development of conscious thought. Given that many others adopted this observational approach, a rich database on the course of (early) cognitive development was available around the turn of the century. This new knowledge base was represented in several important textbooks written by leading German child psychologists of that time (e.g., BuÈhler, 1918; Koffka, 1921; Stern, 1914; Werner, 1926), which all contained chapters on children's memory development. The emphasis was on long-term memory, recognition, and imagery. Statements about short-term memory were rare and not very speci®c. All of the textbook authors agreed that imagination is the result of memory processes, conceivable as replicable memory trace that is independent of actual perception but can be in¯uenced by knowledge structures. Most textbook writers also made a distinction between verbal and visual longterm memory. Regarding developmental trends, all of the authors agreed that ``stimulusbound'' recognition memory can be observed in young infants, indicating that this skill develops very early in life (but not before the age of 4 months). Its main function was seen in the categorisation of input stimuli and the reduction of input complexity. Recognition memory was assumed to improve considerably during the second year of life, accompanied by a dramatic increase in retention span, which could last for several months (cf. BuÈhler, 1930). In contrast, ``free'' memory activities such as recall were assumed to develop considerably later, starting by about 4 years of age. Most authors also agreed that visual memory functions were less de®cient in young children than verbal ones. Accordingly, they assumed that developmental changes were more pronounced for verbal than for visual memory. Several studies investigated what would be called ``infantile amnesia'' today (see Offner, 1924). That is, researchers were interested in establishing the age at which memories of speci®c events can ®rst be permanently stored, a question that was examined using samples of older children and adults. Early research by the French scientists Henri and Henri (1897) suggested that memories could date back to the second year of life where the events in question were particularly dramatic. However, subsequent studies did not con®rm such an early onset, indicating that the earliest events that participants could remember occurred when they were 3 or 4 years old. This ®nding from retrospective analyses squared well with the

8. Research on memory development

145

observation that children who had moved away from their birthplace before the age of 4 could not remember any details about their original homes (BuÈhler, 1930; Offner, 1924). Findings from early experimental studies In experimental work, the major goal was to describe changes in memory performance across the lifespan. Children were treated as a special population in order to evaluate the generalisability of ®ndings from studies with adults. Large samples of children from different ages were included in most of these studies. In contrast to the child psychology studies summarised in the previous subsection, the majority of these experimental studies focused on short-term (immediate) memory. Ebbinghaus (1885, 1897) himself was concerned with memory span capacity at different age levels, with particular interest in identifying the developmental memory-span curves for various types of materials. His ®ndings, based on children and young adults, indicated that age effects could be observed across various materials, even though age differences were least pronounced for meaningless words and syllables. Although there was broad agreement in subsequent studies that children's immediate memory is less developed than that of adults, researchers disagreed regarding the course of memory development from childhood to adulthood. BuÈhler (1930) assumed that developmental gains are most pronounced in 10±12-year-old children and then again after puberty. Experimental studies by Bourdon (1894) and Chamberlain (1915) suggested considerable increases in memory capacity during the late elementary school years, with only minor further improvements until early adulthood (see also Nagy, 1930). On the other hand, Meumann (1907) reported ®ndings from his own studies with children and adults that showed rather slow improvements of immediate memory until the age of 13, followed by rapid improvements between the ages of 13 and 16. Peak performance was reached by the age of 25, followed by a period of stagnation and stability until the age of 46 (i.e., the oldest participants in his study). Despite these differences in view, researchers agreed that memory development was characterised by nonlinear trends, and that developmental patterns varied as a function of material and memory task. Although studies on short-term memory dominated the ®eld, there were also investigations of long-term memory and forgetting. These developmental studies revealed that learning and forgetting rates differed as a function of age. Whereas it took children considerably longer than adults to learn the stimulus lists, they did not forget at faster rates. The fact that Ebbinghaus's (1885) forgetting curves could not be replicated was assumed to be largely due to the relatively rapid rate of self-presentation in that study, and the fact that Ebbinghaus was subject and experimenter at the same time. An alternative explanation is that Ebbinghaus forgot so much so quickly because he experienced a high amount of interference from

146

Schneider

the many lists he had learned previously, that is, proactive interference (Underwood, 1957). Children's eyewitness memory Although there was little interest in children's testimonial competence in the United States at the start of the twentieth century, systematic research on adults' and children's testimony ¯ourished in Europe, particularly in Germany and France (for a detailed review, see Ceci & Bruck, 1993). One of the earliest studies on children's testimony was carried out by Binet (1900). Binet assumed that suggestibility re¯ected the impact of two classes of factors. The ®rst class was labelled ``autosuggestion'' and referred to the in¯uence of a prominent thought that develops within the individual without obvious external cause. Autosuggestion was supposed to paralyze the critical senses and memories of past events. The second class of factors was assumed to be external to the individual, re¯ecting mental obedience to signi®cant others, particularly adults. As emphasised by Ceci and Bruck (1993), Binet's major contribution to the ®eld was (1) that he distinguished between errors of reporting caused by actual memory changes versus those caused by social conformity, and (2) that his early research is of high importance for the current debate as to whether the original memory trace is itself impaired or simply allowed to ``coexist'' with traces of suggestion. In Germany, early research on eyewitness testimony was stimulated by Clara and William Stern's observations of their own children (Stern, 1910, 1914; Stern & Stern, 1913). They distinguished ``memory deceptions'' from intentional lies, de®ning such deceptions as objectively incorrect statements that were based on subjective con®dence. Problems of young children correctly recalling the circumstances and contents of previous experiences were attributed to their weakly developed understanding of time. Preschool and kindergarten children not only lack the ability to coordinate experienced events with speci®c time markers, but they are also not interested in such relations. According to Stern (1914), the major ®nding of research on young children's testimony is that spontaneous reports give much more correct information than speci®c questions, even though children do not offer that much spontaneously. Stern also warned about repeated questioning of the same event, claiming that a person's answers during the ®rst memory assessment may be better remembered than the actual events (Stern, 1910). Also, his research indicated that although younger children are most suggestible to misleading information, even adults are misled by suggestive questions. Assessment of ``general '' memory development Given the broad disagreement regarding the general course of memory development described here, Brunswik, Goldscheider and Pilek (1932)

8. Research on memory development

147

conducted a developmental study that aimed at providing a general description of short-term and long-term memory in school-age children and adolescents. This study also differed from earlier investigations in that the issues addressed were directly derived from developmental theory, that is, Charlotte and Karl BuÈhler's doctrine of phases and stages (e.g., BuÈhler, 1930). A large variety of memory tasks were presented to a sample of about 700 participants, ranging from 6 to 18 years of age. Tests involved short- and long-term memory for nonsense and meaningful words, colours, and numbers, as well as memory for poems. Moreover, several nonverbal memory tasks such as memory for motor actions and their correct sequence were included. The disparate growth curves obtained for different memory functions were consistent with the data in previous studies (e.g., Netschajeff, 1900, 1902; Offner, 1924). The attempt to construct a curve of the general development of immediate memory (``memory strength'') is particularly interesting. The curve was based on scores of all participants and represented an aggregation across all measures included in the study (see Figure 2.1 in Schneider & Pressley, 1997, for a reconstruction of the graphical representation). The outcomes were compatible with ®ndings from other early studies in that linear and steep rises in memory performance were found from 6 to 11 years of age, and that there was a plateau in performance during pre- and early adolescence. The authors claimed to have found support for BuÈhler's position that memory development during the early years is dominated by roteassociative processes (``mechanical learning''), whereas learning and memory predominant in older children and adults is based on the creation of meaning (``logical memory''). A closer inspection of ®ndings revealed that the theoretical position was not entirely supported by the data: In fact, younger children required more practice to learn nonsense syllables than meaningful words, with continuous improvement in the learning of nonsense syllables up until age 18. Nonetheless, subsequent reports on the study of Brunswik and co-workers emphasised the qualitative shift from mechanical to logical memory (for instance, see Werner, 1953).

1936±1965 German research The great progress made by German researchers in the early twentieth century came to a halt as war exploded over Europe. Many leading child psychologists such as the BuÈhlers, Koffka, the Sterns, and Heinz Werner left the country and started new careers in the United States. Theoretical perspectives that predominated in the immediate postwar period (i.e., Gestaltist, phenomenological) did not encourage analytical research on memory development. As a consequence, no empirical studies on memory development were published in Germany between 1933 and 1961.

148

Schneider

American research The situation was very different in America. Behaviourism dominated the ®eld, and theories of verbal learning were very popular. Most of the verbal learning studies with children were descriptive, with studies often conducted at a single age level. However, there were a few notable exceptions. For instance, Koppenaal, Krull, and Katz (1964) used a paired-associate learning paradigm to test the developmental hypothesis that there should be greater interference with older than with younger children for highly associative learning materials. This assumption was based on the observation that verbal associate strength is determined by children's prior knowledge, and that prior knowledge is richer for older compared to younger children. The results were consistent with the developmental hypothesis in that third-graders showed signi®cant retroactive and proactive inhibition, which was not true for the younger children. Another important developmental study focusing on free recall was carried out by Bous®eld, Esterson, and Whitmarsh (1958). These authors compared the degree of associative clustering in free recall of third-grade, fourth-grade, and college students. The theoretical rationale was derived from Werner's assumption that organisational principles change during development, in that perceptual-sensory categorisation is initially dominant but is replaced later by a tendency to organise stimuli conceptually. Both recall and conceptual clustering increased as a function of age. Contrary to expectations, colour clustering was low in all age groups. Although the ®ndings concerning colour clustering were not in accord with Werner's theory, this did not necessarily contradict the theory, given that the age range was restricted and that younger children (e.g., pre-schoolers and kindergarten children) were not included in the sample. Russian research Russian developmental researchers were particularly interested in the development of ``logical'' (meaningful) memory compared to rote ``mechanical'' learning. They emphasised the relevance of conscious, independent, and goal-oriented memory activity as the causal origin of memory development (see reviews by Meacham, 1977; Smirnov & Zinchenko, 1969). Two major concerns can be distinguished in the empirical studies. One was the study of involuntary memory, which occurs when the ultimate goal of the person is not memory but something else, often comprehension in the Russian studies. In such cases, memory is involuntary in that it is a byproduct of comprehension. The second focus was voluntary memory, which was de®ned as a product of activities that are driven by the goal to remember. Most Russian researchers believed that superior (higher-order) forms of memory develop on the basis of the transition from natural, involuntary memorising to subject-controlled and voluntary memory involving the use of mediating processes and cues.

8. Research on memory development

149

One typical feature of the Russian studies on voluntary memory was the focus on learning from text. For example, Korman (1944, 1945; cited in Yendovitskaya, 1971) studied pre-school children's memory for connected material (fairy tales). Korman was impressed by the children's ability to recall the main events of the story correctly. Recall differences between 4±5year-olds and 6-year-olds were primarily quantitative. Subsequent research by Smirnov (1948; cited in Smirnov, 1973) with schoolchildren focused on the impact of text structure on children's recall. Participants (second-, fourth-, and sixth-graders) were presented two texts, with sentences organised in a coherent fashion or randomly. After two attempts at recalling the two texts, children were instructed to break up the texts into pieces of information that went together (done from memory). The most important ®nding was that the youngest children were not able to reorganise the random text, and that only a small percentage of the older participants were able to do so. Interestingly, although most of the children could identify that random texts were harder to break up than coherent texts, they were not consciously aware of the structural differences between the texts. Smirnov concluded that meaningful grouping processes ®rst proceed in an unconscious way before becoming conscious activities. In sum, it seems fair to conclude that most of the research on memory development conducted between 1936 and 1965 took place in Russia. In particular, the outcome of this research improved our knowledge on pre-schoolers' memory activities, foreshadowing later investigations on children's implicit and explicit memory. Other major accomplishments of this period include the analysis of developmental differences in children's text processing and recall, and the study of processes that in¯uence encoding and storage of verbal materials, such as repetition and selfchecking. Whereas the repertoire of memorising methods seemed rather restricted in kindergarteners and young school children, the number and quality of mediating activities considerably increased between the secondand ®fth-grade levels. In accord with earlier ®ndings (e.g., Brunswik et al., 1932), memory development appeared to be more dramatic for the elementary school years than for later periods. THE MODERN ERA How does modern research on memory development differ from the historical approaches? One of the crucial differences involves a shift from describing developmental differences in memory to identifying the underlying mechanisms of change. Another difference concerns the theoretical framework used. Since the mid-1960s, research on memory development has been in¯uenced strongly by theoretical models derived from informationprocessing and neuroscience approaches (see the reviews by Kail, 1990; Schneider & Bjorklund, 1998; Schneider & Pressley, 1997). Developmental

150

Schneider

psychologists began looking at changes in children's thinking in terms of a computer metaphor. From this perspective, memory development can be seen as re¯ecting either hardware (the capacity of memory systems and the speed with which information can be processed) or software (e.g., use of strategies). Developmental research was strongly in¯uenced by multistore memory models that distinguished between a sensory register, a short-term store, and a long-term store. Regarding the content of the long-term store, it was proposed that information can be represented in two ways. Explicit or declarative memory refers to our capacity for conscious recollection of names, places, dates, or events and comes in two types, episodic and semantic. Whereas episodic memory concerns events and experiences that can be consciously retrieved, semantic memory refers to our knowledge of language, rules, and concepts. In contrast, implicit or procedural memory represents a variety of nonconscious abilities, including the capacity for learning habits and skills, and some forms of classical conditioning. In the following sections, major accomplishments regarding memory changes in infants and the very young are presented ®rst. Next, major determinants of memory development in older children and adolescents are described.

Infancy and early childhood What do we know about memory development in very young children? Recent ®ndings demonstrate that from birth on, infants can remember things (e.g., faces, pictures, objects) for rather long periods of time. This is apparent from new-borns' patterns of habituation and dishabituation. Basic memory activities such as the association of a stimulus with a response and the distinction between old and new stimuli (i.e., recognition processes) are especially dominant early in life. Although experiments with new-borns indicate that neonates can form at least a brief memory of perceptual experiences, other research with very young infants indicates that their recognition of objects is surprisingly durable. For instance, 2month-olds continue to prefer to look at other forms that they have not seen previously even 2 weeks after they were habituated to a speci®c form. Also, recent research using the preference-for-novelty technique has shown that infants as young as 3 months can retain a memory for object motions across intervals of 1 and 3 months (Bahrick & Pickens, 1995). Moreover, young infants can also remember activities that they had performed at an earlier point in time. Research focusing on older infants (between 10 and 20 months of age) has used deferred (delayed) imitation techniques to measure memory boundaries. In several of these experiments, infants watched as an experimenter demonstrated some novel, unusual behaviour with an unfamiliar toy. Infants later imitated such strange behaviours, which indicates that these events had been stored in long-term memory. Although levels of imitation were typically greater for older than

8. Research on memory development

151

for younger infants, the ®ndings demonstrate that the neurological systems underlying long-term recall are present at the beginning of the second year of life (Meltzoff, 1995). How does memory for speci®c events develop after the ®rst 2 years of life? What are important determinants of long-term memory? Interestingly, chronological age does not seem to be the primary determinant of whether or for how long an event will be remembered once the capacity is in place. Rather, the literature shows that the organisation of events and the availability of cues or reminders determine young children's long-term memory (see the review by Bauer, 1997). Young children tend to organise events in terms of ``scripts'', which are a form of schematic organisation with realworld events structured in terms of their causal and temporal characteristics. Scripts develop most routinely for common, repeated events. Children learn what ``usually happens'' in a situation, for instance, a birthday party or a visit to a restaurant. Memory for routine events enables infants and toddlers to anticipate events and to take part in, and possibly take control of, these events (Nelson, 1996). Repeated experience clearly facilitates long-term recall in preverbal and early-verbal children. Social context and memory development What are the important mechanisms of young children's (verbal) memory development? Why is it that our childhood memories are so poor? Although research on ``infantile amnesia'' has repeatedly shown that adults almost never remember anything that happened to them in their early years, this does not imply that children younger than 2 or 3 years of age are not capable of remembering speci®c events from the past. Whereas there is little doubt that episodic memory develops very early in life, the speci®c form of episodic memory known as autobiographical (personal) memory may not have been established at the same time. Recent research illustrates that the emergence of autobiographical memory is linked to the acquisition of narrative form to use as a frame for structuring and remembering experience (see Nelson, 1996). Thus parents play an important role. Interchanges between parents and their young children seem highly relevant for children's recall pro®ciency. Children learn to remember by interacting with their parents, jointly carrying out activities that are later performed by the child alone. Parents begin talking with their young children about things that happened in the past. In particular, parents' repeated questioning helps in understanding how remembering is done.

Memory development between 5 and 15 The vast majority of studies on memory development since the mid-1960s has been been carried out with older children, mainly dealing with explicit memory, that is, conscious remembering of facts and events. Particularly

152

Schneider

clear improvements in declarative memory can be observed for the age range between 6 and 12 years, which roughly corresponds to the elementary school period in most countries. In order to explain these rapid increases in memory performance over time, different sources or determinants of memory development have been identi®ed. According to most contemporary memory researchers, changes in basic capacities, memory strategies, metacognitive knowledge, and domain knowledge all contribute to developmental changes in memory performance. There is also broad agreement that some of these sources of development contribute more than others, and that some play an important role in certain periods of childhood but not in others (see Schneider & Pressley, 1997; Siegler, 1998). The role of basic capacities In an extensive review of the literature, Dempster (1981) reported that the memory span of 2-year-olds is about two items; of 5-year-olds about four items; of 7-year-olds about ®ve items; and of 9-year-olds about six items. The average memory span of adults is about seven items. Although these ®ndings indicate continuous increases in memory capacity, critics of the approach refer to the fact that individual differences in knowledge about the items and age-related differences in strategy use were not suf®ciently controlled for. Thus these ®ndings were still compatible with the model of memory capacity developed by Case (1985) that assumed an age-invariant total processing store but developmental changes in its two components (executive processing and storage). Another, more recent approach was based on Baddeley's (1986) concept of working memory. Tests of working memory are similar to memory span tasks in that participants must remember a series of items in exact order, but these are embedded in an additional task in which participants must transform information held in the short-term store. In general, children's performance on working memory tasks shows the same age-related increase in their performance as their performance on memory span tasks, although the absolute level is somewhat reduced in working memory tasks. Working memory models emphasise the fact that information processing is restricted by capacity limits and time limits. That is, the amount of information that one can bring to mind all at once is quite limited, and this limited amount of information that is brought to mind at any one time can be kept in mind only brie¯y unless the person invests enormous effort in order to retain the information. Accordingly, models of working memory place substantial emphasis on speed of processing. For instance, the working memory model developed by Baddeley, Gathercole, Hitch, Hulme, and their colleagues assumes that age differences in verbal memory span are primarily the result of age differences in a phonological subsystem called the articulatory loop. Verbal information is stored in the articulatory loop, and these traces decay rapidly if verbal rehearsal is not performed.

8. Research on memory development

153

Although some researchers have reported age differences in the rate at which information decays (see the review by Cowan, 1997), most researchers propose that age differences in rehearsal rate are primarily responsible for developmental differences in memory span. Overall, there is converging evidence that developmental changes in memory capacity are due to signi®cant increases in information processing speed, which are most obvious in early ages, with the rate of change slowing thereafter. Age differences in speed of processing in¯uence a wide range of tasks and may represent a cognitive primitive (see Kail, 1991, 1993; Kail & Salthouse, 1994). However, considerable controversy has arisen over whether the increased speed is due to greater use of strategies, to greater familiarity with the items used, or to speed per se. Recent evidence suggests that age differences in speed of processing are primarily in¯uenced by maturational factors. For instance, Cowan and colleagues (Cowan, Nugent, Elliott, Ponomarev, & Saults, 1999) evaluated age differences in the span of apprehension, which refers to the amount of information that people can attend to at a single time or the number of items that people can keep in mind at any one time. Span of apprehension be can assessed only when factors such as focused attention, knowledge of the target items, and encoding strategies can be eliminated, factors that are not eliminated on memory span or working memory span tasks. Cowan and his colleagues reported that the average span of apprehension increased signi®cantly with age (2.41, 3.13, and 3.56 digits for the 7-year-old, 10-year-old, and adult participants, respectively), and interpreted these differences as re¯ecting a true developmental difference in the capacity of the short-term store. Effects of memory strategies Memory strategies have been de®ned as mental or behavioural activities that achieve cognitive purposes and are effort-consuming, potentially conscious, and controllable (Flavell, Miller, & Miller, 1993). Since the mid1960s numerous studies have investigated the role of strategies in memory development. Strategic memory was at the centre of early investigations in this area (see Harnishfeger & Bjorklund, 1990, for an historical sketch) and continues to be so today, although the topic no longer dominates the ®eld as it once did. Research interest in strategies such as rehearsal and semantic organisation was partially motivated by the crucial role of such control processes in general multistore memory models, and by John Flavell's work establishing rehearsal and organisation as strategies that develop between 5 and 10 years of age. Strategies can be executed either at the time of learning (encoding) or later on when information is accessed in long-term memory (retrieval). The example of organisational strategies may illustrate the case. Early research by Flavell and colleagues as well as subsequent studies con®rmed a developmental trend which paralled that of rehearsal strategies (for reviews, see

154

Schneider

Kail, 1990; Schneider & Pressley, 1997). Typically, organisational strategies were not observed in children younger than 5 or 6 years of age. The lack of strategic behaviours in very young children was labelled mediation de®ciency, indicating that children of a particular (preschool) age do not bene®t from memory strategies, even after having been instructed how to use them. Kindergarten children and young schoolchildren show a production de®ciency. That is, they fail to use organisational strategies when given ``neutral'' instructions but can be easily trained to do so, usually with corresponding improvements in memory performance. Much research in strategy development has concerned the factors responsible for production de®ciencies, the subsequent failure to transfer an acquired strategy to a new situation, and ways in which to improve children's strategy effectiveness (e.g., Pressley & Van Meter, 1993; Schneider & Pressley, 1997). It could be demonstrated that insuf®cient mental capacity and lack of domain-speci®c knowledge were partially responsible for production de®ciencies in young children. Numerous studies showed that memory strategies develop most rapidly over the elementary school years. Older children are more likely to group items on the basis of meaning and to study same-category items together, with higher levels of sorting and clustering yielding higher levels of recall (e.g., Hasselhorn, 1992). However, the ages of strategy acquisition are relative, and variable within and between strategies. Even preschoolers and kindergarten children are able to use intentional strategies, both in ecologically valid settings, such as hide-and-seek tasks, and in the context of a laboratory task. More recently, research on strategy development has focused on two interesting phenomena. First, memory researchers have observed that higher levels of organisation are not always accompanied by higher levels of recall. Such utilisation de®ciencies have been shown in several studies with both preschoolers and elementary school children (for reviews, see Bjorklund & Douglas, 1997; Miller & Seier, 1994). It appears from this research that using a strategy such as organisation is only a ®rst step. Once the mechanics of the strategy are learned, children need more time before they can execute it pro®ciently. Although there is plenty of evidence supporting the utilisation de®ciency paradigm, it is important to note that ®ndings are not always consistent and are dependent on task characteristics (Schneider & Sodian, 1997; Waters, 2000). A second new line of research based on longitudinal data has shown that our understanding of strategy development based on cross-sectional studies and the analysis of group data does not tell us the whole story about individual developmental trends. Schneider and Sodian (1997; Sodian & Schneider, 1999) explored the issue whether the impression of gradual developmental increases in strategy use and recall derived from crosssectional studies could be con®rmed by the analysis of longitudinal data. Although they con®rmed earlier results on the group level, they were struck

8. Research on memory development

155

by low individual stabilities for the strategy and recall variables over time. If individual children change their relative positions in the sample considerably between measurement points (as indicated by the low stabilities), the model of gradual improvement does not seem to hold. In fact, Schneider and Sodian found that gradual, steady increases were rarely observed. About 80% of the children ``jumped'' from chance level to near-perfection between subsequent measurement points. About 8% were perfect from the start, and only 8% showed a pattern of steady increase as suggested by the group data. Taken together, research conducted during the last three decades has convincingly shown that age-related problems in the frequency of use and quality of children's strategies play a large role in memory development between the early school years and adolescence. However, there is now an increasing realisation that the use of encoding and retrieval strategies largely depends on children's strategic as well as nonstrategic knowledge. There is broad consensus that the narrow focus on developmental changes in strategy use should be replaced by an approach that takes into account the effects of various forms of knowledge on strategy execution. The impact of metacognitive knowledge Almost 30 years ago, John Flavell (1971) introduced the term metamemory to refer to knowledge about memory processes and contents (see for a similar idea Kuhlmann, 1908). Two broad categories of metacognitive knowledge have been distinguished in the literature (Flavell & Wellman, 1977). Declarative metacognitive knowledge refers to what children factually know about their memory. This type of knowledge is explicit and verbalisable and includes knowledge about the importance of person variables (e.g., age or IQ), task characteristics such as task dif®culty, or strategy knowledge. In comparison, procedural metacognitive knowledge is mostly implicit (subconscious) and refers to children's self-monitoring and selfregulation activities while solving memory problems. Empirical research exploring the development of declarative metamemory revealed that children's knowledge of facts about memory increases considerably over the primary-grade years, but is still incomplete by the end of childhood. Recent studies also showed that increases in knowledge about strategies are paralleled by the acquisition of strategies, and that metamemory±memory behaviour relationships tend to be moderately strong (cf. reviews by Schneider, 1999; Schneider & Pressley, 1997). Thus what children know about their memory obviously in¯uences how they try to remember. However, although late-grade-school children know much about strategies, there is increasing evidence that many adolescents and adults (including college students) have little or no knowledge of some important and powerful memory strategies. The situation regarding developmental trends in procedural metacognitive knowledge is not entirely clear. There is evidence that older children are

156

Schneider

better able to predict future performance on memory tasks than younger children, and that there are similar age trends when the task is to judge performance accuracy after the test has been taken. Also, older children seem better able to judge whether the name of an object that they currently cannot recall would be recognised later if the experimenter provided it (feeling-of-knowing judgements). However, although monitoring skills seem to improve continuously across childhood and adolescence, it seems important to note that developmental trends in self-monitoring are less pronounced than those observed for declarative metamemory. Taken together, research on the role of metamemory in memory development has created a large body of evidence supporting the utility of the concept. Mainly due to methodological improvements, more recent work on the metamemory±memory link has provided evidence for rather strong relations among metamemory, memory behaviour, and memory performance (see Schneider, 1999). Effects of domain knowledge Surprisingly, one of the most obvious sources of individual differences in memory performance, prior knowledge of task-related content, was discovered rather late. Since the late 1970s, however, a large number of developmental studies have demonstrated that the amount of knowledge in a particular domain such as chess, physics, or sports determines how much new information from the same domain can be stored and retrieved (see the reviews by Bjorklund, 1985; Bjorklund & Schneider, 1996; Chi & Ceci, 1987; Schneider & Bjorklund, 1998). Prior knowledge of related content affects memory in several ways. It not only in¯uences how much and what children recall, but also affects their execution of basic processes and strategies, their metacognitive knowledge, and their acquisition of new strategies. Rich domain knowledge can also have nonstrategic effects, that is, diminish the need for strategy activation. Interestingly, domain knowledge can serve as an explanation for other memory changes. Increasing domain knowledge improves ef®ciency of basic processes, acquisition and execution of strategies, and metacognitive knowledge. Evidence for the powerful effects of domain knowledge on memory performance comes from studies using the expert±novice paradigm. These studies compared experts and novices in a given domain (e.g., baseball, chess, or soccer) on a memory task related to that domain. It was demonstrated that rich domain knowledge enabled a child expert to perform much like an adult expert and better than an adult noviceÐthus showing a disappearance and sometimes reversal of usual developmental trends. Experts and novices not only differed with regard to quantity of knowledge but also regarding the quality of knowledge, that is, in the way their knowledge is represented in the mind. Moreover, several studies also

8. Research on memory development

157

con®rmed the assumption that rich domain knowledge can compensate for low overall aptitude on domain-related memory tasks, as no differences were found between high- and low-aptitude experts on various recall and comprehension measures (Bjorklund & Schneider, 1996).

New research trends New lines of basic research on memory development have focused on nonstrategic factors, not only reactivating old issues such as unintentional/ implicit memory or long-term retention and forgetting, but also developing conceptual frameworks such as fuzzy-trace theory that question the importance of higher-order cognitive processes for memory development and thus challenge the strategic memory view. Another new exciting line of research that has received substantial attention during the last 15 years and also represents the revival of a classic theme concerns autobiographical memory, in particular, children's eyewitness memory. In the following, the main outcomes of these new research trends will be brie¯y reviewed. Unintentional and implicit memory By de®nition, implicit memory is ``memory without awareness''. Research on implicit memory with children involved the use of fragmented pictures, perhaps a dog, that should be identi®ed by the participants. This is very dif®cult to do initially, but as more of the picture is completed, it becomes increasingly easier to identify the pictures. After a series of such pictureidenti®cation tasks has been given, children are provided with degraded pictures of both previously seen versus unseen objects. The typical ®nding is that repetition priming is observed. That is, children identify fragmented pictures of previously seen pictures much faster than fragmented pictures of previously unseen objects (see Ausley & Guttentag, 1993). Findings from developmental studies further indicate age-related dissociations. That is, age-related differences between younger and older children typically observed on explicit memory tests are greatly reduced with implicit memory testings (Parkin, 1993). More speci®cally, the priming effects seem comparable for older and younger participants. Although older children typically show better explicit memory (i.e., they recognise more pictures than the younger children), no developmental differences in implicit memory are found. This ®nding is not only true for perceptual priming, but seems generalisable to implicit memory tests based on conceptual priming, that is, tests emphasising the semantic relationships between studied and tested items (Anooshian, 1997). Overall, the implicit memory results to date point to an independent memory system. It seems that the brain systems mediating perceptual and

158

Schneider

conceptual priming are fully developed early in life, which clearly contrasts with the continuous development of the explicit memory system (cf. Squire, Knowlton, & Musen, 1993). Fuzzy trace theory An increasingly popular theory that attempts to explain memory and cognitive development without invoking conscious awareness is fuzzy trace theory (see Brainerd & Reyna, 1993; Reyna & Brainerd, 1995). The general idea is that memory representations exist on a continuum from literal, verbatim representations to fuzzy, gist-like traces and that multiple-memory traces exist for any event one experiences. The type of trace people will use depends on the nature of the task, the durability of the trace, and the processing preferences of the individual. Although the assumption is that people of all ages extract the gist of a message and thus build up fuzzy traces, there are developmental differences in gist extraction. That is, children of all ages extract gist, but young children are biased toward storing and retrieving verbatim traces, resulting in a verbatim to gist shift that occurs sometime during the early elementary school years. This implies that the memory system for processing verbatim information develops earlier than the system for processing gist. The theory also assumes that relying on fuzzy traces enables children to resist the adverse effects of interference. Thus, Brainerd and Reyna propose that age differences in cognitive processes can be attributed to age-correlated changes in the construction of memory traces, with an increasing reliance on gist-like fuzzy traces. Although fuzzy trace theory has been tested using various cognitive tasks, investigations using memory tasks focused on cognitive triage, which conceptualises free recall as a process in which the most dif®cult items are recalled ®rst, analogous to the battle®eld surgical practice of treating the most dif®cult cases ®rst. Brainerd, Reyna, and colleagues developed an ``optimisation model'', which predicted that recall of a word list should begin with the recall of items with relatively weak memory strength, followed by strong items, and completed by weaker items. This model de®nitively is not in accord with assumptions of the traditional ``strategic memory model'', which would predict that recall of stronger items should precede that of weaker items. Although further details of the model cannot be presented, owing to space restrictions, it seems important to note that several experiments conducted by Brainerd, Reyna, and colleagues have directly contrasted strategic versus triage models of free recall, with most results yielding a closer ®t with the triage than with the strategic model. Thus it appears that cognitive triage is real, and that this extension of fuzzy trace theory resulted in the discovery and explanation of counterintuitive patterns of order of output in free recall.

8. Research on memory development

159

Long-term retention and forgetting The paucity of research on long-term retention and forgetting since the early 1930s may be due to the failure to ®nd age differences in the early studies conducted around the turn of the last century (see the earlier discussion). Only recently, Brainerd, Howe, Reyna, and colleagues implemented a research programme that tried to overcome certain weaknesses of the classic studies (cf. Brainerd, Reyna, Harnishfeger, & Howe, 1993; Howe & Brainerd, 1989). To do so, they replaced age-insensitive recognition measures by cued or free recall tasks, and also avoided ¯oor effects in forgetting by using less memorable materials and longer forgetting intervals. Finally, the ``levels-of-learning'' problem of earlier research (i.e., the fact that age differences in immediate levels of learning can mask age differences in subsequent amounts of forgetting) was dealt with by requiring participants of all age levels to learn the material to a perfect-recall criterion. As a main result, Brainerd, Howe, Reyna, and colleagues found substantial developmental differences in long-term retention in that older children retained more (and thus forgot less) than younger ones. Consistent improvements in retention were documented from the early elementary school years to adolescence. Thus, the use of more sophisticated experimental paradigms indicates that long-lived assumptions concerning the developmental invariance of long-term retention and forgetting need to be revised. It should be noted, however, that the new interpretations are restricted to the particular methodology used in this research paradigm. The development of eyewitness memory With children being increasingly called upon to provide testimony in legal cases, issues about how much and how accurately children remember, and the degree to which they are in¯uenced by suggestion, has become a highpriority research interest (for recent reviews, see Bruck & Ceci, 1999; Ceci & Bruck, 1998). Eyewitness memory represents one speci®c form of event memory that emphasises the accuracy and not so much the amount of information recalled about an experienced event. In most experimental studies, children witnessed either events staged at their schools, or videotaped events, or were asked to recollect some potentially traumatic experiences, such as visits to the doctor or dentist. During the interview period following the event, children were typically ®rst questioned using general prompts, usually followed by more speci®c questions. Regarding age differences in eyewitness memory, most investigations revealed that young children's free recall levels were generally low, and that free recall increased with age. However, despite low levels of free recall, what preschoolers and kindergarteners do recall is usually accurate. Thus, despite age differences, most young children possess the cognitive capacity

160

Schneider

necessary for accurate testimony. Age differences may diminish or even disappear under certain conditions, for instance, if an event is particularly salient or personally meaningful to young children. In most studies, levels of correct recall increased when more speci®c cues were provided, unfortunately in most cases also accompanied by an increase in the number of inaccurate responses, which reduced overall accuracy particularly in younger children. Although eyewitness memory improves with age, differences seem most pronounced for the period between 4 and 7 years of age. Most studies that included samples of adults illustrated that elementary school children's free recall of witnessed events is comparable to that of adults (see the review by Bjorklund, 2000). An important topic in children's eyewitness memory concerns age differences in susceptibility to suggestion. In most suggestibility paradigms, participants witness an event and are later asked sets of misleading questions, suggesting an inaccurate ``fact''. Most studies that have looked for developmental differences in suggestibility have found them, with preschool children being particularly suggestible to suggestion, much more than schoolchildren and adults. It appears that young children's erroneous answers do not always re¯ect an actual change in memory representation. It is likely that some of the young children's compliance with misleading questions is related to the social demand characteristics of the situation. However, other research suggests that misleading questions actually result in changes in the underlying memory representations, with young children being more likely to make such changes than older children (Ceci & Bruck, 1998). How long do memories of witnessed events last? Are there developmental differences in long-term recall and forgetting rates? Several studies have investigated children's memories of speci®c events for periods ranging from several weeks to 2 years. Although the results of these studies are not totally consistent, they indicate that age differences in the accuracy of recall increase with increasing delays, at least when delays are longer than 1 month. This ®nding is in accord with the experimental research on longterm retention and forgetting conducted by Brainerd, Howe, Reyna, and colleagues, and may be explained by fuzzy trace theory. That is, given that verbatim traces, favoured by young children, deteriorate more rapidly than the gist, or fuzzy traces preferred by older children, greater losses of information over delays should be expected for younger as compared to older children. FUTURE DIRECTIONS FOR MEMORY DEVELOPMENT RESEARCH Although memory development is a rather mature ®eld, there is still much to learn, and the centrality of memory to all other aspects of cognition

8. Research on memory development

161

makes it likely that the ontogeny of memory, in one form or another, will continue to be a primary focus of cognitive development. On a more general level, an interesting question is whether the study of memory needs to be situated in a number of broader conceptual and research contexts, or whether future research activities should focus on more speci®c memory functions. Another issue that most researchers agree upon is that the neuropsychology of memory development deserves more attention. The position that future research on memory development should be embedded within the context of broader topics such as comprehension, knowledge, context, and strategy has been recently proposed by Kuhn (2000). The basic claim is that although memory research has contributed signi®cantly to our understanding in each of these four areas, the study of memory development in turn needs to pro®t from what has been accomplished in each of those areas and to be enriched by integration with each of them. Kuhn distinguishes between two increasingly divergent branches of memory research, one connected to the topics of comprehension and knowledge and focusing on the study of ``memories'', and the other invoking the topics of context and strategies and referring to the study of ``memorising''. Whereas Kuhn's position operates with rather general memory concepts such as ``memories'' and ``memorising'', other research seems to indicate that the assumption of general memory concepts may oversimplify the case. For instance, several recent studies concerned the issue whether verbal memory represents a domain-general skill or consists of domain-speci®c verbal abilities (see the review by Schneider & Bjorklund, 1998). The way to assess this question empirically is to administer different verbal memory tasks and assess the intertask correlations. Overall, cross-sectional studies show that intertask correlations are rather low, regardless of age, and that the low correlations are not due to reliability problems. Further con®rming evidence comes from recent longitudinal research (Schneider & Weinert, 1995). Here, intertask consistency was high only for very similar verbal memory tasks (e.g., parallel measures of story recall or memory span). Thus these ®ndings are consistent with the position that there is no ``unitary'' verbal memory construct, and that deliberate verbal memory may be better thought of as set of speci®c abilities rather than a domain-general concept. More research is needed to explore the dimensionality of the construct and possible changes across the lifespan. Finally, many cognitive developmentalists have recognised that the development of information processing cannot be isolated from development of the brain (Bjorklund, 2000; Johnson, 1998). As noted by van der Molen and Ridderinkhof (1998), developmental psychology has been relatively ``brainless'' for a long time, despite the fact that its founding father, Wilhelm Preyer, was concerned speci®cally with brain±behaviour relations. In the ®eld of memory development, there is an increasing emphasis on the need for greater awareness of and interaction with

162

Schneider

neuroscience perspectives (e.g., Gathercole, 1998; Nelson, 1997). However, we still know very little about the neuropsychology of memory development, at the level both of description and of explanation. Although it seems important for memory development researchers to become familiar with neuropsychological techniques and research ®ndings, it is also important not to lose sight of the cognitive and behavioural levels while examining brain functions. Accordingly, biological levels of explication should not replace psychological-level theories, but should be integrated with them to yield a proper perspective on memory development.

REFERENCES Anooshian, L.J. (1997). Distinctions between implicit and explicit memory: Signi®cance for understanding cognitive development. International Journal of Behavioral Development, 21, 453±478. Ausley, J.A., & Guttentag, R.E. (1993). Direct and indirect assessments of memory: Implications for the study of memory development during childhood. In M.L. Howe & R. Pasnak (Eds.), Emerging themes in cognitive development (pp. 234± 264). New York: Springer. Baddeley, A. (1986). Working memory. Oxford: Clarendon Press. Bahrick, L.E., & Pickens, J.N. (1995). Infant memory for object motion across a period of three months: Implications for a four-phase attention function. Journal of Experimental Child Psychology, 59, 343±371. Bauer, P.J. (1997). Development of memory in early childhood. In N. Cowan (Ed.), The development of memory in childhood (pp. 83±112). Hove, UK: Psychology Press. Binet, H. (1900). La suggestibilite [On suggestibility]. Paris: Schleicher FreÁres. Bjorklund, D.F. (1985). The role of conceptual knowledge in the development of organization in children's memory. In C.J. Brainerd & M. Pressley (Eds.), Basic processes in memory development (pp. 103±142). New York: Springer. Bjorklund, D.F. (2000). Children's thinkingÐDevelopmental function and individual differences (3rd ed.). Belmont, CA: Wadsworth. Bjorklund, D.F., & Douglas, R.N. (1997). The development of memory strategies. In N. Cowan (Ed.), The development of memory in children (pp. 201±246). Hove, UK: Psychology Press. Bjorklund, D.F., & Schneider, W. (1996). The interaction of knowledge, aptitude, and strategies in children's memory performance. In H. Reese (Ed.), Advances in child development and behavior, Vol. 26 (pp. 59±89). San Diego, CA: Academic Press. Bourdon, B. (1894). In¯uence de l'age sur la memoire immeÂdiate [The impact of age on immediate memory]. Revue Philosophique de la France et de l 'EÂtranger, 38, 148±167. Bous®eld, W.A., Esterson, J., & Whitmarsh, G.A. (1958). A study of developmental changes in conceptual and perceptual associative clustering. The Journal of Genetic Psychology, 92, 95±102. Brainerd, C.J., & Reyna, V.F. (1993). Domains of fuzzy trace theory. In M.L. Howe

8. Research on memory development

163

& R. Pasnak (Eds.), Emerging themes in cognitive development (pp. 50±93). New York: Springer. Brainerd, C.J., Reyna, V.F., Harnishfeger, K.K., & Howe, M.L. (1993). Is memorystrength grouping good for recall? Unpublished manuscript. Bruck, M., & Ceci, S.J. (1999). The suggestibility of children's memory. Annual Review of Psychology, 50, 419±439. Brunswik, E., Goldscheider, L., & Pilek, E. (1932). Untersuchungen zur Entwicklung des GedaÈchtnisses [Studies on the development of memory]. In W. Stern & O. Lipmann (Eds.), Beihefte zur Zeitschrift fuÈr angewandte Psychologie (Beiheft 64). Leipzig, Germany: Ambrosius Barth. BuÈhler, K. (1918). Die geistige Entwicklung des Kindes. Jena, Germany: Fischer. BuÈhler, K. (1930). Die Erinnerungen des Kindes [Children's long-term recall]. In K. BuÈhler (Ed.), Die geistige Entwicklung des Kindes (6th ed., pp. 318±329). Jena, Germany: Fischer. Case, R. (1985). Intellectual development: A systematic reinterpretation. New York: Academic Press. Ceci, S.J., & Bruck, M. (1993). Suggestibility of the child witness: A historical review and synthesis. Psychological Bulletin, 113, 403±439. Ceci, S.J., & Bruck, M. (1998). Children's testimony: Applied and basic issues. In W. Damon (Series Ed.), I. Siegel & K.A. Renninger (Vol. Eds.), Handbook of child psychology: Vol. 4. Child psychology in practice. New York: John Wiley. Chamberlain, A.H. (1915). Memory test with school children. The Psychological Review, 22, 71±76. Chi, M.T.H., & Ceci, S.J. (1987). Content knowledge: Its role, representation, and restructuring in memory development. In H.W. Reese (Ed.), Advances in child development and behavior, Vol. 20 (pp. 91±142). New York: Academic Press. Cowan, N. (1997). The development of working memory. In N. Cowan (Ed.), The development of memory in childhood (pp. 163±200). Hove, UK: Psychology Press. Cowan, N., Nugent, L., Elliott, E., Ponomarev, I., & Saults, J. (1999). The role of attention in the development of short-term memory: Age differences in the verbal span of apprehension. Child Development, 70, 1082±1097. Dempster, F.N. (1981). Memory span: Sources of individual and developmental differences. Psychological Bulletin, 89, 63±100. È ber das GedaÈchtnis. [On memory]. Darmstadt, Germany: Ebbinghaus, H. (1885). U Wissenschaftliche Buchgesellschaft. È ber eine neue Methode zur PruÈfung geistiger FaÈhigkeiten Ebbinghaus, H. (1897). U und ihre Anwendung bei Schulkindern. Zeitschrift fuÈr Psychologie und Physiologie der Sinnesorgane, 13, 401±459. Flavell, J.H. (1971). First discussant's comments: What is memory development the development of? Human Development, 14, 272±278. Flavell, J.H., Miller, P.H., & Miller, S.A. (1993). Cognitive development (3rd ed.). Englewood Cliffs, NJ: Prentice Hall. Flavell, J.H., & Wellman, H.M. (1977). Metamemory. In R.V. Kail & J.W. Hagen (Eds.), Perspectives on the development of memory and cognition. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Gathercole, S.E. (1998). The development of memory. Journal of Child Psychology and Psychiatry, 39, 3±27. Harnishfeger, K.K., & Bjorklund, D.F. (1990). Children's strategies: A brief history.

164

Schneider

In D.F. Bjorklund (Ed.), Children's strategies: Contemporary views of cognitive development (pp. 1±22). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Hasselhorn, M. (1992). Task dependency and the role of category typicality and metamemory in the development of an organizational strategy. Child Development, 63, 202±214. Howe, M.L., & Brainerd, C.J. (1989). Development of children's long-term retention. Developmental Review, 9, 301±340. Johnson, M. (1998). The neural basis of cognitive development. In W. Damon (Series Ed.), D. Kuhn & R.S. Siegler (Vol. Eds.), Handbook of child psychology: Vol. 2. Cognition, perception and language (5th. ed.). New York: John Wiley. Kail, R.V. (1990). The development of memory in children (3rd. ed.). New York: W.H. Freeman. Kail, R.V. (1991). Processing time declines exponentially during childhood and adolescence. Developmental Psychology, 27, 259±266. Kail, R.V. (1993). The role of a global mechanism in developmental change in speed of processing. In M.L. Howe & R. Pasnak (Eds.), Emerging themes in cognitive development: Vol 1. Foundations (pp. 97±119). New York: Springer. Kail, R.V., & Salthouse, T.A. (1994). Processing speed as a mental capacity. Acta Psychologica, 86, 199±225. Koffka, K. (1921). Die Grundlagen der psychischen Entwicklung. Osterwieck, Germany: Zickfeld. Koppenaal, R.J., Krull, A., & Katz, H. (1964). Age, interference, and forgetting. Journal of Experimental Child Psychology, 1, 360±375. Kuhlmann, F. (1908). On the analysis of memory consciousness for pictures of familiar objects. American Journal of Psychology, 18, 389±420. Kuhn, D. (2000). Does memory development belong on an endangered topic list? Child Development, 71, 21±25. Meacham, J.A. (1977). Soviet investigations of memory development. In R.V. Kail & J.W. Hagen (Eds.), Perspective on the development of memory and cognition, Vol. 9 (pp. 273±295). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Meltzoff, A.N. (1995). What infant memory tells us about infantile amnesia: Longterm recall and deferred imitation. Journal of Experimental Child Psychology, 59, 497±515. Meumann, E. (1907). Entwicklung der einzelnen geistigen FaÈhigkeiten beim Kinde [The development of intellectual abilities in the child]. In E. Meumann (Ed.), Vorlesungen zur EinfuÈhrung in die Experimentelle PaÈdagogik und ihre psychologischen Grundlagen, Vol. 1 (pp. 170±203). Leipzig, Germany: Wilhelm Engelmann. Miller, P.H., & Seier, W.L. (1994). Strategy utilization de®ciencies in children: when, where, and why. In H.W. Reese (Ed.), Advances in child development and behavior, Vol. 25 (pp. 107±156). San Diego, CA: Academic Press. Nagy, L. (1930). Experimentelle Untersuchungen uÈber die Entwicklung der GedaÈchtnis- und Denkformen der Knaben und MaÈdchen zwischen 7±19 Jahren [Experimental studies on the development of memory and thinking in boys and girls between 7 and 19 years of age]. In H. Volkelt (Ed.), Bericht uÈber den XI. Kongress fuÈr experimentelle Psychologie in Wien (pp. 113±122). Jena, Germany: Gustav Fischer. Nelson, C.A. (1997). The neurobiological basis of early memory development. In N.

8. Research on memory development

165

Cowan (Ed.), The development of memory in childhood (pp. 41±82). Hove, UK: Psychology Press. Nelson, K. (1996). Language in cognitive development: The emergence of the mediated mind. New York: Cambridge University Press. Netschajeff, A. (1900). Experimentelle Untersuchungen uÈber die GedaÈchtnisentwicklung bei Schulkindern [Experimental studies on memory development in school children]. Zeitschrift fuÈr Psychologie und Physiologie der Sinnesorgane, 24, 321±351. È ber Memorieren. Eine Skizze aus dem Gebiete der experiNetschajeff, A. (1902). U mentellen PaÈdagogischen Psychologie [On memorising: An example from experimental educational psychology]. In H. Schiller & Th. Ziehen (Eds.), Sammlung von Abhandlungen aus dem Gebiete der PaÈdagogischen Psychologie und Physiologie, Vol. 5. Berlin: Reuther & Reichard. Offner, M. (1924). Individuelle und sexuelle Unterschiede des GedaÈchtnisses [Individual and sex differences in memory]. In M. Offner (Ed.), Das GedaÈchtnis. Die Ergebnisse der experimentellen Psychologie und ihre Anwendung in Unterricht und Erziehung (4th ed., pp. 168±200). Berlin: Reuther & Reichard. Parkin, A.J. (1993). Implicit memory across the life-span. In P. Graf & M. Masson (Eds.), Implicit memory: New directions in cognition, development, and neuropsychology (pp. 63±131). Amsterdam: Elsevier. Pressley, M., & Van Meter, P. (1993). Memory strategies: Natural development and use following instruction. In R. Pasnak & M.L. Howe (Eds.), Emerging themes in cognitive development, Vol. 2 (pp. 128±165). New York: Springer. Preyer, W. (1882). Die Seele des Kindes [The mind of the child]. Leipzig, Germany: Grieben. Reyna, V.F., & Brainerd, C.J. (1995). Fuzzy-trace theory: An interim synthesis. Learning and Individual Differences, 7, 1±75. Schneider, W. (1999). The development of metamemory in children. In D. Gopher & A. Koriat (Eds.), Attention and performance XVII: Cognitive regulation of performance: Interaction of theory and application. Cambridge, MA: MIT Press. Schneider, W., & Bjorklund, D.F. (1998). Memory. In W. Damon (Series Ed.), D. Kuhn & R.S. Siegler (Vol. Eds.), Handbook of child psychology: Vol 2. Cognition, perception, and language (5th ed.). New York: John Wiley. Schneider, W., & Pressley, M. (1997). Memory development between 2 and 20. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Schneider, W., & Sodian, B. (1997). Memory strategy development: Lessons from longitudinal research. Developmental Review, 17, 442±461. Schneider, W., & Weinert, F.E. (1995). Memory development during early and middle childhood: Findings from the Munich Longitudinal Study (LOGIC). In W. Schneider & F.E. Weinert (Eds.), Memory performance and competencies. Issues in growth and development (pp. 263±279). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Siegler, R.S. (1998). Children's thinking (3rd ed.). Upper Saddle River, NJ: Prentice Hall. Smirnov, A.A. (1973). Problems of the psychology of memory. New York: Plenum Press. Smirnov, A.A., & Zinchenko, P.I. (1969). Problems in the psychology of memory. In M. Cole & I. Maltzman (Eds.), A handbook of contemporary Soviet psychology (pp. 452±502). New York: Basic Books.

166

Schneider

Sodian, B., & Schneider, W. (1999). Memory strategy developmentÐGradual increase, sudden insight or roller coaster? In F.E. Weinert & W. Schneider (Eds.), Individual development from 3 to 12: Findings from the Munich Longitudinal Study (pp. 61±77). Cambridge: Cambridge University Press. Squire, L., Knowlton, B., & Musen, G. (1993). The structure and organization of memory. Annual Review of Psychology, 44, 453±495. Stern, W. (1910). Abstracts of lectures on the psychology of testimony and on the study of individuality, American Journal of Psychology, 21, 270±282. Stern, W. (1914). Psychologie der fruÈhen Kindheit. Leipzig, Germany: Quelle & Meyer. Stern, W., & Stern, C. (1913). Die Kindersprache. Leipzig, Germany: Ambrosius Barth. Underwood, B.J. (1957). Interference and forgetting. Psychological Review, 64, 49± 60. Van der Molen, M.W., & Ridderinkhof, K.R. (1998). The growing and aging brain: Life-span changes in brain and cognitive functioning. In A. Demetriou, W. Doise & C. van Lieshout (Eds.), Life-span developmental psychology (pp. 35±99). New York: John Wiley. Waters, H.S. (2000). Memory strategy development: Do we need yet another de®ciency? Child Development, 71, 1001±1012. Werner, H. (1926). EinfuÈhrung in die Entwicklungspsychologie. Leipzig, Germany: Ambrosius Barth. Werner, H. (1953). Comparative psychology of mental development. New York: Science. Yendovitskaya, T.V. (1971). Development of memory. In A.V. Zaporozhets & D.B. Elkonin (Eds.), The psychology of preschool children (pp. 89±110). Cambridge, MA: MIT Press.

9

Mindreading, emotion understanding, and relationships Judy Dunn

Institute of Psychiatry, London, UK

Interest in children's understanding of mind has blossomed in an extraordinary way over the last decade (e.g., Carrothers & Smith, 1996; Zelazo, Astington, & Olson, 1999). The mapping of young children's growing grasp of the links between what people desire, think, or believe and their actions has come to dominate much of cognitive developmental psychology. A recent meta-analysis of such research strongly supports the theoretical accounts that propose that understanding of mindÐat least as re¯ected in understanding of beliefÐshow ``genuine conceptual change'' in the preschool years (Wellman, Cross, & Watson, 2001). One key set of questions for the future, in this ¯ourishing ®eld, centres on the implications of understanding mind for children's social and emotional development. Three issues stand out: ®rst, the study of individual differences in mindreading and understanding of emotions, and the broadening of the age range and cultural groups within which these are studied; second, the issue of how far the various aspects of cognitive development brought under the general label of ``social understanding'' should be differentiated; third, and most importantly, the relation of such developments in cognition to children's social relationships and their emotional experiencesÐboth as antecedents and as sequelae. Each of these issues represents a broadening of interest away from the focus on analysing the normative age changes in understanding of mind between the ages of 2 and 5 years, which has been dominant in the ®eld. INDIVIDUAL DIFFERENCES There is now an increasing recognition of the marked individual differences in young children's ability to understand mental states and the links between such understanding and human action, as assessed in standard theory of mind and emotion understanding testsÐa new interest in pursuing the implications of these differences, and their origins (e.g., Astington, 1993). Equally striking and important is the evidence that individual differences are marked too in those aspects of real-life behaviour that re¯ect relatively sophisticated or mature understanding of inner states (e.g., deception,

168

Dunn

management of con¯ict, engagement in shared imaginative play), and that these relate to differences in test performance (Astington, 1993, 2001; Dunn, 1996; Newton, 1994). This evidence on individual differences opens up a whole series of new directions for research on mindreadingÐfor instance, the links between such differences and social experiences (see later). But what about beyond the early years? It is likely that there will be increasing interest in studying such individual differences in children in the middle childhood years and adolescents. In common-sense terms it seems obvious that understanding the subtleties of links between what people feel, intend, think, and believe and how they behave continues to develop and change as children grow from preschoolers to adolescenceÐand indeed adulthood. An example of how fruitful such research can be is described later, in considering links between understanding and relationships between adolescents and their teachers (O'Connor & Hirsch, 1999). The developmental issues raised will be of great interest: To what extent are there continuities in individual differences in these domains of understanding as people move from one developmental phase to another, from one social world and peer group to another? A start on such research has been made for the early years (Dunn, 1995; Hughes & Dunn, 1998), but it is only a ®rst step. In early childhood, the signi®cance of the affective context for the use children make of their developing understanding of others' emotions is clear (Dunn, 2001). How does this change as children grow up? Their powers of metacognition clearly increase as they grow upÐdoes this mean that they are less at the mercy of their emotions? Some of the research on children's behaviour in con¯ict with others suggest that this may be so (e.g., Dunn & Herrera, 1997). Does the signi®cance of emotion for what is learned in particular interactions change, then, as they grow up? Studies that examine and incorporate cultural variation will be especially valuable and welcome (see Lillard, 1999). Most research into the development of understanding and emotion has been on middle-class children in the USA and in Europe. Yet we know that educational and social class differences are implicated in individual differences in affective understanding, in socialisation practices, and in children's theory of mind abilities (e.g., Hughes & Cutting, 1999). Moreover, cultural groups differ notably in emotional regulation and expression, and in shared concepts of their social world (Lutz & White, 1986). These cultural in¯uences on how experiences are perceived and classi®ed are importantly mediated through relationships (Hinde, 1987). DIFFERENTIATION OF UNDERSTANDING OF MIND AND EMOTION A set of questions that are likely to feature increasingly in research on children's social understanding concerns differentiation of various aspects

9. Mindreading

169

of social understanding. How is children's understanding of mental life related to other aspects of their cognitive development? How is children's growing grasp of other people's emotions, for example, linked to their understanding of mental life? Do individual differences in children's discovery of the mind in fact re¯ect differences and development in a more general domain of cognition? And how are these aspects of cognitive development related developmentally? Tackling these questions means facing some central issues in psychology (including, for instance, a topic of much current controversyÐthe modularity of brain function, see Moore, 1996). Several lines of research are beginning to develop here. For example, in relation to the intriguing questions about the general pattern of developmental change, Bartsch and her colleagues have used evidence on children's talk about the mind (Bartsch & Wellman, 1995), and experimental studies, to make a key developmental point. Very young children explain people's actions at ®rst in terms of emotions and desires. Then through their social experiences, Bartsch and Wellman argue, they come to incorporate the notion of belief into their understanding of why people behave the way they do. An important implication of this is that an understanding of cognitive states arises through an earlier understanding of emotional states (Bartsch & Estes, 1996). This, they point out, has wide implications for our understanding of developmentÐthat accounts of metacognition will have to be anchored in a much broader understanding of development, and will require ``a better understanding of the relationship between cognitive and noncognitive psychological phenomena'' (Bartsch & Estes, 1996, p. 299). Two promising lines of research concerning the issue of differentiation both link up with the new interest in individual differences. First, there are ®ndings from longitudinal research, which suggest that the sequelae of individual differences in success on mindreading tasks may differ from those of early success on emotion understanding tasks (note that these are by no means always found to be strongly relatedÐthough the ®ndings of different studies vary). Early mindreading skills have been found to be related to a high degree of connected communication with peers, to sophisticated and frequent role play (Dunn, 1996), and, in two separate studies, to increased sensitivity to criticism when ®rst adjusting to the world of school (Cutting & Dunn, 2001; Donelan-McCall & Dunn, 1997). In contrast, early skills at understanding emotions have been reported to be associated with later peer popularity, sophisticated understanding of feelings, and aspects of moral sensibility (Dunn, 1995; Dunn, Cutting, & Demetriou, 1999). Second, research into mindreading within an attachment frameworkÐ which is likely to be a very active ®eld over the next few yearsÐsuggests that differences in early attachment security may be associated with some aspects of later mindreading skills but not others (Meins, 1997; Meins, Fernyhough, Russell, & Clarke-Carter, 1998). Clearly the question of what

170

Dunn

processes might mediate such connections deserves research attention (intractable though it may be). LINKS BETWEEN EMOTION, RELATIONSHIPS, AND SOCIAL UNDERSTANDING The issue of links between attachment relations and later understanding brings us to a centrally important future topic of research. What children understand about the feelings and inner states of those with whom they have close relationships clearly carries great importance for those relationships; their grasp of how others' behaviour is linked to their feelings, beliefs, thoughts, and intentions is likely to be of major importance in their social lives. But the development of their understanding of others may also be in¯uenced by their emotional and social experiences, as the attachment research indicates. Exciting new areas of research are opening up as investigators explore the links between children's social and emotional experiences, and their abilities to understand mind and emotion, both as antecedents and as sequelae. The results of the ®rst studies have exciting implications for both domains.

Antecedents and sequelae Naturalistic observational studies, combined with standard assessments of theory of mind and emotion understanding, have highlighted several kinds of interactive events that not only reveal children's social understanding but may be implicated in fostering its development: shared imaginative play, discourse about inner states, collaborative narratives, arguments in disputes, and exposure to and engagement in deception. What is needed now is more work to help us to clarify the direction of effects in such associations: not only longitudinal studies but experimental studies tooÐto move beyond correlational studies. Further sensitive observational studies are also likely to prove very illuminating (Newton, 1994), particularly in the very early stages of children's discovery of the mindÐstudies of the second year, for instance, that include observations of teasing, deception, and joking are likely to pay dividends, if they are conducted in the real worlds of children's social relationships. Studies of discourse about inner states are likely to be very fruitful. Such research has repeatedly shown associations between children's conversational experience and their success on mindreading tasks (Dunn, 1996). We need more longitudinal studies to tease apart the direction of effects here (see, for example, Hughes & Dunn, 1998). A whole range of questions have been raised that merit study: Is it the ``content'' of such conversations that is all-important, or do the pragmatics also play a role? What precipitates these conversations in children's lives? Is it the case that the social partner

9. Mindreading

171

matters? Do other children, for instance, play a special role (see, for instance, Dunn, 1996; Piaget, 1932/1965)? Does the child's emotional state, while engaging in such discourse, play a role in what is learned, as has been suggested (but by no means established)? The importance of language in the development of mindreading is gaining increasing attention (Astington, 2001). Astington points out that language ability is key to the growth of understanding mind in two ways: It is both the means by which children become aware of others' beliefs and thoughts, and it also makes possible the representation of false belief in contradistinction to the evidence of reality. Among those that focus on the representational signi®cance, the de Villiers have argued, for instance, that it is acquiring the syntax of complementation that provides the representational format needed for the understanding of false belief (de Villiers & de Villiers, 2000). As Astington notes, these aspects of the importance of language are not competing, but complementary accounts. For children to develop understanding of mind, both the representation of mental states made possible through language and the experience of participating in discourse about inner states are crucial. Studies that investigate the issue of antecedents and sequelae of social understanding in terms of relationships (rather than solely interactive events) are likely to be importantÐas we noted earlier in considering attachment and differentiation of aspects of mindreading (see also Fonagy, Redfern, & Charman, 1997). Two interesting new directions in developmental research here have begun to link the role of conversational experiences with the quality of close relationships. For instance, Thompson and his colleagues have shown that mother±child pairs with secure attachment relationships engage in discourse that includes more frequent references to emotion, suggesting that a secure attachment fosters the understanding of negative emotions (Laible & Thompson, 2000). Thompson proposed that the emotion-laden discourse in which the securely attached pairs engaged was likely to make emotions more accessible and less threatening for a child when he or she re¯ected on past personal experiences. Links between discourse about inner states, the security of early attachment relationships, and children's later social understanding were also highlighted in the research of Meins and her colleagues (Meins et al., 1998): Children who were securely attached were more successful on mindreading tasks at 4 years, and at 5 years showed greater ``mentalising'' abilities. Their mothers, apparently, had shown particular sensitivity to the children's current level of understanding, and had used more mental state terms in their conversations with themÐa quality that Meins terms ``mind-mindedness'' (Meins, 1997). Not only is the research on attachment promising here; there is new interest in the friendships of very young children, shown to be associated with individual differences in understanding of inner states (Dunn & Cutting, 1999). A second interesting new direction in research on understanding of mind is the study of understanding of the past. This brings together research on

172

Dunn

developments in memory, and emotion, in children's explanations for people's behaviour. Lagattuta and Wellman (2001), for example, have argued on the basis of their experimental studies that children recognise that recollections of powerful emotions (especially negative feelings) can induce emotions that do not match realityÐthat angry thoughts and memories can induce feelings of anger. A key theme here is that emotion understanding provides a context for later crucial developments in understanding others (Lagattuta, Wellman, & Flavell, 1997). Finally, it should be noted that genetic in¯uences on individual differences must be acknowledged; this is again a growing area of research. Individual characteristics that contribute to the quality of close relationshipsÐand thus in turn play a part in children's growing understanding of othersÐare importantly in¯uenced by genetics. For example, differences in emotion understanding and mindreading (Hughes & Cutting, 1999), in empathy (Zahn-Waxler, Robinson & Emde, 1992), in humour (Manke et al., 2002), and in self-esteem and self-con®dence (McGuire et al., 1999), in addition to differences in aspects of temperament (Goldsmith, Lemery, Buss & Campos, 1999), have all been shown to be in¯uenced by genetics.

Differential use of understanding in different relationships The issue of how children's mindreading and emotion understanding are linked to differences in the quality of their relationships is highlighted in a new way by some intriguing ®ndings that suggest another fruitful direction for research. In a study in which children were observed within their various different relationshipsÐwith their mothers, siblings, and friendsÐit was found that the children used their powers of understanding differently in these relationships (Dunn, 1996). In three different domains (engaging in discourse about mental states, sharing imaginative worlds, handling disputes) there were no signi®cant correlations between the same child's behaviour with mother, sibling, and friend. The key point suggested by these ®ndings is that whether children use their understanding of other people's inner states depends on the nature of their relationship with the other. Of course, what happens within real-life interactions will depend on both partners, and is unlikely simply to be linked to the sociocognitive skills of either partner in the interaction. This implies that different kinds of social relationships will play different roles in in¯uencing the development of social understandingÐa proposal that certainly deserves further research. The potential importance of such study is well illustrated by a recent study of adolescents by O'Connor and his colleagues (O'Connor & Hirsch, 1999), in which adolescents' relationships with their various teachers, and their understanding and attributions concerning these individuals, were investigated. The ``accuracy'' with which the adolescents mentalised about

9. Mindreading

173

their teachers was relationship-speci®c, and predicted from the affective quality of their relationship.

LINKS BETWEEN UNDERSTANDING AND SOCIAL RELATIONS IN EXTREME GROUPS How far do the patterns of connections between social relationships and understanding of mind and emotion hold for children who have problems in relationships, or who suffer from cognitive dif®culties? With the notable exception of studies of autism, until very recently most research on the links between understanding and social relationships was focused on normal children. However, an exciting new direction in such research is to ask whether the nature and patterns of early connections between understanding, emotional experience, and social relationships differs for children with problems such as hyperactivity, conduct problems, language delay, or anxiety. In all of these groups, we know that, by middle childhood, children are at risk for problems in both close relationships and social understanding (e.g., Dodge, 1991). But how are these related in early childhood? Little is currently known, though a promising start has been made in the case of hyperactive children by Hughes and her colleagues (Hughes, Dunn, & White, 1998). In a longitudinal study of ``hard-to-manage'' children, the children's cognitive differences were associated with their observed antisocial behaviour; cognitive, emotion regulation, and social problems were closely related. To address these issues longitudinally, with children who have other problems in cognition and relationships, will be to make a contribution that is both theoretically and socially valuable. Parallel arguments about the signi®cance of emotion regulation and understanding in social information processing of bullies and aggressive children have highlighted the importance of further research into these links (Arsenio & Lemerise, 2001).

IMPLICATIONS FOR THEORY In summary, the current research on children's discovery of the mind presents us with both challenges and opportunities; to address the central theoretical issue of how children's emotional and social experiences are linked to their cognitive development will be of particular signi®cance. The cognitive research that has clari®ed children's growing understanding of mental life has opened up a new range of issues, whereas the naturalistic research has highlighted the signi®cance of the emotional settings in which children's understanding is fostered. We now must address the exciting developmental questions raised by the current work, focus not only on

174

Dunn

children in their pre-school years, and include cultural variation in our studies, if we are to make progress in understanding the nature of the links between social and cognitive change in children. REFERENCES Astington, J.W. (1993). The child 's discovery of the mind. Cambridge, MA: Harvard University Press. Astington, J.W. (2001). The future of theory-of-mind research: Understanding motivational states, the role of language, and real-world consequences. Child Development, 72, 685±687. Arsenio, W.F., & Lemerise, E.A. (2001). Varieties of childhood bullying: Values, emotion processes, and social competence. Social Development, 10, 59±73. Bartsch, K., & Estes, D. (1996). Individual differences in children's developing theory of mind and implications for metacognition. Learning and Individual Differences, 8, 281±304. Bartsch, K., & Wellman, H.M. (1995). Children talk about the mind. Oxford: Oxford University Press. Carrothers, P., & Smith, P.K. (1996). Theories of theories of mind. Cambridge: Cambridge University Press. Cutting, A., & Dunn, J. (2001). Making the transition to school: Preschool predictors and personal perceptions. Manuscript submitted for publication. de Villiers, J.G., & de Villiers, P.A. (2000). Linguistic determinism and the understanding of false beliefs. In P. Mitchell & K.J. Riggs (Eds.), Children's reasoning and the mind (pp. 191±228). Hove, UK: Psychology Press. Dodge, K.A. (1991). Emotion and social information processing. In J. Garber & K. Dodge (Eds.), Emotion regulation (pp. 159±181). Cambridge: Cambridge University Press. Donelan-McCall, N., & Dunn, J. (1997). School work, teachers, and peers: The world of ®rst grade. International Journal of Behavioral Development, 21, 155±178. Dunn, J. (1995). Children as psychologists: The later correlates of individual differences in understanding of emotions and other minds. Cognition and Emotion, 9, 187±201. Dunn, J. (1996). The Emanuel Miller Memorial Lecture 1995: Children's relationships: Bridging the divide between cognitive and social development. Journal of Child Psychology and Psychiatry and Allied Disciplines, 37, 507±518. Dunn, J. (2002). Emotional development in early childhood: A social relationship perspective. In R. Davidson, H.H. Goldsmith, & K. Scherer (Eds.), The handbook of affective science. Oxford: Oxford University Press. Dunn, J., & Cutting, A. (1999). Understanding others, and individual differences in friendship interactions in young children. Social Development, 8, 201±219. Dunn, J., Cutting, A., & Demetriou, H. (1999). Moral sensibility, understanding other, and children's friendship interactions in the preschool period. British Journal of Developmental Psychology, 18, 159±178. Dunn, J., & Herrera, C. (1997). Con¯ict resolution with friends, siblings, and mothers: A developmental perspective. Aggressive Behavior, 23, 343±357.

9. Mindreading

175

Fonagy, P., Redfern, S., & Charman, A. (1997). The relationship between beliefdesire reasoning and projective measure of attachment security. British Journal of Developmental Psychology, 15, 51±61. Goldsmith, H.H,, Lemery, K.S., Buss, K.A., & Campos, J. (1999). Genetic analyses of focal aspects of infant temperament. Developmental Psychology, 35, 972±985. Hinde, R.A. (1987). Individuals, relationships, cultures. Cambridge: Cambridge University Press. Hughes, C., & Cutting, A.L. (1999). Nature, nurture and individual differences in early understanding of mind. Psychological Science, 10, 429±432. Hughes, C., & Dunn, J. (1998). Understanding mind and emotion: Longitudinal associations with mental-state talk between young friends. Developmental Psychology, 34, 1026±1037. Hughes, C., Dunn, J., & White, A. (1998). Trick or treat?: Uneven understanding of mind and emotion and executive function among ``hard to manage'' preschoolers. Journal of Child Psychology and Psychiatry, 39, 981±994. Lagattuta, K.H., & Wellman, H.M. (2001). Thinking about the past: Early knowledge about links between prior experience, thinking and emotion. Child Development, 72, 82±102. Lagattuta, K.H., Wellman, H.M., & Flavell, J.H. (1997). Preschoolers' understanding of the link between thinking and feeling: Cognitive cuing and emotional change. Child Development, 68, 1081±1104. Laible, D.J., & Thompson, R.A. (2000). Mother±child discourse, attachment security, shared positive affect, and early conscience development. Child Development, 71, 1424±1440. Lillard, A. (1999). Developing a cultural theory of mind: The CIAO approach. Current Directions in Psychological Science, 8, 57±61. Lutz, C., & White, G.M. (1986). The anthropology of emotions. Annual Review of Anthropology, 15, 405±436. Manke, B., Pike, A., Carson, C., Robertson, R., Dunn, J., & Plomin, R. (2002). Associative and distancing humor in adolescent family relations: A genetic study of individual and dyadic relationship effects. Manuscript submitted for publication. McGuire, S., Manke, B., Saudino, K.J., Reiss, D., Hetherington, E.M., & Plomin, R. (1999). Perceived competence and self worth during adolescence. Child Development, 70, 1283±1296. Meins, E. (1997). Security of attachment and the social development of cognition. Hove, UK: Psychology Press. Meins, E., Fernyhough, C., Russell, J.T., & Clarke-Carter, D. (1998). Security of attachment as a predictor of symbolic and mentalising abilities: A longitudinal study. Social Development, 7, 1±24. Moore, C. (1996). Evolution and the modularity of mindreading. Cognitive Development, 11, 605±621. Newton, P. (1994). Preschool prevarication: An investigation of the cognitive prerequisites for deception. Unpublished PhD thesis, Portsmouth University, UK. O'Connor, T., & Hirsch, N. (1999). Intra-individual differences and relationshipspeci®city of mentalising in early adolescence. Social Development, 8, 256±274. Piaget, J. (1932/1965). The moral judgement of the child. New York: Academic Press. Wellman, H.M., Cross, D., & Watson, J. (2001). Meta analysis of theory-of-mind development: The truth about false belief. Child Development, 72, 655±684.

176

Dunn

Zahn-Waxler, C. Robinson, J.L., & Emde, R.N. (1992). The development of empathy in twins. Developmental Psychology, 28, 1038±1047. Zelazo, P.D., Astington, J.W., & Olson, D.R. (1999). Developing theories of intention: Social understanding and self control. Mahwah, NJ: Lawrence Erlbaum Associates Inc.

10 Lifespan personality development: Individual differences among goaloriented agents and developmental outcome Cornelis F.M. van Lieshout

University of Nijmegen, The Netherlands The study of personality development is the study of the development of individual differences and, in addition, of individuals as goal-oriented agents in their striving across the life course for self-ful®lment in convergence with the environment. Across time, some consistency has been maintained in ideas concerning personality, but vast differences have also existed. In this chapter, some historical trends in the study of personality and personality development will be discussed ®rst (cf. Graumann, 1959). Second, a model of persons as self-organising, goal-oriented agents will be related to the development of basic dimensions in personality and temperament across the life course. Third, some results of studies will be presented of pattern descriptions and pattern dynamics representing the type-oriented approach to personality. Finally, from an agent-oriented view, the development of separate personality dimensions will be related to developmental outcomes in various domains across the lifespan, and future directions will be sketched.

HISTORICAL TRENDS

Towards a differential developmental psychology In 1894, Windelband (1998) distinguished the idiographic and nomothetic methods as two different, but complementary, scienti®c approaches for the psychological description of individuals. This distinction was represented by Stern (1911) as a matrix of individuals (in the columns) by person characteristics (in the rows). Each row represented the variability of a characteristic across individuals and can be correlated with the variability of the other characteristics in different rows. Each column represented the idiographic ``psychography'' or personality pro®le of an individual across the characteristics investigated; comparisons between different columns allowed estimation of the similarity or difference of individuals on the investigated characteristics or a comparison of personality pro®les. An individual's personality pro®le can be highly speci®c, but personality

178

van Lieshout

pro®les of several individuals can also be similar or comparable to a prototypical personality pro®le or represent a personality type. A third dimension, sampling behaviour across situations, was added to this matrix by Guilford (1959), transforming it into a three-dimensional matrix and allowing the systematic study of individuals, person characteristics, and behaviour across situations over time.

Consistency of personality Whether consistency in an individual's behaviour is a characteristic of the individual or a result of similarity in the situational context, distinguishes trait theorists from anti-trait theorists. As exponents of anti-trait theory, learning theorists have argued that habitual behaviour patterns, even across the life course, are the result of consistency in the stimulus quality of situations; such consistency is rather low and typically results in correlations between .20 to .30 across situations and over time (Mischel, 1968). According to this view, personality consists of a collection of habits and habit systems that are stable and resistant to change and are the result of learning. These ``social learning'' views reached an apex with the publication of Mischel's (1968) book, a publication that elicited a large group of new careful empirical studies (e.g., Bem & Allen, 1974; but also Mischel, Shoda, & Peake, 1988) and the reiteration of theoretical arguments supporting trait theory (Magnusson & Endler, 1977). Viewing personality development in terms of person±situation interactions (Endler & Magnusson, 1976), long-term longitudinal studies began to be used for demonstrating differential trajectories in personality development (Block, 1971; Block & Block, 1980). Models for the comprehensive personality description (Block, 1961) were proposed; behaviour genetic studies (Plomin & Rende, 1991; Loehlin, 1992) regained interest, and powerful statistical procedures were applied in personality research such as factor analysis, cluster analysis, and LISREL. These developments, in turn, resulted in a quiet extinction of the anti-trait theorists and a revival of personality psychology.

Taxonomies of traits During the past century, theorists all over the world proposed a variety of taxonomies of traits, emphasising different facets of personality and some of them with a developmental ¯avour (Murphy, 1947; Thomae, 1955). Only during the last 15 years, however, has support for a convergent taxonomy accumulated. The origins of this convergent taxonomy are rather diverse, stemming ®rst from the lexicographic approaches of Allport and Odbert (1936). These investigators argued that the lexicon is the main psychological archive of mankind and the foremost instrument for communicating about person perceptions; in fact, they culled nearly 18,000 person-descriptive

10. Lifespan personality development

179

adjectives from Webster's Dictionary. After many years, this approach led to the establishment and cross-language replication of the ``®ve-factor model'' or the ``big-®ve'' personality factors of extraversion, agreeableness, conscientiousness, emotional stability, and openness (cf. Goldberg, 1990). Similarly, early clinical assessments such as the Minnesota Multiphasic Personality Inventory (Hathaway & McKinley, 1943) and nonclinical personality descriptions such as the California Q-set (Block, 1961) aimed at a comprehensive description of personality. These person-descriptive instruments were also subjected to factor analyses and, despite their heterogeneous theoretical origins, again and again resulted in the big-®ve personality factors (cf. NEO-PI; Costa & McCrae, 1989). Other approaches to personality assessment, e.g., Eysenck (1967) and Cloninger (1987), who linked personality description to psychophysiological processes, or Tellegen (1991), were not incompatible with the big-®ve model (cf. Shiner, 1998). Moreover, the big-®ve model has been extended to person descriptions of children and adolescents (cf. Halverson, Kohnstamm, & Martin, 1994), and related to early temperament (Rothbart, Ahadi, & Evans, 2000; Rothbart & Bates, 1998) and to spontaneous person descriptions by parents of their children in several language areas (Kohnstamm, Halverson, Mervielde, & Havill, 1998). In sum, the temperament and personality dimensions reveal continuity across the lifespan and may represent a comprehensive set of person characteristics that ®t into Stern's (1911) and Guilford's (1959) three-component model of individuals, person characteristics, and situations-across-time, and allow the systematic study of inter-individual differences of intra-individual stability and change.

From the particular to the general Most representations of the organisation of personality in the past century have been rather static and descriptive. Eysenck (1954) and later Digman (1990) proposed a four-tiered hierarchic model, with speci®c reactions to stimuli in speci®c situations at the lowest level, habitual or modal reaction patterns in prototypical situations represented by traits at the next higher level, personality facets aggregating related traits at the next higher level, and superordinate personality dimensions like the big-®ve personality factors at the highest level. Some researchers objected to these representations, arguing that they do not properly represent the loadings of speci®c traits on one, two, and sometimes three factors in factor analyses: Circumplex models have been proposed alternatively with the big-®ve personality factors as the orthogonal axes of the circumplex, personality facets as overlapping domains, and several hundred speci®c traits as the basic elements comparable in function to the solar system or the periodic table of chemical elements (Hofstee, de Raad, & Goldberg, 1992). These structural models themselves seem rather static and insensitive to change, but have the great advantage of being comprehensive and speci®c with

180

van Lieshout

respect to the complex intertwining of personality dimensions and concomitant developmental outcomes.

Persons, types, and agents With their three-component model, Stern (1911) and Guilford (1959) were offering a perspective on differential developmental psychology as well as on the idiographic and nomothetic study of personality and personality types. In addition, Stern characterised persons as active agents. In his view, personality is a goal-oriented, self-organising system: Persons have the ``capacity'' and ``orientation'' to self-realisation, that is, a striving for completion or ful®lment in convergence with the environment. Dispositions are not self-suf®cient but function in the service of the goal-oriented selforganisation of the person. Dispositions are not ®xed entities: they can be identi®ed and function in self-preservation, but are changing and dynamical in self-realisation. Dispositions are not inborn nor acquired; as sources for future development, however, dispositions have an inborn inner causality (Stern, 1911; 1923). Accordingly, a distinction is made in this chapter between person-centred approaches to personality that are typological and others that are agent-centred, although the difference between the two may appear to be small. Stern's ideas concerning the person can be recognised much in present-day theorising. For example, David Magnusson and his colleagues (cf. Magnusson & ToÈrestad, 1993) have followed a holistic person-oriented approach in their longitudinal studies of personality development. Magnusson and coworkers have stressed the integrated functioning of individuals as intentional and active beings, as biological beings, and as social beings. In addition, psychological factors and physiological indicators in personality have been related to developmental outcome in speci®ed social contexts. Also consistent with the Stern (1911) tradition, personality types have been distinguished as patterns or pro®les of personality traits that can be mutually compared and related to developmental outcomes. This work was started by Block (1971; Block & Block, 1980) and gained renewed interest when evidence accumulated that supported characteristic and consistent individual patterns of personality traits or pro®les (cf. Caspi, 1998).

Conclusions and additional views on personality Several conclusions can be drawn from the study of personality over the past century. First, a worthwhile distinction has been made between person-centred and trait-oriented approaches to personality; both kinds of studies should continue. More than ever, these orientations can be considered as complementary (Caspi, 1998). Second, personality theories over the past century may be distinguished according to their emphasis on various elements, such as global somatic constitution (e.g., somatic body

10. Lifespan personality development

181

build, neuroendocrinological and biophysiological correlates), internal stimuli (e.g., emotional and motivational orientation), self-regulatory and intentional orientations, the perception of others and self, and the consistency of behaviour across situations and over time. Nowadays, theorists are aware of the complex interdependence of these personality facets, especially across time. A third distinctive feature is the manner in which consistency and continuity in personality and behaviour are explained. Continuity has sometimes been seen exclusively in the continuity of the personality substrate (for example, body build, biophysiological structures and processes, or genetic endowment; see McCrae et al., 2000). Such substrate has been considered as more objective, more basic, and causal in personality development than the more subjective impressions of individual differences, situational variations, and behavioural outcome. Person perception, however, has been a central feature in social relations throughout the evolution of mankind, and self-perception seems to be the backbone of the individual's sense of identity across the life course. Fourth, consistency and variability in behavioural outcome has been considered by many theorists to derive from consistency or variability in the situational context. Thus far, the intrinsic continuity of behavioural development across the life course has seldom been assumed as the main source of continuity in personality development (for an exception, see Thomae, 1955). Fifth, several personality psychologists have emphasised the functioning of persons as active agents in their own personality development. In this chapter, I aim to contribute further to that understanding. AN AGENT-CENTRED VIEW OF PERSONALITY DEVELOPMENT The goal-orientation and self-organisation of individuals as active agents seems most manifest when they are confronted with novelty in situations that appeal to their problem-solving capacities. In the great variety of those situations across long periods of time, personality development becomes visible.

Personality and problem solving Block and Block (1980) considered persons to be problem solvers, ``problem solving being de®ned to include the social and personal domains as well as the cognitive'' (p. 48). The repertoire of problem-solving strategies available to a person is de®ned by the personality dimensions egoresiliency and ego-control. Ego-control refers to the tendency to contain versus express impulses, desires, and emotions and can be represented as a continuum between overcontrol and undercontrol. Ego-resiliency is de®ned as the dynamic capacity of individuals to modify their modal levels of

182

van Lieshout

impulse control as a function of situational demands. Ego-resiliency is a continuum, extending from resourceful, ¯exible, and persistent adaptations to changing circumstances, on the one hand, to ego-brittleness, that is, the inability to respond to the dynamic requirements of the situation and a tendency to become disorganised under stress, on the other. Some people tend to control or suppress impulses, needs, and feelings (overcontrollers) whereas others give their impulses free rein (undercontrollers). Individuals with a moderate ego-control have the highest ego-resiliency (resilients); both overcontrollers and undercontrollers tend to ego-brittleness, resulting in a curvilinear relation between ego-resiliency and ego-control. Recently, Block (2000) suggests that ego-control is congruent with personality dimensions identi®ed in certain other theoretical traditions, wherein different terms are used for the same or almost the same underlying construct, for example ``sensation seeking'' (Zuckerman, Kuhlman, Joireman, Teta, & Kraft, 1993), ``disinhibition'' (Watson & Clark, 1993), ``novelty seeking'' (Cloninger, 1987), or ``constraint'' (Tellegen, 1991). The common denominator of these constructs is that they all relate to the way impulses are monitored in various attentional, emotional, behaviourexecutive, reward-oriented, and social domains, and affect differences in the degree to which individuals are activated or inhibited. Block (2000) further argues that similarity between ego-resiliency and personality dimensions contained in other theories does not exist. In the present chapter, however, I examine whether the other four big-®ve personality dimensions (excluding extraversion, which is similar to ego-control) represent facets of ego-resiliency in speci®c problem domains. That is, the question is whether ego-resiliency can be further conceptually differentiated into the other four big-®ve personality factors: agreeableness, conscientiousness, emotional stability, and openness, in relation to problem-solving in the interpersonal domain, the performance domain, the self domain, and the intellectual, creative, and artistic domain, respectively. The choice of these four problem domains will be further elaborated later in this chapter. Like ego-resiliency, at one end of the continuum, each of the four big-®ve dimensions is de®ned by ``resourceful and ¯exible adaptation to changing circumstances and environmental contingencies'' in problem solving in a speci®c domain; at the opposite end of the continuum, like ego-brittleness, ``little adaptive ¯exibility and an inability to respond to the dynamic requirements of the situation'' is implied in each of the four big-®ve dimensions (Block & Block, 1980, p. 48). As mentioned, ego-control and ego-resiliency are assumed to be curvilinearly related. Accordingly, we expect a curvilinear relation between the big-®ve dimensions extraversion and agreeableness in problem-solving in the interpersonal domain, between extraversion and conscientiousness in the performance domain, between extraversion and emotional stability regarding the development of the self, and, ®nally, between extraversion and openness in the domain of intellectual, creative, and artistic problem

10. Lifespan personality development

183

solving. Of course, there will be speci®c overlap across these problemsolving domains, as well as in the functioning of the various personality dimensions, which is quite in agreement with the circumplex character of the ®ve-factor model (cf. Hofstee, de Raad, & Goldberg, 1992).

Temperament and infant problem solving Temperament is often de®ned as a substrate for personality development, consisting of simple, basic styles that emerge early and that are closely tied to later personality dimensions (cf. Hartup & van Lieshout, 1995). From different theoretical perspectives, a large variety of temperamental dimensions have been proposed for the comprehensive description of temperament. Increasing support has been evinced recently for Rothbart's (1989) distinction between reactivity, comprising the more negatively valenced dimensions of distress-to-novelty (fear proneness) and distress-to-limitations (anger proneness) and activity level, and self-regulation, referring to positive emotionality, duration of orientation, and soothability. Reactivity refers to the onset, duration, and intensity of affective reactions as well as variability in arousability and distress in overstimulation, activity, and attention. Rothbart and Bates (1998) have sketched the development of temperament in relation to later personality development. Referring to Gray's (1975) model of brain functioning, Rothbart (1989) described the parallel between reactivity, activation, and arousability of infants with the personality dimension extraversion±introversion. Reactive infants and extraverted individuals at later ages are seen as more activated than inhibited, overcontrolling, or introverted individuals. In this model, arousability is seen as arising from the relative in¯uence of two different systemsÐthe Behavioural Activation System (BAS) and the Behavioural Inhibition System (BIS). From a very early age onwards, reactive individuals have a relatively strong BAS and a strong orientation on reward; they are described as sensation- and reward-oriented, active, energetic and outgoing, seeking stimulation or novelty, and actively avoiding aversive and pain stimuli. They are highly sociable but tend to overreact with anger and irritation if frustrated. Inhibited infants, in contrast, have a relatively strong BIS that becomes most evident at the age of about 9 months when they start being wary of novel persons and objects, show behavioural inhibition, and are more withdrawn and passive in avoiding aversive and pain stimuli. Moreover, these infants tend to react with high anxiety and fear under conditions of high arousal. Self-regulatory processes, such as executive attention and effortful control, serve to modulate reactivity, including attention, approach and withdrawal, attack and af®liation, behavioural inhibition, and self-soothing. The relation between reactivity and self-regulation incorporates the idea of

184

van Lieshout

optimal level of impulse regulation (Ahadi, Rothbart, & Ye, 1993), which, in turn, resembles Block and Block's (1980) assumption concerning the curvilinear relation between ego-control and ego-resiliency in later personality functioning. The self-regulatory dimensions of temperament have also been related to the other big-®ve personality factors (i.e., agreeableness, conscientiousness, emotional stability, and openness). First, in the affective domain in temperament research, Rothbart and Bates (1998) have differentiated positive from negative affect. In opposition to more self-regulated and resilient positive affect, two different kinds of negative affect have been identi®ed, that is, fearful distress and anxiety, when infants are confronted with novelty, and angry distress and irritability, when infants are confronted with limitations and frustration, respectively. This dimension of positive versus negative affect bears resemblance to the big-®ve personality dimension emotional stability, where resilient positive emotional balance and pleasure are opposed to anxiety and fear, emotional disorganisation under stress, shame and guilt, and anger and irritation. In terms of ego-brittleness, when under stress overcontrollers tend to fearful distress and anxiety and undercontrollers tend to angry distress and irritability. A second dimension in temperamental self-regulation concerns attention span and task persistence, also labelled duration of orienting and interest. This dimension refers to infants' regulation of information processing, their capacity to shift and focus attention, and to control action, particularly in the face of external distractions. Attention span and task persistence maps onto the big-®ve factor of conscientiousness, which refers to a person's ability to exercise resilient, organised, planful, and reliable control of action in tasks requiring achievement versus impulsive carelessness, negligence, and sloppiness in the case of undercontrol, and undue and compulsive precision in the case of overcontrol. A third dimension in self-regulation in temperament concerns soothability/af®liativeness versus irritability/anger. On this dimension, prosocial af®liation is opposed to aggressive behaviour in undercontrollers and aloof and glum withdrawal in overcontrollers, in keeping with the agreeableness±hostility dimension of the big-®ve personality model. The cognitive, creative, and artistic problem-solving domain is not explicitly covered in temperamental research (cf. Halverson, Kohnstamm, & Martin, 1994), but is represented by the big-®ve dimension of openness to new experience or intellect. In infants and toddlers, it may be hard to distinguish attention span, task persistence, duration of orientation, and interest from intellectual, creative, and artistic involvement. In sum, across the lifespan we assume a similar curvilinear relation between reactivity and self-regulation in temperament as between egocontrol and ego-resiliency in personality. Moreover, ego-control reveals similarity with extraversion whereas ego-resiliency manifests itself in various self-regulatory big-®ve factors depending on the problem-solving domain with which an individual is confronted.

10. Lifespan personality development

185

More evidence concerning the congruence between temperament and personality Additional evidence supports the distinction between reactivity and extraversion, on the one hand, and self-regulation and the other big-®ve personality factors, on the other hand. For example, the dimensions within each of the temperamental domains of reactivity and self-regulation reveal some commonality, but, in addition, they each have somewhat different genetic and environmental aetiologies. In a large study of twins, Goldsmith, Lemery, Buss, and Campos (1999) concluded that the aetiology of the reactivity dimensions of anger, fear, and activity level are best represented by a model incorporating additive genetic and nonshared environmental in¯uences, in which all similarity between co-twins is believed to be mediated genetically. In contrast, the self-regulatory dimensions of positive affect and duration of attention were best represented by a model incorporating additive genetic, shared environmental, and nonshared environmental in¯uences. In addition, soothability was best modelled without a genetic in¯uence; shared environmental in¯uences accounted entirely for the similarity between co-twins. These ®ndings suggest that infants are more open to caregiver input on the self-regulatory temperamental domains of positive affect and duration of attention, and, in particular, to different soothing techniques, whereas the receptiveness to caregiver input was least on the negatively valenced affectivity dimensions and activity level (i.e., reactivity). Like the reactivity dimension, extraversion reveals a remarkable heritability. Of all big-®ve personality dimensions, extraversion (and openness to new experience) accounts for the largest proportion of additive and nonadditive genetic variance (Loehlin, 1992). Not only are the heritability of reactivity and extraversion similar, but also the heritabilities of selfregulatory dimensions of temperament (positive affect, duration of attention, and soothability) and the personality dimensions emotional stability, conscientiousness, and agreeableness. In comparison to extraversion, emotional stability, conscientiousness, and agreeableness are more susceptible to shared and nonshared environmental in¯uences (Loehlin, 1992). Evidence was further established for the increasing stability of temperament from an early age onwards, suggesting its importance as early antecedent of later personality development. For example, Lemery, Goldsmith, Klinnert, and Mrazek (1999) concluded, from the ®ndings of ®ve longitudinal studies across the ages from infancy to 12 years, that several dimensions of temperament reveal substantial continuity, if assessed with age-appropriate measures. In infancy, at 3-month intervals between the ages of 3 to 18 months, substantial age-to-age stability was demonstrated, but also considerable change, in the position of individual infants on the temperamental dimensions of positive emotionality, distress-anger, fear, and activity level. Between the ages of 4 months and 4 years, Fox, Henderson,

186

van Lieshout

Rubin, Calkins, and Schmidt (2001) demonstrated differential associations of stability and change in behavioural inhibition and exuberance with neurophysiological correlates and quality of parental and extrafamilial nonparental care. Together, it is assumed that the stability and change of early temperament could be explained by the dynamic nature of children's interaction with the environment. From 3 to 48 months of age, these dimensions of temperament appeared to be increasingly stable. Apparently, it is harder to establish stability in the more self-regulated temperamental dimensions of duration of attention, task persistence, and soothability. A meta-analysis (Roberts & DelVecchio, 2000) also showed increasing stability of personality across later ages, with stability coef®cients from about .3 in childhood, about .5 in young adults, and increasing to a peak of .7 between the ages of 50 and 70. Stability was slightly higher for extraversion, but also for agreeableness, and lowest for emotional stability. DEVELOPMENT OF PERSONALITY TYPES Discovering continuity across the life course in the structure of temperament and personality enabled the study of the similarity in trait patterns across individuals (i.e., personality and temperament types). It was further possible to investigate whether and how these types re¯ect the curvilinear relation between activation and self-regulation. A last question concerned individual differences in patterns of adjustment or problem solving related to different personality types.

Personality types Recently, Caspi (1998) summarised the ®ndings of eight studies in which a typological approach had been adopted. In these studies, three replicable personality types emerged. One type is well-adjusted (often labelled resilients) made up of individuals who are well-functioning intellectually, in their achievements, emotionally, and interpersonally. The second type is less well-adjusted and overcontrolled (often labelled overcontrollers). These individuals tend to be shy, inhibited, restrained, inward-looking, and easily upset. The third type is less well-adjusted and undercontrolled (often labelled undercontrollers), who tend to be more impulsive, restless, distractible, disagreeable, and antagonistic and hostile. When ego-resiliency and egocontrol were measured in these studies, results ®tted with Block and Block's ideas concerning the curvilinear relation between ego-resiliency and egocontrol, that is, resilients scored high on ego-resiliency and intermediate on ego-control, whereas both overcontrollers and undercontrollers scored low on ego-resiliency, but low and high, respectively, on ego-undercontrol. These three personality ``types'' consistently show characteristic pro®les on the big-®ve personality factors. Resilients score moderately high on all

10. Lifespan personality development

187

big-®ve factors; the personality pro®les of overcontrollers and undercontrollers are mirror images of one another, that is, overcontrollers score low on extraversion and emotional stability, and average on agreeableness, conscientiousness, and openness. Undercontrollers score high on extraversion, particularly low on agreeableness, low on conscientiousness, and average on emotional stability and openness. Caspi (1998) concluded: ``At this point, these three types are good candidates to become an integral part of any generalisable personality typology'' (p. 322). Indeed, in the eight studies cited, these three types were generalisable across age (i.e., from 3 years to middle adulthood), gender, ethnic background (i.e., EuropeanAmerican and African-American adolescent boys), birth cohort (i.e., across samples born in the past century), geographical regions (i.e., North America, Northern Europe, and New Zealand), data sources (e.g., selfreport, parent report, teacher report, examiner observations, and clinical judgements), methods of personality assessment (e.g., Q-sorts, behaviour ratings, and questionnaires), and statistical procedures used in type derivation (i.e., inverted or Q-factor analysis and cluster analysis). More recent studies have supported and re®ned these conclusions. Cluster analysis on self-report big-®ve personality scale scores of 3284 adolescents (van Lieshout, van Aken, & Scholte, 1999) resulted in the same three basic personality types. Within each type, the big-®ve scale scores were con®gured similarly to the patterns found in the earlier studies (Robins, John, Caspi, Mof®tt, & Stouthamer-Loeber, 1996; van Lieshout, Haselager, Riksen-Walraven, & van Aken, 1995). Elsewhere, van Lieshout, Haselager, Scholte, and van Aken (1998) compared two statistical procedures for derivation of personality types. The ®rst typology consisted of an inverse or Q-factor analysis (cf. Robins et al., 1996) on 100 California Child Q-set items (CCQ; Block & Block, 1980) and the second of a cluster analysis (cf. Caspi & Silva, 1995) on big-®ve scale scores derived from the same CCQs (cf. van Lieshout & Haselager, 1994). Generally, cluster and inverted Q-factor analysis as methods for type derivation resulted in very similar typologies; despite substantial similarity, however, considerable amounts of variability are found, leaving room for developmental change and variability in the manifestation of personality in various settings. Asendorpf and van Aken (1999) applied separate inverted or Q-factor analysis on teacher Q-sorts with a longitudinal sample of 151 preschoolers (at the ages 4 to 6 years) and again on parent Q-sorts of 124 of these same children at age 10, resulting in three replicable personality pro®les. Comparison with similar factor pro®les found in North American, Dutch, and Icelandic children (Robins et al., 1996; Hart, Hofman, Edelstein, & Keller, 1997; van Lieshout et al., 1995) revealed that the personality pro®les in this study could be unambiguously interpreted as resilient, overcontrolled, and undercontrolled. Relating the three Q-types at both earlier ages with parental ratings of the children at age 12 showed that similar big-®ve personality pro®les were predicted as in earlier studies. High continuity of the

188

van Lieshout

personality pro®les was found across age, despite different judges. Moreover, highly signi®cant stability of Q-type membership was demonstrated between the ages of 4±6 and 10. At the same time, however, many children changed their classi®cations between pre-school and middle childhood. ``Temperamental'' types have also been identi®ed. Using a different statistical procedure (i.e., con®gural frequency analysis), Aksan et al. (1999) revealed two small ``temperamental'' types in 488 children, aged 3 years and 6 months, on the basis of a broad set of temperamental characteristics, including positive and negative emotionality and the attentional and behavioural control domains. The con®guration of temperamental characteristics, as well as a wide range of problem behaviours, showed the controllednonexpressive temperamental type to be similar to the overcontrolled personality type and the noncontrolled-expressive type to be similar to the undercontrolled personality type. Both types could be predicted very consistently on the basis of several infant and toddler temperamental characteristics as early as 1 month after birth. A moderate degree of stability in categorical standing of the children for these types could be detected after a year. Aksan and co-workers did not ®nd a third type, comparable to the resilients. Because resilience is closely linked with self-regulatory temperament, relevant characteristics may not be suf®ciently differentiated in toddlers' behaviour to be visible for caretakers. Moreover, identifying a subcategory of resilients may have been dif®cult because of the extremely small categories established in this investigation. In sum, Caspi's (1998) hypothesis concerning a generalisable three-type personality typology has been supported extensively in personality as well as in temperament research. The three types demonstrate considerable continuity across age from a very early age onwards but also substantial variability, thus, manifesting both developmental stability as well as change. Early antecedents of the three types are found in longitudinal studies and, in turn, the three types predict differential adjustment in problem-solving behaviour after several years.

Personality subtypes The three-personality-type model is rather undifferentiated. For example, the numbers of individuals marked by problem behaviours, that is, internalising problems in overcontrollers and externalising problems in undercontrollers, seem far too large for samples from normal populations. Therefore, the three-type model asks for more subtle typological research. In three studies in children and adolescents (Robins, John, & Caspi, 1998; van Lieshout et al., 1995; van Lieshout et al., 1998), each of the three main types were subdivided into two subtypes. That is, resilients were subdivided into communal resilients and agentic resilients; overcontrollers into achieving overcontrollers and vulnerable overcontrollers; and undercontrollers into impulsive undercontrollers and antisocial undercontrollers. Each of the two

10. Lifespan personality development

189

subtypes was marked by only slightly different personality pro®les, but substantially different adjustment patterns. Agency refers to the striving of individuals to separate from others, to be autonomous and independent, to master the environment, and to assert, protect, and expand the self. Communion refers to the individual's striving to lose one's own individuality by merging with others in social harmony, by being interdependent, by participating in social networks which are larger than the self, and by relating to others in a warm, close, intimate, and loving manner (Bakan, 1966). The communal and agentic resilient children and adolescents were above average on all big-®ve personality dimensions. Compared to the communal resilients, the agentic resilients were more extravert, more emotionally stable, and more open to new experience, whereas the communal resilients showed higher agreeableness and conscientiousness. Both resilient subtypes appeared to be well adjusted, but they differed in a number of adjustment domains, resulting in two distinguishable behavioural pro®les. The agentic resilients scored highest on school achievement, presented themselves as more autonomous, but also as more risk-taking (i.e., they smoked and gambled more, used more drugs, and were more involved in antisocial behaviour). The communal resilients tended to be more prosocial and concerned than the agentic resilients. Achieving overcontrollers revealed average scores on most personality dimensions (except lower scores on extraversion) and average adjustment on various domains. Compared to the achieving overcontrollers, the vulnerable overcontrollers scored lower on emotional stability, but also on conscientiousness and openness to new experience; no differences were found on introversion and agreeableness. Overcontrollers in earlier studies (cf. Caspi, 1998) were marked by internalising problems, but these ®ndings show that achieving overcontrollers are high-achieving, reliable, and somewhat withdrawn individuals without internalising problems, whereas vulnerable overcontrollers, in contrast, reveal high levels of internalising problems that are not noticed by their peers. Compared to the total sample, the impulsive undercontrollers scored clearly above average on extraversion, just below average on agreeableness, far below average on conscientiousness, average on emotional stability, and below average on openness. In contrast, the antisocial undercontrollers scored lower on extraversion, far lower on agreeableness, and higher on conscientiousness, above average on emotional stability, and about average on openness. Except for agreeableness, the antisocial undercontrollers were better able to control their impulses than the impulsive undercontrollers. Patterns of adjustment behaviour among both subtypes of children and adolescents were in accordance with these different personality patterns. The antisocial undercontrollers were aggressive and delinquent and perceived very low relational support and high peer rejection, whereas the impulsive undercontrollers were not very well able to monitor their impulses and had very low school achievement, but were not particularly

190

van Lieshout

antisocial. That is, impulsive undercontrollers revealed fewer externalising problems compared to antisocial undercontrollers and maintained effective relationships as could be seen from their perceived relational support and the peer evaluations by their classmates. In addition to more externalising problems, antisocial undercontrollers reported con¯icting relationships at home and with best friends as well as in evaluations by classmates. In sum, the personality subtypes reveal slightly different personality pro®les, but very distinctive adjustment patterns that seem highly similar across middle childhood, adolescence, and adulthood. These patterns of (mal)adjustment seem to vary in accordance with different problem-solving domains that mark the various societal roles of males and females at different ages. MALE AND FEMALE PERSONALITY AND ADJUSTMENT Across age, the personality pro®les revealed in males and females as well as their adjustment patterns seem rather similar, suggesting similar developmental patterns. The numbers of males and females, however, in different personality types and subtypes are far from equal. Thus, similar developmental personality patterns may occur in males and females, but many females demonstrate a very different personality development from males. Generally, the distribution of individuals across the three main personality types is not equal, despite the tendency of Ward's cluster analysis (cf. Caspi & Silva, 1995) to divide samples in clusters of nearly equal size. Ego-resilients tend to be more numerous than the other types in most recent studies. For example, when aggregated across a number of studies (van Lieshout et al., 1998), about 45% of the individual participants were resilients, about 35% were overcontrollers, and about 25% were undercontrollers, although these percentages vary depending on the sample. Moreover, in typological studies, gender by age interactions are commonly reported, even though, in trait-centred studies, average gender differences are generally small (Feingold, 1994; McCrae et al., 1999). Gender differences are even more striking and far from equal among the different personality subtypes. Again, frequencies differ at various ages. In pre-school and elementary school (Asendorpf & van Aken, 1999; van Lieshout et al., 1995) girls are over-represented among resilients but underrepresented among undercontrollers; girls were particularly over-represented in the communal resilient subtype. Among adolescents (van Lieshout et al., 1998) the numbers of resilient boys and girls are nearly equal. Now, the agentic resilient adolescents were mainly boys (73%), whereas the communal resilient adolescents were mainly girls (60%). These age differences are not easy to interpret. The higher proportion of resilient girls observed in elementary school may be due to the tendency of teachers and parents to evaluate well-adapted, well-balanced, achievement-oriented, sociable, and

10. Lifespan personality development

191

agreeable girls as resilient. The evaluations of adolescents were based on self-descriptions, which may be coloured by a less steady self-concept and gender stereotypes. Moreover, among adolescents the onset of puberty in girls, on average, is 2 years earlier than in boys, and is more often related to emotional turmoil and social turbulence than in boys (Caspi, Lynam, Mof®tt, & Silva, 1993) and with ambiguous relationships with older boys (Stattin & Magnusson, 1990). In contrast, boys frequently welcome the increased maturity, physical growth, strength, and relative independence associated with the early onset of puberty (Simmons & Blyth, 1987), and this may promote a more assertive and autonomous personality pro®le in resilient boys and a more boisterous personality pro®le among undercontrollers. Moreover, such changing differences in gender identity and gender role may become more visible in the more speci®c personality pro®les of the subtypes than in the more general personality pro®les of the three-type model, let alone in separate traits. The distribution of males and females among overcontrollers was similar in children and adolescents. In adolescence, girls were somewhat overrepresented among the overcontrollers. However, for the achieving subtype, the numbers of boys and girls are nearly equal at both ages, whereas about two-thirds of the vulnerable overcontrollers are girls and one-third are boys. Among undercontrollers, boys outnumber girls by far, particularly in adolescence and within the antisocial undercontrolling subtype. About 71% of the undercontrollers are boys, but 85% of the antisocial undercontrollers are boys and 15% are girls. The different numbers of boys and girls in the personality subtypes in children and adolescents are also re¯ected in the adjustment patterns of adults (Pulkkinen, MaÈnnikoÈ, & Nurmi, 2000). In sum, the personality pro®les of children, adolescents, and adults seem rather similar but many males and females tend to follow different pathways in their personality development. DEVELOPMENTAL OUTCOME In this section, outcomes in various developmental domains over the life course will be associated with separate temperament and personality dimensions, and examined with respect to personality type and subtype. In earlier reviews, predictions were made from early temperament and personality to a great variety of later adjustment variables, such as internalising or externalising problem behaviour, as well as later outcome domains, such as quality of attachment (cf. Caspi, 1998; Rothbart & Bates, 1998). A somewhat different approach will be followed in this instance.

Four developmental domains A taxonomy of developmental domains is proposed. Similar to the dimensions of temperament and personality, each developmental domain is

192

van Lieshout

assumed to have continuity as well as variability across the life course. Each domain consists of a lifelong and thematically coherent, age-graded sequence of problems, demands, and opportunities (Baltes, 1987). Four developmental task domains are distinguished that every individual must master across the life course. First, the interpersonal domain concerns social competence in the formation and maintenance of various close dyadic relationships such as child± parent, sibling, friendship, partner, and parent±child relationships. Second, individuals across the life course must complete a sequence of performance tasks, such as mastery of the physical environment, acquisition of the mother tongue, pursuing school and professional careers, educating children, or providing care for dependent parents. Third, the acquisition and maintenance of a coherent sense of self is another developmental domain continuing across the life course for every individual. The development of gender identity and a sense of identity related to other biological, societal, and personal roles also ®ts into this domain. Violation of the sense of self results in emotional destabilisation in the ®rst place. Fourth, for many individuals, the realisation of intellectual, artistic, emotional, or creative talents are also continuous goals (van Lieshout & Heymans, 2000). These domains are sketched here as distinctive in terms of their goal orientations, but achievement (or failure) in each of these domains also affects adjustment in the other domains. To a certain degree, the four domains are mutually interdependent. For example, creative musical performance requires excellent technical musical skills (Subotnik, 2000) and moral wisdom requires high levels of moral reasoning (Olthof, 2000); that is, attainment of intellectual, artistic, emotional, and creative levels of excellence assume achievement of preliminary skills in the speci®c domain of excellence. Moreover, across the life course, individuals meet problems, demands, and opportunities in each of these domains in various degrees, depending on different historical time frames and cultural contexts, and whether they are males and females.

Meeting problems, demands, and opportunities The problems, demands, and opportunities that an individual faces at any time during his or her life can be assigned globally to the four domains mentioned. In order to solve problems, meet demands, and use opportunities an individual must be activated. Activation is a prerequisite for problem solving to take place. Activation is related to the activity level versus inhibition and the regulation of impulse, but may also be manifest in an individual's emotional experience of being challenged or constrained by fear and anxiety or anger and irritability. In addition, four self-regulatory and volitional problem-solving modalities are involved, that is, cognitions for the analysis of problems, demands, and opportunities, and for the generation of alternative solutions, planful and controlled execution of

10. Lifespan personality development

193

problem-solving behaviour, affect regulation for the control of emotions in problem solving, and goal orientation in selecting speci®c solutions. These problem-solving modalities are assumed to be speci®cally related to the big®ve personality dimensions, that is activation to extraversion, cognitions to openness, behaviour execution to conscientiousness, affect regulation to emotional stability, and goal orientation to one of each of the four selfregulatory big-®ve personality dimensions depending on the developmental domain (see Table 10.1; for an earlier version of this model, see van Lieshout, Cillessen, & Haselager, 1998). The four developmental domains seem speci®cally to be associated with each of the four self-regulatory big-®ve dimensions (see Table 10.1, bottom line). Agreeableness concerns mainly the interrelatedness of one's own interests with those of one's partners and thus refers to the interpersonal domain. Conscientiousness concerns the attainment of performance standards and thus refers to the domain of achievement performances. Emotional stability concerns the regulation of emotions and is particularly at stake in the development and maintenance of a coherent self-concept and self-esteem. Openness to new experience concerns the ¯exibility of information processing and particularly relates to the domain of intellectual, artistic, emotional, and creative excellence. The four developmental domains are not unitary. Facets of one domain are intertwined with facets of the others. It is far beyond the scope of this chapter to outline in detail the diversity of the four domains and to relate them to temperament and personality development. Except for ®ndings concerning relations between personality and developmental outcomes, empirical data are scarce, whereas a multiplicity of complex relations must be assumed. For example, Rothbart and Bates (1998) distinguished 14 different processes linking temperament and personality to developmental outcomes. Most relevant ®ndings have been reported regarding the interpersonal domain. Kochanska and her co-workers related the ``moral'' agency of infants and toddlers to early development of temperament. Even at the early age of 13±15 months, there is a distinction between committed and situational compliance as precursors of early conscience development (Kochanska, Tjebkes, & Forman, 1998). Infants' committed compliance, in contrast to situational and supervised compliance, may be considered an early form of the internalisation of maternal prohibitions, values, and rules and was correlated with infants' temperamental restraint as an indicator of self-regulation a few months earlier. Moreover, committed compliance was related to different maternal socialisation strategies for mothers of toddlers who tended to anger and fear (Kochanska, 1993), supporting the idea of a curvilinear relation between reactivity and self-regulation. Continuity of other-oriented prosocial interactions was further related to personality development in a longitudinal study extending between 4±5 years and early adulthood. Prosocial behaviours that did not re¯ect an

Problem-solving modality

Activation vs. passivity

Cognition

Executive regulation

Affect regulation

Goal orientation

Personality dimension

Extraversion

Openness

Conscientiousness

Emotional stability

Four big-®ve dimensions

Performance

Initiative vs. passivity

Expressive vs. reserved

Self

Attainment of standard (Conscientiousness)

(Agreeableness)

Failure vs. success; proud vs. shame

Controlled, planful action vs. sloppiness

Cognitive problem analysis

Comparison of interests

Trust vs. mistrust; positive affect vs. fear or anger

Cooperation, altruism, proactive aggression

Social cognition

(Emotional stability)

Balance of self and ideal self

Emotional balance vs. vulnerability

Exploration of self

Re¯ection about self

Speci®c modalities of self-regulated problem solving

Sociability vs. withdrawal

Facets of activation of problem solving

Interpersonal

Developmental domains

(Openness)

Degree of creative excellence

Aesthetic euphoria vs. disgust

Intellectual, artistic, creative performance

Creative problem analysis

Intellectual energy vs. dull

Intellectual; creative

Table 10.1 The big-®ve personality dimensions, activation of problem solving, and self-regulated problem solving in four developmental domains

194 van Lieshout

10. Lifespan personality development

195

other-orientation, such as accidental helping and compliance, generally did not predict later prosocial behaviour (Eisenberg et al., 1999). Such results suggest that individual differences in early self-regulated conscience development in infants and toddlers may be linked to a sympathy-mediated prosocial disposition as revealed in consistent prosocial behaviour from early childhood into early adulthood (cf. Kochanska et al., 1998), consistent with the idea that agreeableness and prosocial development are related (Graziano, 1994; Graziano, Jensen-Campbell, & Hair, 1996). Using seven assessments across 18 months, Asendorpf and Wilpers (1998) predicted development in the early adult years of diverse aspects of students' relationships after entry to university from various speci®c big-®ve personality dimensions. Controlling for the initial correlation between personality and relationship quality, highly stable personality affected involvement in smooth opposite-sex relationships, but not vice versa. Even strong changes in relationships did not affect changes in personality. Extraversion (sociability, shyness) was related to size of the social network and interaction rate in close opposite-sex relationships; network size and interaction rate refer to the degree of activation. Agreeableness was associated with the quality of these relationships, especially lowered con¯ict with opposite-sex peers, and refers to the quality of this age-graded interpersonal problem solving. Conscientiousness predicted high involvement in family relationships and, in particular, was associated with successful maintenance of long-term relationships. Conscientiousness refers to compliance with social standards in already existing relationships. Lastly, emotional stability and openness did not interfere with the development of relationships; according to our model, these two personality dimensions are not at stake in interpersonal problem solving. In sum, these ®ndings support our model concerning the speci®c relation between personality and interpersonal problem solving of these university students as goal-oriented agents. Developmental outcome is often measured as a single output measure and growth curve modelling is rarely used (for exceptions, however, see Asendorpf & van Aken, 1999; Hart et al., 1997). In their longitudinal study, Asendorpf and van Aken (1999) predicted growth curves based on six measurements between 4 and 12 years of four developmental outcome variables for the three personality types of resilient, overcontrolling, and undercontrolling preschoolers. Their outcome variables conform closely to the developmental domains speci®ed in this chapter: the ®rst, cognitive competence, concerns aspects of school performance and intellectual functioning (IQ); the second, inhibition, is very much related to activation; the third, aggression, refers to the interpersonal domain; and the fourth, selfesteem, taps directly into the self-domain. Findings show consistent differences between the three pre-school personality types in the four domain trajectories across the elementary school years as expected. Summarising, the evidence links descriptive personality assessments as well as speci®c types of problem solving to developmental domains. Most

196

van Lieshout

research concerns the interpersonal domain. Information on the relation between personality and performance, the self, and creative domains is very much lacking.

CONCLUSIONS AND FUTURE DIRECTIONS In the years to come, a number of issues touched upon in this chapter must be clari®ed. First, in the past few decades, our greatest achievement has been that congruence has been established between temperament and personality description and among various theoretical models for the study of temperament and personality (Halverson et al., 1994; Rothbart & Bates, 1998; Shiner, 1998). Separate dimensions have been identi®ed describing temperament, on the one hand, and personality, on the other, revealing continuities across the life course (cf. Lemery et al., 1999). Similarly, support is increasing for a ®ve-factor model for describing personality development in children, adolescents, and adults. However, the similarity among various schemes for the description and functioning of temperament and personality needs to be better documented. Second, the curvilinear relation between activation and impulse control (i.e., reactivity, ego-undercontrol, extraversion), on the one hand, and ¯exible, resourceful, and volitional regulation (i.e., self-regulation or egoresiliency), on the other, provides a model for understanding personality functioning in problem solving in both early temperament and later personality. Con®rmation of these assumptions is a major research agenda on its own. Third, although the distinction between the agent-centred and the typological person-centred approaches can be traced back nearly a century to the ideas of Stern (1911, 1923), this distinction must be conceptually and empirically clari®ed. The distinction may be smaller than suggested in this chapter. The personality pro®les and the variety of associated adjustment patterns may be just descriptions of different types of agents, including differences in personality between males and females and their different orientations on the various developmental domains. The type-centred as well as the agent-centred approach has hardly been studied in infants or in longitudinal examination. Fourth, in contrast to the structure of temperament and personality development, developmental outcomes are rather unexplored. Most outcome studies include a large number of unorganised outcome variables, loosely assigned to developmental domains. In this chapter, an ``outcome sketch'' has been provided, consisting of four distinctive developmental domains, mainly marked by distinctive goal orientations. But these assumptions concerning the development of personality and developmental outcomes are rather speculative thus far and await empirical testing.

10. Lifespan personality development

197

ACKNOWLEDGEMENT I wish to thank Willard W. Hartup, Marcel A.G. van Aken, and my wife and colleague psychologist Irene A.A. van Lieshout-Rammeloo. Their stimulating comments, suggestions, and editorial comments have helped me to express my ideas more clearly. REFERENCES Ahadi, S.A., Rothbart, M.K., & Ye, R.M. (1993). Child temperament in the US and China: Similarities and differences. European Journal of Personality, 7, 359±378. Aksan, N., Goldsmith, H.H., Smider, N.A., Essex, M.J., Clark, R., Hyde, J.S., Klein, M.H., Vandell, D.L. (1999). Derivation of temperamental types among preschoolers. Developmental Psychology, 35, 958±971. Allport, G.W. & Odbert, H.S. (1936). Trait-names: A psycho-lexical study. Psychological Monographs, 47 (1, Whole No. 211). Asendorpf, J.B., & van Aken, M.A.G. (1999). Resilient, overcontrolled, and undercontrolled personality prototypes in childhood: Replicability, predictive power, and the trait-type issue. Journal of Personality and Social Psychology, 77, 815± 832. Asendorpf, J.B., & Wilpers, S. (1998). Personality effects on social relationships. Journal of Personality and Social Psychology, 74, 1531±1544. Bakan, D. (1966). The duality of human existence: Isolation and communion in Western man. Boston: Beacon. Baltes, P.B. (1987). Theoretical propositions of life-span developmental psychology: On the dynamics between growth and decline. Developmental Psychology, 23, 611±626. Bem, D.J., & Allen, A. (1974). On predicting some of the people some of the time: The search for cross-situational consistencies in behaviour. Psychological Review, 81, 506±520. Block, J. (1961). The Q-sort method in personality assessment and psychiatric research. Spring®eld, IL: Thomas. Block, J. (1971). Lives through time. Berkeley, CA: Bancroft Books. Block, J. (2000). Three tasks for personality psychology. In L.R. Bergman, R.B. Cairns, L.-G. Nilson, & L. Nystedt (Eds.), Developmental science and the holistic approach (pp. 154±164). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Block, J.H., & Block, J. (1980). The role of ego-control and ego-resiliency in the organisation of behaviour. In W.A. Collins (Ed.), The Minnesota Symposia on Child Psychology, Vol. 13 (pp. 39±101). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Caspi, A. (1998). Personality development across the life course. In W. Damon (Series Ed.), N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (5th ed., pp. 311±388). New York: John Wiley. Caspi, A., Lynam, D., Mof®tt, T.E., & Silva, P.A. (1993). Unraveling girls' delinquency: Biological, dispositional, and contextual contributors to adolescent misbehaviour. Developmental Psychology, 29, 19±30.

198

van Lieshout

Caspi, A., & Silva, P.A. (1995). Temperamental qualities at age 3 predict personality traits in young adulthood: Longitudinal evidence from a birth cohort. Child Development, 66, 486±498. Cloninger, C.R. (1987). A systematic method for clinical description and classi®cation of personality variants: A proposal. Archives of General Psychiatry, 44, 573±588. Costa, P.T., Jr., & McCrae, R.R. (1989). The NEO PI/FFI manual supplement. Odessa, FL: Psychological Assessment Resources. Costa, P.T., Jr., & McCrae, R.R. (1994). Stability and change in personality from adolescence through adulthood. In C.F. Halverson, G.A. Kohnstamm, & R.P. Martin (Eds.), The developing structure of temperament and personality from infancy to adulthood (pp. 139±150). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Digman, J.M. (1990). Personality structure: Emergence of the ®ve-factor model. Annual Review of Psychology, 41, 417±440. Eisenberg, N., Fabes, R.A., Guthrie, I.K., & Reiser, M. (2000). Dispositional emotionality and regulation: Their role in predicting quality of social functioning. Journal of Personality and Social Psychology, 78, 136±157. Eisenberg, N., Guthrie, I.K., Murphy, B.C., Shepard, S.A., Cumberland, A., & Carlo, G. (1999). Consistency and development of prosocial dispositions: A longitudinal study. Child Development, 70, 1390±1372. Endler, N.S., & Magnusson, D. (1976). Toward an interactional psychology of personality. Psychological Bulletin, 83, 956±979. Eysenck, H.J. (1954). The science of personality: Nomothetic! Psychological Review, 61, 339±342. Eysenck, H.J. (1967). The biological basis of personality. Spring®eld, IL: Thomas. Feingold, A. (1994). Gender differences in personality: Meta-analysis. Psychological Bulletin, 116, 429±456. Fox, N.A., Henderson, H.A., Rubin, K.H., Calkins, S.D., & Schmidt, L.A. (2001). Continuity and discontinuity of behavioral inhibition and exuberance: Psychophysiological and behavioral in¯uences across the ®rst four years of life. Child Development, 72, 1±21. Goldberg, L.R. (1990). An alternative ``description of personality'': The Big-Five factor structure. Journal of Personality and Social Psychology, 59, 1216±1229. Goldsmith, H.H., Lemery, K.S., Buss, K.A., & Campos, J.J. (1999). Genetic analyses of focal aspects of infant temperament. Developmental Psychology, 35, 972±985. Graumann, C.F. (1959). Eigenschaften als Problem der Persoenlichkeitsforschung [Dispositions as a problem in personality research]. In Ph. Lersch, F. Sander, & H. Thomae (Series Eds.), Ph. Lersch & H. Thomae (Vol. Eds) Handbuch der Psychologie: Vol. 4. Persoenlichkeitsforschung und Persoenlichkeitstheorie. Goettingen: Hogrefe. Gray, J.A. (1975). Elements of a two-process theory of learning. New York: Academic Press. Graziano, W.G. (1994). The development of agreeableness as a dimension of personality. In C.F. Halverson, Jr., G.A. Kohnstamm, & R.P. Martin (Eds.), The developing structure of temperament and personality from infancy to adulthood (pp. 339±354). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Graziano, W.G., Jensen-Campbell, L.A., & Hair, E.C. (1996). Perceiving interper-

10. Lifespan personality development

199

sonal con¯ict and reacting to it: The case for agreeableness. Journal of Personality and Social Psychology, 70, 820±835. Guilford, J.P. (1959). Personality. New York: McGraw. Halverson, C.F. Jr., Kohnstamm, G.A., & Martin, R.P. (Eds.) (1994). The developing structure of temperament and personality from infancy to adulthood. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Hart, D., Hofmann, V., Edelstein, W., & Keller, M. (1997). The relation of childhood personality types to adolescent behavior and development: A longitudinal study of Icelandic children. Developmental Psychology, 33, 195±205. Hartup, W.W., & van Lieshout, C.F.M. (1995). Personality development in social context. Annual Review of Psychology, 46, 655±687. Hathaway, S.R., & McKinley, J.C. (1943). MMPI Manual. New York: Psychological Corporation. Hofstee, W.K., de Raad, B., & Goldberg, L.R., (1992). Integration of the Big Five and circumplex approaches to trait structure. Journal of Personality and Social Psychology, 63, 146±163. Kochanska, G. (1993). Toward a synthesis of parental socialization and child temperament in early development of conscience. Child Development, 64, 325± 347. Kochanska, G., Tjebkes, T.L., & Forman, D.R. (1998). Children's emerging regulation of conduct: Restraint, compliance, and internalization from infancy to second year. Child Development, 69, 1378±1389. Kohnstamm, G.A., Halverson, C.F., Jr., Mervielde, I., & Havill, V.L. (1998). Parental descriptions of child personality. Developmental antecedents of the Big Five? Mahwah, NJ: Lawrence Erlbaum Associates Inc. Lemery, K.S., Goldsmith, H.H., Klinnert, M.D., & Mrazek, D.A. (1999). Developmental models of infant and childhood temperament. Developmental Psychology, 35, 189±204. Loehlin, J.C. (1992). Genes and environment in personality development. Newbury Park, CA: Sage. Magnusson, D., & Endler, N.S. (1977). Interactional psychology: Present status and future prospects. In D. Magnusson & N.S. Endler (Eds.), Personality at the crossroads: Current issues in interactional psychology (pp. 3±31). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Magnusson, D., & ToÈrestad, B. (1993). A holistic view of personality: a model revisited. Annual Review of Psychology, 44, 427±452. McCrae, R.R., & Costa, P.T. Jr. (1987). Validation of the ®ve-factor model of personality across instruments and observers. Journal of Personality and Social Psychology, 52, 81±90. McCrae, R.R., Costa, P.T. Jr., Ostendorf, F., Angleitner, A., Hrebickova, M., Avia, M.D., et al., (2000). Nature over nurture: Temperament, personality, and lifespan development. Journal of Personality and Social Psychology, 78, 173±186. McCrae, R.R., Costa, P.T., Pedroso de Lima, M., Simoes, A., Ostendorf, F, Angleitner, A. et al. (1999). Age differences in personality across the adult life span: Parallels in ®ve cultures. Developmental Psychology, 35, 466±477. Mischel, W. (1968). Personality and assessment. New York: John Wiley. Mischel, W., Shoda, Y., & Peake, P.K. (1988). The nature of adolescent competencies predicted by preschool delay of grati®cation. Journal of Personality and Social Psychology, 54, 687±696.

200

van Lieshout

Murphy, G. (1947). Personality: a biosocial approach to origins and structure. New York: Harper. Olthof, T. (2000). The morality paradox: Choosing not to be moral as a component of moral excellence. In C.F.M. van Lieshout & P.G. Heymans (Eds.), Developing talents across the life span (pp. 125±149). Hove, UK: Psychology Press. Plomin, R., & Rende, R. (1991). Human behavioral genetics. Annual Review of Psychology, 42, 161±190. Pulkkinen, L., MaÈnnikoÈ, K., & Nurmi, J.E. (2000). Self-description and personality styles. In L.R. Bergman, R.B. Cairns, L.-G. Nilson, & L. Nystedt (Eds.), Developmental science and the holistic approach (pp. 265±280). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Roberts, B.W., & DelVecchio, W.F. (2000). The rank-order consistency of personality traits from childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126, 3±25. Robins, R.W., John, O.P., Caspi, A., Mof®tt, T.E., & Stouthamer-Loeber, M. (1996). Resilient, overcontrolled, and undercontrolled boys: Three replicable personality types. Journal of Personality and Social Psychology, 70, 157±171. Robins, R.W., John, O.P., & Caspi, A. (1998). A typological approach to studying personality. In R.B. Cairns, L. Bergman, & J. Kagan (Eds.), Methods and models for studying the individual: Essays in honor of Marian Radke-Yarrow (pp. 135± 160). Thousand Oaks, CA: Sage. Rothbart, M.K. (1989). Temperament and development. In G.A. Kohnstamm, J.E. Bates, & M.K. Rothbart (Eds.), Temperament in childhood (pp. 187±247). Chichester, UK: John Wiley. Rothbart, M.K., Ahadi, S.A., & Evans, D.E. (2000). Temperament and personality: Origins and outcomes. Journal of Personality and Social Psychology, 78, 122±135. Rothbart, M.K., & Bates, J.E. (1998). Temperament. In W. Damon (Series Ed.), N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (5th ed., pp. 105±176). New York: John Wiley. Shiner, R.L. (1998). How shall we speak of children's personalities in middle childhood? A preliminary taxonomy. Psychological Bulletin, 124, 308±332. Simmons, R.G., & Blythe, D.A. (1987). Moving into adolescence: The impact of pubertal change and school context. New York: Aldine. Stattin, H., & Magnusson, D. (1990). Paths through life: Vol. 2. Pubertal maturation in female development. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Stern, W. (1911). Die differentielle Psychologie in ihren methodischen Grundlagen [Methodological foundations of differential psychology]. Leipzig, Germany: Ambrosius Barth. Stern, W. (1923). Die menschliche PersoÈnlichkeit [Human Personality] (3rd ed.). Leipzig, Germany: Ambrosius Barth. Subotnik, R. (2000). The Juilliard model for developing young adolescent performers: An educational prototype. In C.F.M. van Lieshout & P.G. Heymans (Eds.), Developing talents across the life span (pp. 249±276). Hove, UK: Psychology Press. Tellegen, A. (1991). Personality traits: Issues of de®nition, evidence, and assessment. In W.M. Grove & D. Cicchetti (Eds.), Thinking clearly about psychology: Vol. 2. Personality and psychopathology (pp. 10±35). Minneapolis, MN: University of Minnesota.

10. Lifespan personality development

201

Thomae, H. (1955). PersoÈnlichkeit. Eine dynamische Interpretation [Personality. A dynamical interpretation]. Bonn, Germany: Weinheim. van Lieshout, C.F.M., Cillessen, A.H.N., & Haselager, G.J.T. (1998). Interpersonal support and individual development. In W. A. Collins & B. Laursen (Eds.), Relationships as developmental contexts. The Minnesota Symposia on Child Psychology, Vol. 30 (pp. 37±60). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. van Lieshout, C.F.M., & Haselager, G.J.T. (1994). The big-®ve personality factors in Q-sort descriptions of children and adolescents. In C.F. Halverson, G.A. Kohnstamm, & R.P. Martin (Eds.), The developing structure of temperament and personality from infancy to adulthood (pp. 293±318). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. van Lieshout, C.F.M., & Heymans, P.G. (Eds.). (2000). Developing talents across the life span. Hove, UK: Psychology Press. van Lieshout, C.F.M., Haselager, G.J.T., Riksen-Walraven, J.M., & van Aken, M.A.G. (1995). Personality development in middle childhood. Paper presented at the Biennial Meetings of the Society for Research in Child Development, Indianapolis, USA, 1995. van Lieshout, C.F.M., Haselager, G.J.T., Scholte, R.H.J., & van Aken, M.A.G. (1998). Generalizability of personality types and subtypes across age, gender, and methods of assessment. Paper presented at the Biennial Meetings of the International Society for the Study of Behavioral Development, Berne, Switzerland, 1998. van Lieshout, C.F.M., van Aken, M.A.G., & Scholte, R.H.J. (1998). Adolescenten met verschillende persoonlijkheidstypen. Hun sociale relaties en hun psychosociaal functioneren. [Adolescents with different personality types. Their social relationships and psychosocial functioning.] Nederlands Tijdschrift voor Opvoeding, Vorming en Onderwijs, 14, 114±133. Watson. D., & Clark, L.A. (1993). Behavioral disinhibition versus constraint: A dispositional perspective. In D.M. Wegner & J.W. Pennebaker (Eds.), Handbook of mental control (pp. 506±527). Englewood Cliffs, NJ: Prentice Hall. Windelband, W. (1998). History and natural science. Theory and Psychology, 8, 5± 22. Zuckerman, M., Kuhlman, D.M., Joireman, J., Teta, P., & Kraft, M. (1993). A comparison of three structural models for personality: The Big Three, the Big Five, and the Alternative Five. Journal of Personality and Social Psychology, 65, 757±768.

11 Perspectives on gender development Eleanor E. Maccoby Stanford University, USA

In the last several decades there have been important shifts in psychologists' thinking about gender development. There were empirical questions ®rst of all: In what ways, to what degree, and how consistently, did boys and girls differ in the developmental pathways typically taken? And there have been notable changes in the points of view psychologists have brought to bear in their efforts to understand and explain whatever gender differentiation was thought to occur. Two viewpoints about gender development were dominant for many years: the socialisation perspective and the gender-cognitive perspective. These perspectives are ®rst described in this chapter, and then their limitations are pointed out, with stress on how narrow these views were concerning the nature of the gender-differentiated phenomena that need to be understood. The chapter turns then to considering how much an ethological perspective, when added to the traditional pair, can contribute towards achieving a more comprehensive view of gender development. Finally, some more recent thinking from psychobiology is brought to bear in the interests of moving towards an integration of the several perspectives. THE DIRECT SOCIALISATION PERSPECTIVE In the middle of the twentieth century, psychologists asked: By what processes do children become ``sex-typed?'' By sex-typing, they usually meant that children take on the attributes that are typical and/or valued (expected, normative) for their own sex. In seeking answers, they worked from the stimulus±response principles of the reinforcement learning theories that dominated the ®eld of psychology at that time. From this point of view, sex-typed behaviours were a set of habits. Boys and girls would develop different sex-typed habits if socialisation agentsÐparents, teachers, older childrenÐreinforced girls for ``feminine'' behaviours and provided negative consequences when they displayed behaviours thought to be more appropriate for boys. Similarly, boys were thought to be ``shaped'' toward the version of ``masculine'' behaviour deemed proper in the particular society where the children were growing up.

11. Perspectives on gender development

203

Implicit in this viewpoint is the idea that individual differences within each sex, and mean differences between the sexes on any given trait, are essentially re¯ections of the same processes. Thus it would be assumed that if one aspect of becoming masculine for a boy is to learn not to cry, then boys on the average would be subject to more socialisation pressuresÐi.e., would be told, more often than girls, ``That didn't really hurt,'' or ``It's only a scratch'' or ``Don't be a crybaby,'' or ``Oh, toughen up!''Ðand would develop stronger inhibitions against crying than would girls. At the same time, some boys would receive stronger, more consistent pressures of this kind than others, and so some would develop stronger crying inhibitions, and become more ``masculine'' than others. Differential socialisation pressures on boys and girls could take a variety of additional forms, beyond differential positive and negative reinforcement: for example, via providing different toys for children of the two sexes, or attributing different characteristics to them. And by giving boys and girls distinctive names and dressing them differently, their gender was announced to all comers with the message: ``This child is a boy/girl, to be treated accordingly.'' Distinctive socialisation pressures on boys and girls were thought to begin at birth, to be fairly strong, and to be consistent in some respects throughout a given culture. A secondary perspective which in¯uenced some of the developmental research at mid-century was derived from psychodynamic theory. This theory was more developmental than S±R learning theory, which did not specify particular periods during growth that would be optimal for children's acquisition of sex-typed behaviours. Freud's theory of psychosexual development posited a transition at about the age of 4 or 5, when the resolution of Oedipal con¯icts would involve children's identifying with their same-sex parent and thus taking on the appropriate sex-typed characteristics of that parent. This transition was thought to be more dif®cult for boys, who had to break a prior identi®cation with their mothers, and in becoming masculine would need to distance themselves from all things feminine. Echoes of this theory are found in the more recent work of Chodorow (1978). In the 1950s and 1960s, Robert Sears and colleagues (Sears, Maccoby, & Levin, 1957; Sears, Rau, & Alpert, 1965) attempted to integrate S±R and psychodynamic theories in their studies of the relationships between parental child-rearing practices and the degree of children's sex-typing. The results failed to con®rm Freudian theory. As Sears and colleagues wrote, in summarising their ®ndings: ``The box score for primary identi®cation theory as an explanation of gender role is poor.'' (Sears et al., 1965, p. 194). How well have the assumptions about children of the two sexes being socialised differently stood up empirically? It is certainly true that parents give their children sex-typed names, dress them differently, and decorate their rooms differently. Also, all languages or the world provide different ways of speaking about male and female persons. In English and other

204

Maccoby

Indo-European languages, the sexes are distinguished by pronounsÐhe, she, his, hersÐwhich surely facilitates children's learning to code themselves and others as to gender. In addition there are distinctive ways of speaking to children that emphasise stereotypic qualities, such as saying to a 4-year-old ``That's my sweet little girl'', or ``There's my big strong boy''. Still, it has proved surprisingly dif®cult to document differential treatment of boys and girls by their parents, especially when children are young. Several reviews have summarised child-rearing practices used by parents with sons and daughters (mainly in modern Westernised societies). When it comes to the traditional dimensions of child-rearing (e.g., permissiveness, restrictiveness, monitoring, responsiveness, warmth) few differences have been found in the way parents deal with sons as compared with daughters (Maccoby & Jacklin, 1974; Huston, 1983; Lytton & Romney, 1991). Although it is often assumed that parents react differently to assertive behaviour by sons and daughters, and draw young daughters into greater emotional closeness with themselves than they do with young sons, evidence to date does not support these assumptions. There are some ways in which parents do consistently differentiate: They do more roughhousing with sons, offer dolls more often to girls (and toy trucks to boys), and talk about feelings more with girls. (See Maccoby, 1998, for summary of studies.) And fathers in particular show negative reactions to any behaviour by their sons that seems effeminate. Of course, when socialisation differences are found, the ubiquitous issue of direction of effects arises. Do parents offer dolls to girls, trucks to boys, because they want their children to be appropriately ``masculine'' or ``feminine'', or because they have discovered that these are the toys the children prefer? If they roughhouse more with boys, is this because they have a stereotypical view of what kind of play boys ought to like, or because boys actually do like it more than girls and ``train'' their parents over time to play in ways that boys ®nd most enjoyable? There is good evidence that boys' and girls' different initiatives can indeed evoke different reactions from their parents. (See summary in Maccoby, 1998.) But these reactions, in their turn, can then in¯uence the children, so the existence of child-to-parent effects do not by any means preclude parent-to-child effects (Ge et al., 1996; O'Connor, Deater-Deckard, Fulker, Rutter, & Plomin, 1998). Experimental studies have been done in which unfamiliar children are given an arbitrary gender label. In such studies, an infant is chosen whose sex cannot easily be identi®ed when the child is dressed. Unfamiliar adults are then offered the opportunity to interact with the infant, or view the infant on videotape, the child having been introduced to some of the subjects as a girl, to others as a boy. The adults' reactions to the child, and interpretations of the child's behaviour, are then recorded. Such studies control for the direction-of-effects problem, since differential eliciting properties of male and female infants are ruled out by design. Whereas

11. Perspectives on gender development

205

early studies suggested that adults' reactions and interpretations were indeed in¯uenced by the child's gender label, a review of 23 gender labelling studies (Stern & Karraker, 1989) found that overall effects were quite weak and quite inconsistent from study to study. No doubt there are subtle differences in parental treatment that are not captured in the rather coarse-grained net that researchers have cast, and certainly what seem to be rather minor differences can accumulate over many repetitions into signi®cant in¯uences on children. And surely, socialisation differences would be more apparent in traditional societies where there are more rigid status differences between men and women. Still, we would have to say that to date, the socialisation theory that grew out of the S±R learning perspective of the mid-century and stressed the role of parents as ``shapers'' of sex-typed behaviour rests on a weak empirical foundation. Of course, a socialisation perspective does not need to focus so exclusively on parents as the developmental psychology of the 1950s, 1960s and 1970s tended to do. There were a number of reasons for this focus. First of all, much of the research at that time involved infants and children of preschool age, in families in which few mothers worked outside the home. Thus, it was natural to assume that parents were indeed the most in¯uential socialisation agents. In addition, there was a pervasive assumption that the early years were a time of great plasticity, when children were especially subject to ``shaping'' with respect to characteristics thought to be pervasive and long-lasting, such as the gendered aspects of the self. In recent years much more attention has been given to the in¯uence of other socialisation agents, such as out-of-home caregivers, peers, teachers, and coaches, whose positive or negative reactions to children's behaviour can provide additional shaping for children's sex-role development in ways that can supplementÐor sometimes even contradictÐthe in¯uence of parents. Actually, to test a theory of direct socialisation adequately, it would be necessary not only to demonstrate that socialisation agents deal differentially with the two sexes, but that these different socialisation experiences are related to any differences in developmental trajectories that appear in male and female children. Much depends, then, on what aspects of gender enactment are chosen as ``outcomes'' in efforts to test socialisation theory. This issue will be considered more fully below. THE INDIRECT SOCIALISATION PERSPECTIVE Beginning in the 1960s, social learning theory added learning by imitation (often called modelling) to reinforcement as a powerful process involved in gender socialisation. In social learning theory, children were still seen as being shaped by direct positive and negative reinforcement. But it was also shown that children could learn vicariously from seeing how other children's gender-appropriate or gender inappropriate behaviour were reacted

206

Maccoby

to by others (Bandura, 1965). Apart from learning about the consequences of such behaviour, children could also learn by observation what behaviours were characteristic of each sex. And of course, they could learn these things not only from what they observed within their families, schools, and neighbourhoods, but from ®lms or TV, or as depicted in stories (Mischel, 1966). The introduction of observational learning gave great added power to learning theory. But it also brought some complications. In daily life, boys and girls are exposed to models of both sexes. Both boys and girls presumably would learn the same facts concerning what behaviour is appropriate for boys, what for girls. The theory called for selective imitation, such that boys would adopt behaviour depicted by male models, or adapt their own behaviour according to the reinforcement patterns they saw being provided for male, rather than female, children. At the least, this required that children would know their own gender and that of the people whom they observed. Then too, it required that children should be able to summarise and generalise from multiple exemplars, and deal with exceptions. In addition, it called for some motivation to adopt the behaviour patterns of people who are ``the same as me'' with respect to gender. Clearly, the theory called for the extensive incorporation of cognitive elements in gender development. THE COGNITIVE PERSPECTIVE In the 1970s and 1980s, there was a strong surge of interest in gender cognitions. Although a cognitive perspective on developmental processes had been strongly anticipated in Europe for many decades, it was the ``cognitive revolution'' that took place in American psychology beginning in the 1960s that set the stage for an active revision of American views about child development. Children were increasingly seen not as passive recipients of socialisation pressures, but as active selectors and users of information pertinent to their developmental levels and personal goals. And children began to be seen as developing the capacity to adopt standards, and regulate their own behaviour in conformity to these standards, thus contributing to their own socialisation. Vast amounts of information are available to children concerning the way gender is enacted in the world around them. In the 1970s and 1980s, and continuing into the 1990s, research focused heavily on how children acquire knowledge and develop stereotypes and scripts concerning what is usual, or considered ``appropriate'', for people of the two sexes. (See Ruble & Martin, 1998, for a review.) It became evident that gender is a highly salient category for children, perhaps because it is neatly binary, because it is so heavily culturally emphasised, and because socially ascribed sex and biological sex are so completely redundant. The distinction, in other words, is easy to make, and there is good evidence that children do indeed make it

11. Perspectives on gender development

207

very early in life. Gender categories, once applied, have been shown to be a convenient hook on which children can easily hang stereotypes about gender attributes (Gelman, Coleman, & Maccoby, 1986) and assimilate new incoming information to these stereotypes. Gender schema theories introduced in the early 1980s held that children form cognitive structures that organise their gender knowledge into a set of expectations that guide and organise their social perceptions (Bem, 1981; Martin & Halverson, 1981). Knowledge about the characteristics of the two sexes, and about the norms for their behaviour, is clearly necessary, but is it suf®cient by itself to motivate children to adopt socially prescribed roles and standards? Several hypotheses have been advanced concerning the ways in which children's knowledge of their own gender identity and that of others could function to set motivational processes in motion: 1

2

3

When observing the contingencies that are experienced by other children, the observer is able to select the experiences of children of their own sex as most relevant to inferences about what might happen to themselves. This allows observational learning to be focused speci®cally on the acquisition of behaviours and standards that apply differentially to the child's own sex. Out of a need for cognitive consistency, children want to adapt themselvesÐthat is, conformÐto what they believe is appropriate for their own sex. Kohlberg (1966) proposed that this motivation would not set in until approximately the age of 5±7, when he thought children achieve a ®rm level of gender constancy. Since Kohlberg urged the importance of gender constancy, evidence has accumulated that the functional elements of gender constancy (namely, identity and stabilityÐsee Ruble & Martin, 1998; Maccoby, 1990) are achieved at considerably younger ages than Kohlberg believed. When they have achieved a stable gender identity, children classify themselves as members of a same-sex group. They identify with this group, see members of the other sex as belonging to an out-group, and want to be like members of their own-sex group.

These possibilities are not mutually exclusive, and may all combine to generate the motivation for children to (a) selectively imitate same-sex models (if they are known to be good exemplars of their gender categoryÐ see Perry & Bussey, 1979); (b) seek, select, and remember preferentially information that is relevant to, and consistent with, children's own-sex schemas, and (c) reject, ignore, forget, or distort schema-inconsistent material. (See Ruble and Martin, 1998, for a summary.) With the acquisition of a stable gender identity, children can also begin to monitor their own behaviour with reference to a self-accepted standard of what is appropriate for their own sex (Bussey & Bandura, 1992).

208

Maccoby

The socialisation and cognitive perspectives discussed so far combine into a kind of social constructivist approach. In so far as children of the two sexes are found to differ in their behaviour, interests, and/or values, it is thought that this differentiation comes about for three reasons: because adults shape children in this way, because peers shape each other according to the way they themselves have been socialised, and because childrenÐ once they have established a ®rm gender identityÐsocialise themselves to conform to what they know to be stereotypical for children of their own sex, within the limits of what their own competencies permit (Bussey & Bandura, 1999). Sometimes the use of the word ``stereotypes'' is taken to mean that attributes assimilated to social categories are arbitrary, so that our concepts about the two sexes may be quite arti®cial or distorted. How accurate are our gender stereotypes? In their review, Deaux and LaFrance (1998) note that it is hardly possible to check the reality base of some of our stereotypes, since they refer to characteristics that are very dif®cult to measure objectively. People's beliefs about certain attributes of the two sexes that can be objectively assessed, however, turn out to be quite accurate (e.g., that women are more involved than men in the care of young children). This work has been done with adults, and we can only assume that the accuracy of children's gender stereotypes would improve with age, as they accumulate information about a larger and larger sample of exemplars. At the end of the twentieth century, then, a predominant perspective on gender development is a dual one focusing on individual differences. Its central themes are that children will differ in the degree to which they become sex-typed as a result of (a) the strength of the socialisation pressures they have experienced, and (b) the nature and coherence of their gender schemasÐtheir knowledge about the characteristics stereotypically associated with each sex, and about what the social expectations are for persons of their own sex. Or course, socialisation and cognitive factors in gender development are not truly distinct. For example, socialisation pressures are one source of information enabling children to develop their knowledge concerning the gendered norms that they are expected to adopt. The direct socialisation experiences children have, in other words, constitute a major source of information upon which cognitive structures are built. Indeed, the whole cultural milieu in which a child grows up presents to children an array of cultural beliefs and practices concerning gender, and when children draw on these to construct their gender schemas, it can reasonably be said that they are being socialised by the surrounding culture into becoming co-practitioners of these cultural forms. An emphasis on cognitive and socialisation factors by no means precludes a recognition of possible biological in¯uences that may generate different predispositions in boys and girls. Nowadays there is widespread recognition of the importance of biological factors. However, since

11. Perspectives on gender development

209

biological sex and socially ascribed gender are so completely redundant, it has proved dif®cult to tease them apart. Something is known concerning sex differences in brain structures and functions. For example, in males, more functions are lateralised, so that they are associated with activation primarily in one hemisphere of the brain, whereas in females, the two hemispheres are more likely to be both activated for a speci®c function. However the possible behavioural impacts of these structural differences are far from being understood. A good deal is known concerning the physiological events during gestation that differentiate the genital structures of male and female foetuses. And we know, too, something about the way in which prenatal hormones organise the developing foetal brain so as to create different propensities and sensitivities in the two sexesÐtendencies that will manifest themselves behaviourally at various times during postnatal development, perhaps requiring either a biological or environmental trigger for their activation. As noted earlier, most accounts of gender development do note possible biological underpinnings for some of the gender differentiation that occurs. And there is considerable interest in taking biological explanations one step backward to the genetic factors that may control biological differentiation. But we are a long way from having traced pathways from genes to the behavioural attributes that typically differentiate the sexes. In psychology, the great bulk of work has been concerned with the social and cognitive factors that are thought to underlie this differentiation, over and above what any biological predispositions may call for. LIMITATIONS OF THESE PERSPECTIVES In the 1980s and 1990s, it began to be evident that these perspectives were not serving well, on the whole. First of all, it was increasingly clear that sex differences in children's psychological attributes as usually measured were not substantial, and ®ndings were inconsistent from one study to another. Focusing on individual variation along dimensions such as ``masculinity'', ``femininity'', ``androgyny'', or ``degree of sex-typing'' was not turning out to be a strategy that accounted for much variance in behavioural outcomes. While this might mean that gender simply is not an important factor in children's daily lives, it could also mean that gender matters only in certain contexts, so that aggregating across contexts attenuates gender-related phenomena that are in fact quite strong. If so, it would appear that research should turn to studies of moderating contexts and, in addition, look for gender-related outcome variables that are more robust than ``sex-typing'' as we have measured it with toy- or activity-preference tests or clusters of personality traits. The socialisation and cognitive perspectives have proved disappointing in another respect: Empirical tests have failed to give consistent support for

210

Maccoby

the predicted connections between processes and outcomes. As noted earlier, the similarities in the ways socialisation agents treat boys and girls far outweigh the differences. Still, some differences are found. Ruble and Martin's comprehensive review of studies examining connections between differential socialisation and sex-typed outcomes shows very meagre relationships: ``Although adults and peers treat boys and girls differently in many ways, especially concerning activities and interests, the role of these processes in children's gender-related preferences and behaviours remains to be demonstrated'' (Ruble & Martin, 1998, p. 982). In a similar vein, they note that variations among children in the level of their knowledge of gender stereotypes are generally unconnected to individual differences in sex-typed behaviour (e.g., Powlishta, 1995). This is by no means a fatal blow to cognitive theories of gender development, because in many respects gender cognitions are important in their own right, regardless of how and whether they ``drive'' individual differences in sex-typed behaviour. Furthermore, children's understanding of gender identityÐtheir own and that of other peopleÐdoes appear to be somewhat connected to other aspects of gender development early in life. However, understanding of gender identity is virtually complete in most children by the age of about 3, and hence (since it varies so little) cannot correlate with individual differences in sex-typing that emerge after approximately the ®rst 3 years of life. Similarly, in come contexts and at some ages, there is very little individual variation in the choice of same-sex playmates, so that this aspect of sex-typing cannot be predicted from individual cognitive or personality characteristics. Clearly, a rethinking of the long-dominant perspectives described here is in order, and has begun. A major reorientation is a shift away from a focus on individual differences in outcomes, and an increasing focus on some robust outcomes for groupsÐgender effects that might appear primarily, or only in certain speci®c social contexts. ROBUST GENDER PHENOMENA IN THE CONTEXT OF GROUPS What might these outcomes and contexts be? Indications began to emerge in the late 1970s that the answer could be found in the context of pairs or larger groups of children engaged in social interaction. Jacklin and Maccoby (1978) observed 92 pairs of previously unacquainted children, all close to 33 months of age, as they interacted with each other and with toys. Tallies were kept of the instances of social behaviours (both positive and negative, verbal and nonverbal) each child directed toward his/her partner. Some of the pairs were composed of two boys, some of two girls, and some were mixed-sex pairs. Results were that children of both sexes directed about twice as much social behaviour toward partners who were of their

11. Perspectives on gender development

211

1

own sex as they did to other-sex partners . It is notable that when data were computed without regard to the sex of the child's partner, boys and girls displayed virtually identical levels of social behaviour. In other words there was no overall sex difference in a personality dimension that might be called ``sociability''. This ®nding underlines the fact that analyses which look only at the behaviour of individual children without regard to social context can totally obscure powerful gender phenomena. It is no surprise, then, that many simple comparisons between boys and girls have shown sex differences to be weak or absent. Theorists who have emphasised the importance of context, and variations in the salience of gender from one context to another, might rightly see the ®ndings of the Jacklin and Maccoby study as a vindication of their position. But there is something very speci®c about the context that turned out to be important in this work: It was the sex of a child's interactive partner that mattered, not context construed as environmental setting or prior priming conditions. The study points to the importance of group composition, and/or of relationships, in how gender is enacted. A study of preschoolers, conducted at about the same time, helps us to understand these ®ndings. Wasserman and Stern (1978) laid down a strip of carpet on a playroom ¯oor, and asked a child (sometimes a boy, sometimes a girl) to stand quietly at one end of it. Another child was then placed at the ``starting'' end of the strip, and asked to walk along it to the standing child. It turned out that children would walk up quite close to the standing child, facing directly forward, if the standing child was of their own sex. Approaching an other-sex child, however, both boys and girls would turn away as they approached and stop sooner. Notably, this occurred whether they were acquainted with the standing child or not, indicating that we are seeing here a form of other-sex avoidance that is not driven by previous experience with another child, but rather by the other child's membership in a gender category. Perhaps we are seeing here an instance of the kind of self-regulation pointed to by cognitive social learning theorists, in which a child recognises the sex of another child, knows that it is considered inappropriate to associate with children of the other sex, and hence inhibits the approach to the other child. It could be argued that children begin to show cross-sex avoidance at such a young age that they are not yet able to code their own gender identity and that of other children reliably and explicitly. In traditional theories, there has never been a claim that the cognitions involved are conscious or deliberate, but still, gender-cognitive processes are assumed to occur at some level. An alternative possibility is that there is simply an unrecognised raw emotion connected with gender categories. The early

1 These results were replicated in a study of English children ranging in age from 19 to 39 months (Lloyd & Smith, 1986).

212

Maccoby

paper by Zajonc, Feeling and thinking: Preferences need no inferences (1980), presented evidence that indeed preferences (liking/disliking, approach/avoidance) can be primary, immediate, without (or prior to) activation of related cognitions. Recent writings by Panksepp (1998), stress not only the immediacy of emotional reactions, but their deep mammalian origins. ``Primal'' emotions, including affective reactions to certain characteristics of same- or other-sex conspeci®cs, are claimed to be instinctive and species-wide. Given a pattern of wariness toward children of the other sex, and/or especial interest in, or compatibility with, own-sex other children, it is predictable that when children have a choice of playmates, they will congregate in same-sex pairs or groups. In childhood, there is a clear pattern of cross-sex avoidance and/or same-sex preference that begins in about the third year of life and becomes progressively stronger through middle childhood. Children's tendency to congregate socially with others of their own sex has long been noted in the developmental literature, and has been thoroughly documented, in a variety of cultures and subcultures (see, for reviews, Hartup, 1983; Maccoby & Jacklin, 1987; Ruble & Martin, 1998). In modern Western societies, it is manifested most strongly in situations not structured by adults, though in more traditional societies the structures adults provide for children certainly contribute to it. The phenomenon of gender segregation in childhood is remarkably robust, with very little overlap, by the age of 5 or 6, between the distributions of the two sexes with respect to the gender of the other children with whom they spend their free social time (Maccoby, 1998). The students of gender segregation have been concerned with the factors that bring it about (see Leaper, 1994) and with the implications for how gender development should be studied. Clearly, one implication of gender segregation is that it is important to continue to study and understand the nature of the group processes that occur in all-male as compared to all-female dyads or groups, although much progress has already been made in this work. There is now considerable evidence that the groups or dyads formed by girls, as compared with boys, differ with respect to the agendas they enact, and in their prevailing interaction ``styles''. The nature of these differences has been summarised elsewhere (Maccoby, 1998; Ruble & Martin, 1998). Here it is suf®cient to note a few dominant trends: 1

The themes that appear in boys' fantasies, in the stories they invent, the scenarios they enact when playing with other boys, and the ®ctional fare they prefer (books and TV) involve danger, con¯ict, destruction, heroic actions, and trials of physical strength. Girls' fantasy and play themes tend to be oriented around domestic or romantic scripts, portraying characters who are involved in social relationships and depicting the maintainance or restoration of order and safety.

11. Perspectives on gender development 2

3

213

Interaction among boys, more often than among girls, involves roughand-tumble play, competition, con¯ict, ego displays, risk-taking, and striving for dominance. Girls, by contrast, are more responsive to the inputs of their interactive partners, more likely to use suggestions rather than imperative demands, and are more likely to construct collaborative scripts in which the actions of play characters are reciprocal. (See Leaper, 1991, on girls' collaborative discourse style.) This does not imply that girls do not assert their own objectives, or that their interactions are con¯ict-free, only that they seek their individual goals in the context of also striving to maintain group harmony (see Sheldon, 1992). Girls' and boys' friendships are qualitatively different, girls' friendships being more intimate in the sense that friends share information about the details of their lives and concerns, whereas boys typically know less about their friends' lives and base their friendships on shared activities. The break-up of girls' friendships is more emotionally intense than it is for boys' friendships.

By age 6, too, boys typically play in larger groups. Benenson, Apostolaris, and Parnass (1997) showed, in a study of same-sex six-child groups, that between the ages of 4 and 6, boys greatly increased the time they spent in coordinated group activities, so that by the age of 6, they were spending 74% of their time in such activities. No such increase occurred for girls, whose coordinated group activities dropped below 20% of their time at age 6. And, girls have been found to show more enjoyment than boys when engaged in dyadic interaction (Benenson, 1993), while this differential is not found for interaction in larger groups. Girls, too, are found to sustain longer bouts of interaction in dyads than do boys. In other words, girls actively seek, prefer, and elaborate dyadic interactions, while boys do not appear to ®nd such interactions to be especially gratifying, and instead gravitate toward coordinated activities carried out in larger groups. There is evidence, too, that boy's groups are not only larger, but also stronger in some senseÐi.e., more cohesive, with stronger in-group identi®cation and stronger boundaries, in the sense of more strongly excluding both girls and adults. (See Maccoby, 1998, for elaboration of these processes.) The fact that boys congregate in larger groups has important implications. When in dyads, children of both sexes are relatively noncompetitive, and more emotionally supportive of their interactive partners, than they are when participating in larger groups (see Benenson, Nicholson, Waire, Roy, & Simpson, in press), something that was noted many years ago in interactions among adults (Bales & Borgatta, 1955). Can it be, then, that the greater competitiveness and lesser positive intimacy in male±male interaction can be accounted for by boys being more often in larger play groups? Perhaps so, but the fact that they congregate in larger groups may itself be a re¯ection of their preference for certain forms of competitive but

214

Maccoby

coordinated activity that can only be performed in larger groups. Indeed, boys form coalitions to achieve group goalsÐand gain group power therebyÐto an extent that girls seldom do. This account suggests that the two sexes are pursuing different agendas in their same-sex groups. But children in groups are not always engaged in enacting these differentiated agendas, and when they are not, male± male and female±female interactions can be much alike. It should be noted, too, that researchers have not yet spelled out the developmental time-line for these aspects of gender differentiation. Thus, although there is reason to have a good deal of con®dence in their occurrence, it is not yet clear how they wax and wane, and whether there are privileged sequences such that some processes need to occur before others can come into play.

THE ETHOLOGICAL PERSPECTIVE Ethology originally referred to the study of animal behaviour, and a guiding perspective in ethological research was an evolutionary one. The principle of natural selection was invoked to explain species-wide adaptive behaviours, such that each animal species was equipped to survive in its particular environmental niche. As the perspective began to be applied to human behaviour, it was of great interest to trace similarities between humans and nonhuman primates, their closest relatives on the evolutionary tree. In the 1950s and 1960s, John Bowlby (1969) drew upon the work of Harlow (1961) and Hinde (1966) to show striking resemblances between the patterns of attachment behaviour in human infants or toddlers and what was seen in young monkeys and apes. In the 1970s, work by Blurton-Jones and colleagues (Blurton-Jones, 1972) and Strayer (Strayer, 1977; Strayer & Strayer, 1978) focused on patterns of children's play, social dominance, aggression, and peer af®liation, identifying further parallels with the young of nonhuman primates. In observing sex differences in behaviours of this kind, ethologists do not look for generalised ``trait'' differences, but rather for the situations under which a difference does or does not appear, and the form the behaviour takes. It would not be meaningful to an ethologist, for example, to ask whether male ungulates are more aggressive than female ungulates. Rather, they would note that both males and females attack predators, and do so in the same way: with their hooves. Only males, however, ®ght over territory and mates, and they use their horns, rather than their hooves, to attack or threaten other males. In a similar way, in studying human children, ethologists look for speci®c behavioural topography, and speci®c contexts in which sex differentiation is seen. Striking parallels between human children and the young of other primates have indeed been found, and this

11. Perspectives on gender development

215

fact has been interpreted as pointing to evolved, genetically guided underpinnings for certain elements of human behaviour. The evolutionary history of human development continues to be a matter of great interest up to the present time (see Geary & Bjorklund, 2000, for a review). Throughout, thinkers involved in this work have been concerned with what adaptive purpose an evolved behaviour pattern might serve. In what way might an evolutionary perspective be pertinent to gender development? Most evolutionary adaptations, after all, are seen as specieswide, occurring in both sexes. However, in bisexual species, the distinctive roles of the two sexes in reproduction is thought to have produced different adaptive behaviours: i.e., different strategies for mate selection, and differential involvement in the rearing of young. With respect to these domains, then, the two sexes are seen as distinct subspecies. Might evolution have anything to do with gender differentiation that occurs in childhood, before the age when the activities of mate-selection and care of offspring emerge? The fact that young monkeys and apes separate into same-sex playgroups, and display some of the same sex-differentiated play styles as those seen in human children, strongly suggests that there is indeed an evolutionary basis for these behaviours. So does the fact that there appears to be substantial uniformity in these patterns across human cultures. Geary and Bjorklund (2000) say: ``From an evolutionary perspective, these sex differences are predicted to be a re¯ection of and a preparation for sex differences in adult reproductive activities'' (p. 60). Thus boys' competitiveness and dominance strivings are seen as preparation for adult male competition over mates, whereas girls' greater social responsiveness and cooperativeness with other girls can seen as preparation for participation in the kin-based social groups of females in which most rearing of the young occurs, in nonhuman primates as well asÐpresumablyÐin the huntergatherer bands where some underpinnings of present-day human characteristics evolved. In what possible way could sex segregation in childhood serve an adaptive purpose, from the standpoint of the subsequent successful reproductive activities as adults? One hypothesis is: It functions to prevent incest, with its attendant risks for expression, in offspring, of genetic defects carried on recessive genes (for elaboration and evidence, see Maccoby, 1998; Wolf, 1995). According to this account, humans are predisposed to lose sexual interest in anyone with whom they have been closely associated in childhood, such as siblings and other close relatives. Thus, paradoxical though it may seem, cross-sex avoidance in childhood can be seen as a means of fostering or safeguarding future heterosexual attraction. It is more dif®cult to see how some of the other sex-differentiated patterns we have noted might contribute to individual reproductive success and viability of an individual's offspring. Rather, it would seem timely to expand the ecological perspective to include group success, group viability. Among chimpanzees, and in the human societies closest to the way our

216

Maccoby

ancestors probably lived (Collier & Risaldo, 1981), males form coalitions to engage in cooperative group hunting and group warfare2, enterprises not directly related to individual reproductive success, but relevant to survival of the troop. Some of the processes seen in boys' groups, then, may have the function of regulating hostility and competition among group members in the interests of allowing cooperative group enterprises to emerge. THE PSYCHOBIOLOGICAL PERSPECTIVE The ethological perspective largely describes instinctive elements in a species' behavioural repertoire: i.e. behavioural dispositions assumed to be governed directly by evolved genetic programmes. From this perspective, individual life experience and environmental contexts have little in¯uence, other than to provide the environmental triggers required to ``release'' an instinctive behaviour. In fact, however, there are many aspects of the environment, other than innate releasers, that function jointly with genetic factors to in¯uence behavioural development. Many years of research have revealed that the way genetic (G) instructions are carried out depends on environmental (E) inputs at every stage of development from conception to maturity. The modern psychobiological view is that G and E have a bi-directional, reciprocal relationship, and cannot properly be understood as separate components whose effects can be independently estimated and then compared or summed (Gottleib, Wahlsten, & Lickliter, 1998). G Ò E interactions are widespread, and recent work has begun to ®ll in the gap between genes and behavioural outcomes, by showing how both environment (i.e., conditions of rearing) and an allele of a given gene can affect a speci®c intervening biochemical process, which then in¯uences behavioural outcomes (Anisman, Zaharia, Meany, & Merali, 1998; Suomi, 1999; see Maccoby, 2000, for a summary). The role of environment, as it interacts with genetic predispositions, is much more complex than simply providing a releaser for an instinctive response. In both rodents and primates, it has been shown that animals coming from different genetic strains will manifest their different predispositions only under certain early rearing conditions (i.e., being deprived of contact with maternal animals or peersÐsee Hood & Cairns, 1989; Suomi, 1997). And for some song birds, a young male will not acquire the species-speci®c male courting song unless he is reared in the company of older females, who ``train'' him by responding selectively to the elements of song that females of the species ®nd most compelling (West & King, 2001).

2 For information on hunting and warfare in chimpanzees, see: McGrew, Marchant, and Nichida (1996), Boesch and Boesch (1989), and Stanford, Wallis, Matana, & Goddall (1994).

11. Perspectives on gender development

217

Conditions of rearing matter, too, in how different the two sexes become. Wallen (1996) has summarised a series of studies with rhesus monkeys, showing that males display elevated levels of threat/aggression toward other animals if they were earlier raised with only limited access to peers, while such limited access increases the amount of submissive, not aggressive, behaviour in females. For males, the amount of rough play is also affected by the social conditions of rearing, while for females, the level of such play remains low regardless of conditions of rearing. When females are prenatally exposed to androgens (late in gestation, or over an extended period), they subsequently show elevated levels of rough play, while there is no such effect for males. Clearly, both biology and conditions of rearing are important, but differently for the two sexes. Wallen notes that it is the sexdimorphic behaviours most strongly affected by prenatal hormones in a given sex that show the least effects of rearing conditions for that sex. And he concludes: ``These studies demonstrate that the expression of consistent juvenile behavioural sex differences results from hormonally induced predispositions to engage in speci®c patterns of juvenile behaviour whose expression is shaped by the speci®c social environment experienced by the developing monkey'' (Wallen, 1996, p. 364). Humans are very different from birds, mice, and monkeys, but nevertheless there are useful parallels here. We see that whatever differential predispositions boys and girls may have, it is likely that the way they are enacted will depend greatly on the social conditions provided by the adults and peers with whom they interact. Societies differ with respect to how much time children of each sex spend with adults, with peers of their own sex, and with peers or siblings of the other sex. They also differ with respect to how much autonomy peer groups have at what developmental periods. These cultural variations may be expected to produce variations in the degree and kind of sex differentiation that appears as children grow up in different societies. INTEGRATING PERSPECTIVES The preceding account is meant to show that there is a good case for including the ethological and psychobiological perspectives in any attempt to understand gender differentiation in childhood. These perspectives are not meant to replace the socialisation and cognitive perspectives, only to enrich them by expanding our view of how biological and experiential and cognitive factors work together when it comes to the enactment of gender. The two biological perspectives are especially useful in helping us to understand any characteristics where between-sex differences are robust and consistent across cultures and even across species. However, there is considerable variation within each sex and among cultural groups in the nature of gendered behaviour displayed and the contexts in which it

218

Maccoby

appears. Whereas some boys, for example, establish a network of good male friends, and participate actively in male group activities, others are loners or peripheral ``hangers-on'' to these groups, and still others are the victims of teasing and humiliation by other boys. Some children join peer groups that are basically prosocial; others associate mainly with same-sex peers who engage in risky, antisocial behaviour. Among girls, too, there is variation in how fully they participate in ``girl culture'', and in how much interest they have in less ``feminine'' activities such as team sports. There is evidence that these individual differences re¯ect developed differences in competencies or vulnerabilities acquired at earlier periods of developmentÐ differences which in their turn undoubtedly re¯ect both within-sex genetic variability and individual socialisation histories. Thus the socialisation and cognitive perspectives should be especially pertinent to the understanding of such within-sex variation. In the last century, there have been substantial changes in gender roles and the relationships between the sexes, changes that have occurred much too rapidly to be explained in genetic terms. These changes underscore how large the social factors are in such matters as the relative dominance and power of the two sexes, and make it clear that these matters are indeed open to change and not built into human nature as a result of our species evolution. The foregoing review has suggested that male power in society stems in no small degree from male groups and male alliances, despite the internal con¯icts and competition that characterise male interaction. The paradox is this: We have seen that girls and women have especially strong interactive skills that support collaboration and cooperation in their close relationships. Yet women's social groupings do not appear to yield power in out-of-home contexts to the extent that men's do. Perhaps it will be possible to gain insight into this issue with more focused developmental research on the way power is exercised within same-sex groups and between the sexes. The fuller incorporation of the ethological and psychobiological perspectives on gender into our existing frameworks should enrich the research agenda of students of social development. More detailed observation of children in groups is called for, with more attention to the agendas that children of the two sexes seem motivated to enact, and more careful delineation of how gender-related patterns change with age. Of especial interest are the changes that occur in gender enactment during the transition into adolescence. Useful cues for how such work can be carried out comes from pioneers in neighbouring disciplines (e.g., the sociologists Maltz & Borker, 1982; and Thorne, 1986; and sociolinguist Eckert, 1996, 2000). Developmental psychology is in the process of building a coherent social psychology of childhood. There has been considerable attention to individual children as members of groupsÐtheir acceptance or rejection by peers, their group-entry skillsÐand the causes and consequences of individual

11. Perspectives on gender development

219

variation in these things. In addition, there has been some work on the nature of groups, apart from the individuals who make them up. In their comprehensive review, Rubin, Bukowski and Parker (1998) distinguish group processes from the interactions that occur within dyadic relationships, noting that groups can have properties that individuals and dyads cannot have (e.g., density of relationships, norms). Studies of ``crowds'' and ``cliques'' (e.g., Brown, 1990) underscore the importance of shared norms in the in¯uence peer groups have on their members, and Kinderman (1993) has shown that a group can have its own identity over and above the identity of its members: i.e., a group norm persists through a school year despite considerable turnover in the membership of the group. It is time for gender to take a more central place than it has occupied so far in such work, with more consistent attention to the gender composition of dyads and groups. Only with this knowledge in hand will we be able to understand the role of peers as gender-socialisation agents, and the way in which children build shared gender cognitions that can serve either to amplify or dampen the gender differentiation of roles and status. REFERENCES Anisman, H., Zaharia, M.D., Meany, M.J., & Merali, Z. (1998). Do early life events permanently alter behavioral and hormonal responses to stressors? International Journal of Developmental Neuroscience, 16, 149±164. Bales, R.F., & Borgatta, E.F. (1955). Size of group as a factor in the interaction pro®le. In A.P. Hare, E.F. Borgatta, & R.F. Bales (Eds.), Small groups: Studies in social interaction (pp. 496±512). Toronto, Canada: Random House. Bandura, A. (1965). In¯uence of models' reinforcement contingencies on the acquisition of imitative responses. Journal of Personality and Social Psychology, 1, 589±595. Bem, S.L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354±364. Benenson, J.F. (1993) Greater preference among females than males for dyadic interaction in early childhood. Child Development, 64, 544±555. Benenson, J.F., Apostolaris, N.H., & Parnass, J. (1997). Age and sex differences in dyadic and group interaction. Developmental Psychology, 33, 538±543. Benenson, J.F., Nicholson, C., Waire, A., Roy, R., & Simpson, A. (2001). The in¯uence of size of social context on children's competitive behaviour. Child Development, 72, 921±928. Blurton-Jones, N. (Ed.). (1972). Ethological studies of child behaviour. London: Cambridge University Press. Boesch, C., & Boesch, H. (1989). Hunting behavior of wild chimpanzees in the Tai National Park. American Journal of Physical Anthropology, 78, 547±573. Bowlby, J. (1969). Attachment. New York: Basic Books. Brown, B.B. (1990). Peer groups and peer cultures. In S.S. Feldman & G.R. Elliott (Eds.), At the threshold (pp. 171±196). Cambridge, MA: Harvard University Press.

220

Maccoby

Bussey, K., & Bandura, A. (1992). Self-regulatory mechanisms governing gender development. Child Development, 63, 1236±1250. Bussey, K., & Bandura, A. (1999). Social-cognitive theory of gender development and differentiation. Psychological Review, 106, 676±713. Chodorow, N. (1978). The reproduction of mothering: Psychoanalysis and the sociology of gender. Berkeley, CA: University of California Press. Collier, J., & Risaldo, S. (1981). Politics and gender in simple societies. In S.B. Ortner & H. Whitehead (Eds.), Sexual meanings. New York: Cambridge University Press. Deaux, K., & LaFrance, M. (1998). Gender. In D.T. Gilbert, S.T. Fiske, & G. Lindsey (Eds.), Handbook of social psychology, Vol. 1. New York: McGraw-Hill. Eckert, P. (1996). Vowels and nailpolish: The emergence of linguistic style in the preadolescent heterosexual marketplace. In J. Ahlers, L. Bilmes, M. Chen, M. Oliver, N. Warner, & S. Werhsteim (Eds.), Gender and belief systems. Berkeley, CA: Berkeley Women and Language Group. Ge, X., Conger, R., Cadoret, R., Neiderhiser, J., Yates, W., et al. (1996). The developmental interface between nature and nurture: A mutual in¯uence model of child anti-social behavior and parent behavior. Developmental Psychology, 32, 574±589. Geary, D.C., & Bjorklund, D.F. (2000). Evolutionary developmental psychology. Child Development, 71, 57±65. Gelman, S.A., Coleman, P., & Maccoby, E.E. (1986). Inferring properties from categories, versus inferring categories from properties: The case of gender. Child Development, 57, 396±404. Gottleib, G., Wahlsten, D., & Lickliter, R. (1998). The signi®cance of biology for human development: A developmental psychobiological systems view. In W. Damon & R.M. Lerner (Eds.), The handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed.). New York: John Wiley. Harlow, H.F. (1961). The development of affectional patterns in infant monkeys. In B.M. Foss (Ed.), The determinants of infant behaviour, Vol. 1. London: Methuen. Hartup, W. (1983). Peer relations. In P.H. Mussen (Series Ed.), E.M. Hetherington (Vol. Ed.), Handbook of child psychology, Vol. 4. New York: John Wiley. Hinde, R.A. (1966). Animal behaviour: A synthesis of ethology and comparative psychology. New York: McGraw-Hill. Hood, K.E., & Cairns, R.B. (1989). A developmental genetic analysis of aggressive behavior in mice: 4. Genotype±environment interaction. Aggressive Behavior, 15, 361±380. Huston, A.C. (1983). Sex typing. In P.H. Mussen (Series Ed.), E.M. Hetherington (Vol. Ed.), Handbook of child psychology, Vol. 4, Socialization, personality and social development (4th ed., pp. 387±467). New York: John Wiley. Jacklin, C.N., & Maccoby, E.E. (1978). Social behavior at 33 months in same-sex and mixed-sex dyads. Child Development, 49, 557±569. Kinderman, T.A. (1993). Natural peer groups as contexts for individual development: The case of children's motivation in school. Developmental Psychology, 29, 970±977. Kohlberg, L. (1966). A cognitive-developmental analysis of children's sex-role concepts and attitudes. In E. Maccoby (Ed.), The development of sex differences. Stanford, CA: Stanford University Press.

11. Perspectives on gender development

221

Leaper, C. (1991). In¯uence and involvement in children's discourse: Age, gender and partner effects. Child Development, 62, 797±811. Leaper, C. (1994). Childhood gender segregation: Causes and consequences. New Directions for Child Development. San Francisco, Jossey-Bass. Lloyd, B., & Smith, C. (1986). The effects of age and gender on social behaviour in very young children. British Journal of Social Psychology, 25, 33±41. Lytton, H., & Romney, D.M. (1991). Parents' differential socialization of boys and girls: A meta-analysis. Psychological Bulletin, 109, 267±296. Maccoby, E.E. (1990). The role of gender identity and gender constancy in sexdifferentiated development. In D. Schrader (Ed.), The legacy of Lawrence Kohlberg, No. 47, New Directions for Child Development. San Francisco: JosseyBass. Maccoby, E.E. (1998). The two sexes: Growing up apart, coming together. Cambridge, MA: Harvard University Press. Maccoby, E.E. (2000). Parenting and its effects on children: On reading and misreading behavior genetics. Annual Review of Psychology, 51, 1±27. Maccoby, E.E., & Jacklin, C.N. (1974). The psychology of sex differences. Stanford, CA: Stanford University Press. Maccoby, E.E., & Jacklin, C.N. (1987). Gender segregation in childhood. In H. Reese (Ed.), Advances in child behavior and development. New York: Academic Press. Maltz, D.N., & Borker, R.A. (1983). A cultural approach to male±female miscommunication. In J.A Gumperz (Ed.), Language and social identity (pp. 195± 216). New York: Cambridge University Press. Martin, C.L., & Halverson, C.F. (1981). A schematic processing model of sex typing and stereotyping in children. Child Development, 54, 563±574. McGrew, W.C., Marchant, L.F., & Nichida, T. (1996). Great ape societies. New York: Cambridge University Press. Mischel, W. (1966). A social-learning view of sex differences in behavior. In E.E. Maccoby (Ed.), The development of sex differences. Stanford, CA: Stanford University Press. O'Connor, T.G., Deater-Deckard, K., Fulker, D., Rutter, M., & Plomin, R. (1998). Genotype-environment correlations in late childhood and early adolescence: Antisocial behavioral problems and coercive parenting. Developmental Psychology, 34, 970±981. Panksepp, J. (1998). Affective neuroscience: The foundations of human and animal emotions. New York: Oxford University Press. Perry, D.G., & Bussey, K. (1979). The social-learning theory of sex differences: Imitation is alive and well. Journal of Personality and Social Psychology, 37, 1699±1712. Powlishta, K.K. (1995). Intergroup processes in childhood: Social categorization and sex-role development. Developmental Psychology, 31, 781±788. Rubin, K.H., Bukowski, W., & Parker, J.G. (1998). Peer interactions, relationships, and groups. In W. Damon & N. Eisenberg (Eds.), Handbook of child psychology, Vol. 3 (5th ed., pp. 619±700). New York: John Wiley. Ruble, D.N., & Martin, C.L. (1998). Gender development. In W. Damon & N. Eisenberg (Eds.), Handbook of child psychology, Vol. 3 (5th ed., pp. 933±1016). New York: John Wiley.

222

Maccoby

Sears, R.R., Rau, L., & Alpert, R. (1965). Identi®cation and child rearing. Stanford, CA: Stanford University Press. Sears, R.R., Maccoby, E.E., & Levin, H. (1957). Patterns of child rearing. Evanston, IL: Row Peterson. Sheldon, A. (1992). Con¯ict talk: Sociolinguistic challenges to self-assertion and how young girls meet them. Merrill-Palmer Quarterly, 38, 95±117. Stanford, C.B., Wallis, J., Matama, H., & Goddall, J. (1994). Patterns of predation by chimpanzees on red colobus monkeys in Gombe National Park, 1982±1991. American Journal of Physical Anthropology, 94, 213±228. Stern, M., & Karraker, K.H. (1989). Sex stereotyping of infants: A review of gender labeling studies. Sex Roles, 20, 501±522. Strayer, F.F. (1977). Peer attachment and af®liative subgroups. In F.F. Strayer (Ed.), Ethological perspectives on preschool social organization, Memo de Recherche No. 5. Universite de Quebec. Strayer, F., & Strayer, F.F. (1978). Social aggression and power relations among preschool children. Aggressive Behavior, 4, 173±182. Suomi, S.J. (1997). Long-term effects of different early rearing experiences on social, emotional and physiological development in non-human primates. In M.S. Keshevan & R.M. Murra (Eds.), Neurodevelopmental models of adult psychopathology. Cambridge: Cambridge University Press. Suomi, S.J. (in press). A biobehavioral perspective on developmental psychopathology: Excessive aggression and serotonergic dysfunction in monkeys. In A.J. Samaroff, M. Lewis, & S. Miller (Eds.), Handbook of developmental psychopathology. New York: Plenum Press. Thorne, B. (1986). Boys and girls together . . . but mostly apart: Gender arrangements in elementary schools. In W.H. Hartup & Z. Rubin (Eds.), Relationships and development. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Wallen, K. (1996). Nature needs nurture: The interaction of hormonal and social in¯uences on the development of behavioral sex differences in rhesus monkeys. Hormones and Behavior, 30, 364±378. Wasserman, G.A., & Stern, D.N. (1978). An early manifestation of differential behavior toward children of the same and opposite sex. The Journal of Genetic Psychology, 133, 129±137. West, M.J., & King, A.P. (2001). Science lies its way to the truthÐreally. In E. Blass (Ed.), Handbook of behavioral neurobiology, Vol. 12: Developmental psychobiology (pp. 587±614). New York: Plenum Press. Wolf, A. (1995). Sexual attraction and childhood association: A Chinese brief for Edward Westermark. Stanford, CA: Stanford University Press. Zajonc, R.B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35, 151±175.

12 Agency in development Todd D. Little

University of Kansas, Lawrence, USA

Models and metatheories of human development have evolved signi®cantly over the previous two centuries. Most changes are clearly re¯ections of progress and integration. Early mechanistic models, for example, have incorporated organismic features such as when behaviour is linked to changes in biology (GerieÂpy, 1996) or when social learning theorists speak of agents and agency (Bandura, 1997). Some organismic models, on the other hand, have revealed mechanistic tendencies such as when behaviour interventions are focused on the ®rst few years of life under the belief they will inoculate children from future harm (Lewis, 1997). Because these changes have blurred traditional boundaries among metatheories, I revisit the origins of the organismic model in an effort to revitalise it by highlighting the place and functions of agency in development. In addition, I will examine context's role in in¯uencing the expression of agency throughout development and discuss both statistical advances and methodological limitations for modelling the agentic self in context. My primary goal is to discuss the linkages among the various dimensions that comprise personal agency across the lifespan. Although volitional goaldirected activity (i.e., agency) has an evolutionary basis (Chapman, 1984; see Hawley & Little, 2002, for a detailed exposition), I will focus on the various levels of ontogenetic meanings as they relate to the individual's sense of personal agency. Generally speaking, one's sense of personal agency stems from goal-driven actions. It functions as a personal resource for facing the challenges that emerge throughout development (Little, 1998). This sense of personal agency is particularly relevant given today's developmental milieu of increased challenges and potential disruptions to the successful development of individuals. Even casual observation reinforces the notion that most cultural, societal, social, and familial structures have dramatically transformed during the latter part of the previous century. Because sociocultural features such as educational choices, familial constellations, behavioural norms, career directions, post-retirement lifestyles, and so on will probably continue to become more varied and differentiated, individuals will be faced with new and diverse challenges to successful development at every phase of the lifespan. How individuals

224

Little

meet and overcome such life-course challenges is the operative realm of personal agency. This chapter is organised around three related themes that I feel need to become focal directions of study in order to fully articulate the role that agency plays in development. The ®rst focal direction is to understand better the interrelations among the varied features of the self system and the synergistic connections among the layered self-regulatory processes of development (e.g., motives, goals, volitional control mechanisms, mental representations, subjective interpretations, and so on). As I describe in detail later, an organismic perspective on the individual offers a useful starting point from which to unpack fully the relationships among the qualities and quantities of one's sense of personal agency (e.g., Hawley & Little, 2002; Little, Hawley, Henrich, & Marsland, 2002; Magnusson & Cairns, 1996; Ryan, 1993; Ryan, Sheldon, Kasser, & Deci, 1996). The second focal direction is to examine the in¯uence of context more precisely (e.g., Bronfenbrenner, 1995; Lerner, 1996). Although the resurgence of cross-cultural designs, for example, indicates progress in this direction, I suggest that clearly articulated theories about the nature of contexts need to emerge and that correspondingly compelling designs for contextual studies need to be implemented. Simply demonstrating that a change in context is associated with a change in behaviour provides only rudimentary information about what aspects of contexts are operating and does little to clarify the individual-level mechanisms by which context has its effects. The ®nal focal direction is to increase further the methodological sophistication by which the interplay of context and self is disentangled. The latter quarter of the previous century saw the introduction of an extremely powerful and versatile family of statistical modelling tools (e.g., general structural equation modelling techniques, including speci®c instantiations such as multilevel, growth-curve, and multiple-group models; Little, Schnabel, & Baumert, 2000; Marcoulides & Schumacker, 1996). These modelling tools are continuing to be re®ned and will soon offer even greater ¯exibility for researchers to ask and answer the types of complex questions that are needed for understanding developmental processes across contexts. This chapter ®nishes with a discussion of a number of issues and broader themes that emerge when one considers these factors in the study of agency in development. ORGANISMIC PERSPECTIVE REVISITED Because many research accounts rely on implicit assumptions about the nature of human behaviour, observers have noted that more rapid and ef®cient progress in the social and behavioural sciences will come about if we make explicit the larger models of development that drive our various microtheories (Shanahan, Valsiner, & Gottlieb, 1997). Historically, the organismic

12. Agency in development

225

model has provided a unifying basis for a number of productive and in¯uential research traditions (see Cairns, 1983, 1998). In light of current trends towards larger and more encompassing studies of development, I feel that the organismic model remains a viable and useful vantage point from which to view human development, particularly the development of the self. At the heart of an organismic perspective is the presumption that individuals are active agents and contributors to their own development. An organismic perspective views volitional behaviour as goal-directed actions, where actions are both purposive and self-initiated (BrandtstaÈdter, 1984, 1998; Chapman & Skinner, 1985). By extension, individuals are inherently active and self-regulatingÐthat is, agentic. An organismic approach to understanding the developing individual also involves an explicit focus on the interface between the self and context (Little et al., 2002; Ryan et al., 1996). Contexts re¯ect complex constellations of features, spanning the molar and micro levels, that both constrain and afford behaviour (Bronfrenbrenner, 1995; Gottlieb, 1997). Taken together, such features create a psychological carrying capacity within which the individual operates. As an agent in his/her own development, the individual functions as an integrated organism (e.g., assimilative and accommodative processes are invoked to maintain balance and equilibrium). Although an individual both in¯uences and is in¯uenced by the contexts in which s/he acts and develops, the agentic self must negotiate the affordances and constraints of each context. Developmentally speaking, an individual continually progresses along a predominantly self-guided path, giving form and meaning to his or her actions along the way. Such actions result from selective choices that emanate primarily from the individual. The organismic perspective presumes, then, that individuals are active agents who plot and navigate a chosen course through the uncertainties and challenges of the social and ecological environments; however, like trade winds and currents, environments sometimes hinder, sometimes bolster, and other times change the course of a developmental route (Little et al., 2002). As part of their integrated functioning, individuals engage in a selfevaluative feedback process, continuously interpreting and evaluating actions and their consequences. Across the episodes of activity in the varying constellations of context, the individual continually discovers and re®nes who one is and what one is capable of and, perhaps most importantly, under what conditions one's actions can or cannot have a desired effect. Under optimal circumstances, this continually evolving and actively monitored selfsystem gives rise to a strong and effective sense of personal agency. Different individuals with differing experiences and differing predispositions, however, will yield varying pro®les of personal agency because the sense of agency is a multifaceted and striated system of motives, beliefs, and behaviours (Hawley & Little, 2002; Little et al., 2002; Little & Wanner, 1997). This revitalised organismic perspective on agency necessarily entails a systems approach wherein the functioning of one system, such as the peer

226

Little

system, in¯uences the functioning of another, such as the self (Gottlieb, 1997). Because systems and subsystems are yoked, signi®cant advances in understanding the breadth of human behaviour will require that they be examined together rather than in isolation (Magnusson & Cairns, 1996). Such a view necessitates greater investment in planning a data collection protocol and greater investment in procuring appropriate samples of participants (e.g., samples derived on the basis of theoretical considerations rather than convenience). In my view, such larger-scale integrated studies of development can succeed only if they share a common and unifying metatheoretical perspective. When the pieces of a research programme come from different models and metaphors, they are unlikely to ®t together into a coherent picture. The puzzle pieces would rarely match up and the resulting picture would look more cubist than portraiture. Not only does an organismic perspective encourage broad-based studies of the agentic self, but it also supports a concerted focus on resilience and successful developmental outcomes. As each child begins to discover who s/he is and what s/he is capable of, the evolving competence system contributes to an integrated sense of personal agencyÐan agentic self. The resulting systems of action-control motives, goals, beliefs, and behaviours provide a developmental foundation that, as mentioned, is called upon to negotiate various developmental tasks and challenges throughout the life course. In facing these challenges, an agentic individual is primarily the origin of his or her actions, has high aspirations, perseveres in the face of obstacles, sees more and varied options, learns from failures, has a greater sense of wellbeing, and so on. A nonagentic individual, on the other hand, is primarily the pawn of unknown extra-personal in¯uences, has low aspirations, is hindered with problem-solving blinders, and often feels helpless and unempowered (Little, 1998; Little et al., 2002; Ryan et al., 1996; Skinner, 1995; Weisz, 1990). Such undesirable characteristics can stimulate additional negative feedback from others, such as teasing and victimisation (Graham & Juvonen, 1998). From a revitalised organismic perspective, one's sense of personal agency is not simply a reinvention of constructs such as self-esteem, self-ef®cacy, self-concept, and so on. Rather, it is a layered concept starting with the premise of volitional goal-directed actions. Such actions are motivated by both biological and psychological needs (Hawley, 1999; Hawley & Little, 2002; Little et al., 2002). Such actions are propelled by speci®c action-control beliefs about the links among agents, means, and ends (Chapman, 1984; Little, 1998; Skinner, Chapman & Baltes, 1988). Such actions are guided by general action-control behaviours that entail self-chosen forms and functions, means and ends (Lopez & Little, 1996; Little, Jones, Henrich, & Hawley, 2001; Little, Lopez, & Wanner, 2001). Regarding action-control beliefs, this perspective explicitly differentiates the possible belief relations among the primary constituents of intentional action (Chapman, Skinner, & Baltes, 1990; Little, Oettingen, Stetsenko, &

12. Agency in development Others as agent

Self as agent

Means

Means Cultural agency beliefs

Capacity (agency) beliefs Agent

227

Strategy beliefs

Agent

Causality beliefs

Cultural efficacy

Control expectancy Ends

Ends

Figure 12.1 Relations among the three constituents of human action, both when the self is the agent and when others are the agent. Means can vary from personal attributes such as effort and ability to external aids such as friends and teachers.

Baltes, 1995; Skinner & Chapman, 1987; Skinner et al., 1988; Stetsenko, Little, Gordeeva, Grasshof, & Oetingen, 2000). The three constituents of an action sequence are the actor, the goal, and the various means by which the goal (or end) can be obtained. Agentic action re¯ects an agent's general awareness of goals and the means to the goals, taking personal responsibility in pursuing a chosen goal, and the ability to select and utilise potential means (Chapman & Skinner, 1985). Given the three constituents of volitional activity, a number of belief types about the relations among these constituents are possible. Figure 12.1 displays six of these belief types (see also Skinner, 1996). Control expectancy re¯ects general beliefs about the link between the self as agent and the goal (e.g., ``when I want to do ____, I can.''). Collective ef®cacy, on the other hand, re¯ects the belief link between one's focal cultural group and the goal. Capacity beliefs re¯ect the links between the self and the various means that are relevant for attaining a chosen end (``I'm able to put forth enough effort to do this'', ``I possess the necessary skills to do this''). The counterpart linkage, when others are viewed as the agent of behaviour, can be termed collective capacity beliefs. Similarly, the link between the various means and ends can have two levels of meaning. When the self is the agent, the resulting system of beliefs would re¯ect strategy beliefs (i.e., what means work for me). Causality beliefs, on the other hand, re¯ect general views of the utility or usefulness of a given means such effort, luck, or ability for attaining a particular goal (Chapman et al., 1990; Little et al., 1995; Oettingen et al., 1994; Skinner et al., 1988). Ample research has supported these distinctions (see Little, 1998; Skinner, 1996, for overviews). A de®ning feature of the action-control view that separates it from other similar models of control perceptions is the

228

Little Prosocial means

Indirect ends

Inaction

Direct ends

Asocial means

Figure 12.2 An organismic classi®cation of coping behaviours in terms of their action-control aspects (based on Little, Lopez, & Wanner, 2001).

explicit means±ends analysis of actions. For example, because ability is considered separately from other means such as effort and access to powerful others, Stetsenko et al. (2000) were able to identify a meansspeci®c gender bias in the action-control beliefs of school-aged children. Girls whose actual school performance was on a par with boys, reported lower beliefs in their own ability to accomplish getting good grades than did boys. All other beliefs about their personal agency, such as effort, luck, and access to teachers, were on an equal footing, as would be expected from equally performing groups of children. Only the girls' beliefs in their access to ability revealed the bias. In other words, by focusing on the unique means that are utilised to achieve a goal, Stetsenko and co-workers identi®ed a speci®c source of bias in girls' personal agency pro®le that may hinder them as they progress along the educational ladder. A means±ends analytic framework not only applies to beliefs about behaviour, but also to actual behaviours that are enacted when goal pursuit is challenging. For example, the various strategies of coping behaviours that have permeated the literature can be categorised in terms of the nature of the means that are utilised (e.g., prosocial versus asocial) and the speci®c ends that are pursued (negotiating the stressor directly, or indirectly alleviating the consequences of the stressor; Lopez & Little, 1996). Figure 12.2 displays the conceptual quadrants into which many extant coping strategies can be placed. This multi-axial representation was tested in a short-term longitudinal study and found to be fully adequate in capturing the variability in coping behaviours both within and across measurement occasions (Little, Lopez, & Wanner, 2001). The model suggests that coping behaviours can be classi®ed in terms of their form, such as utilizing others or

12. Agency in development Overt:

Relational:

(dispositional)

(dispositional)

Overt: instrumental

Overt: reactive

Instrumental

Relational: instrumental

229

Relational: reactive

Reactive

Figure 12.3 Structural relations and decompositions among the forms and functions of aggressive behaviour (from Little et al., 2001).

relying on self-resources, and the function, such as taking direct action towards the stressor or indirect action to remediate the impact of the stressor. Such a means±ends analysis can also be applied to other domains of behaviour. For example, the research on aggressive-agonistic behaviour has revealed muddied mixtures of means (e.g., physical, verbal, relations) and ends (instrumental, reactive). Using an organismic metatheoretical perspective, my colleagues and I developed a measurement and analysis system that differentiates among the different forms and functions of aggressive behaviour (Little et al., 2001). Figure 12.3 displays this decomposition of aggressive behaviour in terms of its form (means) and function (ends). With this measurement and analysis system, we have been able to show new and novel relationships among the various aggression dimensions. For example, in the literature, proactive/instrumental measures of aggression typically correlate in the .6 to .8 region with reactive measures of aggression (Dodge & Coie, 1987; Poulin & Boivin, 1999; Price & Dodge, 1989). Our work has shown that, when one accounts for the form of the aggression (e.g., overt, relational), the instrumental and reactive functions of aggression are, in fact, independent (uncorrelated) processes. A number of commonalities, which emanate from the overarching organismic metatheoretical perspective, unite these different aspects of agentic action. First, because actions are coordinated in speci®c domains of goal-directed activity, they are domain speci®c. Second, because a speci®c action is chosen from among a number of possible actions, it is means

230

Little

speci®c. Third, because of the domain-speci®c and means-speci®c nature of actions in service of self-chosen goals, they can be classi®ed and categorised in terms of a means±ends, form±function analytic framework. Understanding the complexities of organismic activity clearly requires an understanding of how aspects of individuals such as their beliefs and behaviours coordinate together in the pursuit of life-course goals. As I see it, the linkages among the layered aspects of the agentic self are best examined when the operational nature of each layer stems from a common metatheoretical perspective, in this case, the organismic perspective. THE ROLE OF CONTEXT A contextual perspective on agency explicitly acknowledges that every developmental epoch represents a unique constellation of in¯uences (Bronfenbrenner, 1995; Lerner, 1995, 1996). Contexts provide both affordances and constraints on the range of individual actions and behaviours that can be expressed. Although some mechanisms of development are clearly universal (i.e., transcend context), many of the mechanisms that will ultimately account for the diversity of human behaviour across developmental contexts and the life course will probably be context-speci®c or context-moderated mechanisms (i.e., mechanisms that function in interaction with contextual features). When viewed in their entirety, both the preand post-millennial changes in the various social structures amount to unprecedented variability among the contexts in which development unfolds. As a result, a primary challenge to understanding the role of agency in development is to examine the organismic processes of change, growth, and transformation as they unfold within and between these varied contexts. As suggested earlier, a given context re¯ects a speci®c constellation of features (i.e., where intra-, inter-, and extra-personal features converge). These multilayered features can provide, if carefully contrasted, a quasiexperimental manipulation of many potential mechanisms that can shape and mould individual development. Given that the features in which developmental processes occur can be quasi-systematically varied, contextual designs can intimate the working mechanisms that in¯uence development and they can dramatically increase our universe of generalisation. In other words, comparing individual development across varying contexts can highlight which facets of the self-system are independent of the contextual variations and which are not (e.g., Little, 1998). Similarly, comparing the same individual across numerous contexts would reveal which processes of the agentic self function similarly (transcend context) and which function uniquely as a result of context-speci®c constraints or affordances (Mischel & Shoda, 1995). Well-conceived contextual studies can systematically vary and thereby quasi-experimentally isolate speci®c features of both contexts and individuals in order to derive the nature and generality

12. Agency in development

231

of their respective in¯uences on developmental outcomes (Hawley & Little, 1999). The dyad, for example, re¯ects a unique and informative context for study. Individuals carry with them orientations, response patterns, abilities, motivations, and so on. As soon as social behaviour begins, the characteristics of one individual collide with the characteristics of other individuals. Much can be learned about the development of the agentic self simply by exploring in detail these individual±individual interactions. With a general bias toward ``large-n studies'', developmentalists sometimes miss opportunities for rich ``small-n'' data designs such as small-group, roundrobin designs. Small groups allow for a microanalysis of the rich contexts of interaction among the different characteristic features of individuals. Research has shown that children will vary their behaviour according to the past interactions and behaviour patterns of their partners. Across the various dyadic interactions, a child can exhibit agency with one child and deference to another according to their history of wins and losses (see Hawley & Little, 1999). Although context-related effects can be readily documented, advances in understanding how context in¯uences individuals have been slow to emerge because of the reliance on context-level comparisons of groups of individuals. Rather than focus on documenting group-level differences in development across various contexts, a challenge of contextual approaches will be to attempt to reduce context to a set of context-level features and individual-differences variables that can effectively explain away the grouplevel differences of the individuals in a given context. The basic action plan involves ®rst de®ning the dimensions of context and second developing measures of those dimensions at both the level of the context and the level of the individual. By explicitly measuring the de®ning features of various contexts, greater precision in isolating the working mechanisms of change, growth, and transformation would be gained. For example, instead of characterising two sociocultural contexts as individualistic versus collectivistic, a researcher could measure the degree to which each individual in each context endorses individualistic and collectivistic values and ideals (i.e., one's personal balance of the cultural values). In this way, one could then examine the relative contribution of individual-level vs. context-level in¯uences on behaviour. The multilevel decomposition of variance from random coef®cient models (Kreft & De Leeuw, 1998; also known as Hierarchical Linear Models [HLM] analyses; see Bryk & Raudenbush, 1992) will be particularly useful for modelling the agentic self across context and time. The contextual perspective also encourages a typological approach to understanding individual development. Given that (a) few developmental processes are likely to operate in the same way for all individuals and that (b) the self is composed of synergistically intertwined layers, one could reasonably expect differing constellations of beliefs and behaviours that

232

Little

characterise subsets of individuals. These subgroups or typologies would also be characterised by a range of individual differences within them. In this sense, a typological pro®le would be considered a speci®c developmental context, with unique affordances and constraints on behaviour. For example, the pro®le of an individual who is both high on individualistic values and collectivistic values suggests constraints on behaviour that would not be the same for an individual who is low on both dimensions or high on one but medium or low on the other. Especially when typologies are more theory-driven rather than data-driven, a typological approach can be quite useful for understanding developmental processes (see, e.g., Hawley, Little, & Pasupahti, in press; Smith & Baltes, 1999). MODELLING THE AGENTIC SELF IN CONTEXT The contextual and organismic perspectives on agency necessitate the use of sophisticated statistical and methodological tools in order to disentangle (model) the various sources of in¯uence. A primary goal of a modelling perspective is to make strong tests of underlying theoretical models (i.e., bring implicit assumptions into the explicit realm of model speci®cation). The goal is not to partition the person into discrete sources of variance, but rather to determine how the pieces interlock and what their boundary conditions are and then to assemble the units into a coherent portrait of the individual. A modelling perspective is not limited to statistical procedures such as structural modelling techniques, it is equally at home within the framework of classical, regression-based techniques (see, for example, Widaman, 2000). The modelling perspective views statistical procedures not as mechanistic ends in themselves (i.e., ®xed and rule-bound), but rather as ¯exible tools that can be adjusted and adapted into an appropriate means for testing a given substantive theory. From a modelling vantage point, statistical methods are used to ®t models that are tested against observed data patterns and are implemented such that issues of effect size, parsimony, and theoretical integration are emphasised (cf. null hypothesis testing). Numerous recent techniques would ®t within this modelling framework. For example, dyadic, triadic, and social relations paradigms are extremely useful to model the interactions among individuals (see, e.g., Hawley & Little, 1999; Kenny, 1990). These designs allow decomposition of the systematic sources of variance that contribute to individual behaviour; they also allow decomposition of the degree to which individuals contribute to the behaviour of the larger unit (i.e., dyad, triad, family, and so on; see, for example, Kreppner, 1992). Multiple-group and multilevel designs are useful to compare and contrast the in¯uences of different contexts as well as to partition the degree to which variability in behaviour is associated with contextual features versus individual differences (Little, 1997; Little et al.,

12. Agency in development

233

2000). These techniques will be particularly useful to examine differences among groups and subgroups (i.e., typologies) of individuals. Because studies of the agentic self in context will often necessitate largescale approaches, the demands on participants can be great. However, recent advances in data imputation techniques and incomplete design estimation have tremendous potential for reshaping how we organise our data collection protocols. Various forms of intentionally missing designs can be conceived that would allow researchers to collect data in a more ef®cient and less costly manner and still be able to draw valid generalisations based on the whole sample of individuals and variables. For example, one could design a study of some large number of individuals, but where any one individual responds to only a subset of the items of interest or where only a randomly selected sub-sample of participants receives additional intensive measurements. With proper care, the missing subset of items or individuals can be imputed veridically (or the missing relationships estimated) such that conclusions and generalisations are drawn on the full set of individuals and variables (Graham & Hofer, 2000; McArdle, 1994). Clearly, at this stage, the boundaries and limitations of such designs still need careful testing, but the potential of such intentionally missing designs is certainly an appealing possibility for large-scale contextual-organismic research programmes. A second future development that will be useful in this regard is a robust technique to evaluate individual ®t of the models that we specify. For example, the Mx program calculates a psuedo-chi-squared value for each individual included in an analysis (see Neale, 2000). This information is useful because it can identify persons for whom the model does not hold. Further advances in this direction will offer information about the source of the mis®t and, ideally, provide an algorithm for clustering individuals based on their concordance with the ®t of the speci®ed model. Current multivariate analytic techniques generally only focus on two dimensions of the data cube at any one time, with one dimension being the focal dimension and the second being the variance-producing dimension (e.g., variables by persons, persons by variables, variables by occasions). Making multiple dimensions be the simultaneous focus of investigation will yield greater understanding of the complexities of the agentic self. Finally, recent advances in general structural equation modelling procedures are beginning to overcome one of the more frustrating limitations of this class of modelling technique: Namely, the restriction to linear relationships among continuous variables. Now, the procedures offer researchers the ability to specify latent interactions and nonlinear relationships as well as to model the relations among a mixture of different types of variables and constructs (see, e.g., Little et al., 2000; Marcoulides & Schumacker, 1996). Such advances will undoubtedly allow much greater precision in the manner by which theoretical propositions are speci®ed, estimated, and tested.

234

Little

SPECIFIC ISSUES A number of issues about the nature of development arise when one considers the organismic model as the metatheory. These issues highlight the probabilistic complexity of development that will continue to challenge our efforts to model the expressions of personal agency throughout development.

Equipotentiality versus omnipotentiality Cicchetti (e.g., Cicchetti & Rogosch, 1999) has championed this critical distinction in the ®eld of developmental psychobiology. Equipotentiality refers to the idea that many paths or antecedent conditions can lead to the same outcome or consequent condition. Omnipotentiality refers to the idea that common pathways and antecedent conditions can give rise to a plethora of outcomes or consequent conditions. Agentic choices and inputs throughout development are features likely to yield greater predictability in the varied directions and outcomes of developmental changes along the life course.

Heterotypic versus homotypic constructs The continuity of the nature of constructs across development poses signi®cant challenges to understanding the role of agency in development. Such challenges to measurement occur when a construct changes forms across development or when it takes on various forms across individuals, but still serves the same function. The nature and function of the layers of personal agency re¯ect a set of constructs that evolve in form throughout development. Similar to the methodological challenge of cross-cultural comparisons (Little, 1997), the challenge to developmentalists will be to develop measurement instruments and statistical tools to ensure the metric comparability of the dimensions of personal agency as they continue to transform across the lifespan. A related and relevant distinction is that between the surface-structure and the deep-structure meaning of behaviours. For example, resource control theory (Hawley, 1999) posits that the deep-structure goals of an evolutionarily adapting organism do not change across developmentÐthat is, the goal is optimal resource control. However, the surface-structure manifestations of that driving force change dramatically with age. In toddlerhood, coercive strategies of resource control are most common and effective. By early childhood and throughout development, more differentiated structures of resource control emerge such as prosocial strategies (Hawley, 1999).

12. Agency in development

235

Horizontal versus vertical cross-disciplinary integration Many researchers have cried out for greater interdisciplinary cooperation. At least two dimensions of cross-disciplinary integration can be brought to bear in understanding agency in development. The vertical dimension focuses on the up±down linkages from genes, through a speci®c class or type of behaviour, to higher levels of sociocultural constraints or in¯uences. From this perspective, a behaviour such as aggression would be examined at the level of hormones, the forms and functions of the behaviour, and the socio-cultural context that promotes or inhibits the degree of aggression. The horizontal dimension focuses on the lateral linkages spanning aspects of affect, behaviour, and cognition. From this perspective, a behaviour such as school performance would be examined by integrating various sources of in¯uence such as beliefs, motivational orientation, teacher-skills, parental support, and so on. In order to understand agency in development, research programmes will need to be highly collaborative and designs will need to be both more encompassing of the various micro-divisions of the ®eld, that is, more multidisciplinary in nature. Regarding collaborative work, a major goal of the next era should be to pool the intellectual and talent capital of numerous researchers in order to ensure that the large-scale research programmes are conducted well. For example, a specialist on control and motivation can join forces with a specialist in personality and a specialist on emotion regulation to design a study that does more than just explore the inter-relations among the separate categories of study, but rather, generates and tests meaningful hypotheses regarding the interface among the areas. Such approaches would advance our science by allowing us to contrast, reconcile, and integrate the plethora of microtheories that abound in developmental psychology. Cross-cultural approaches are just one manifestation of collaborative work. Another type are the interdisciplinary network studies that integrate various disciplines in the social and behavioural sciences and cross over into the biological and neurobiological realms (e.g., the Berlin Aging Study; Baltes & Mayer, 1999). The model of the network will certainly be one instantiation of the successful research programme in the years to come. Such a model will probably impact the culture of psychological enquiry. For example, it is likely that fewer studies will emerge with single-authored credits. Instead, the results of research teams and networks will continue to propagate our journals (e.g., Grob, Little, Wanner, Wearing, and EURONET, 1996). In my view, however, cross-disciplinary projects will suffer the cacophony of asynchronous instruments if there is no consensus or agreement about the nature of development. With a uni®ed view such as the agenticorganismic model outlined here, the many parts would be able to work in concert to yield a clear symphony of human development.

236

Little

Direct versus indirect and reciprocal effects The distinction between direct and indirect in¯uences on development has become more relevant with the advent of structural equation modelling (SEM) procedures. The traditional approach to identifying a mediator (Baron & Kenny, 1986) has given way to the SEM approach, which entails establishing a set of simultaneously estimated linkages (i.e., A to B to C). The magnitude and signi®cance of the parameters are directly interpreted. Mediation is tested by adding an additional path from A to C that bypasses B. If this path is nonsigni®cant and does not change the magnitude of the A to B and B to C parameter estimates, mediation is concluded. Moreover, the signi®cance of the indirect relationships between A and C is directly calculated as part of the procedures. However, full mediation is an idealised situation. In many applications, both indirect and direct effects can occur and partial mediation would be concluded. Proximal versus distal nature of the constructs is used to justify the placement or ordering of constructs. With time-order measurements, direct and indirect effects can be calculated to examine reciprocal effects among a set of constructs.

GENERAL IMPLICATIONS OF THE REVITALISED ORGANISMIC PERSPECTIVE A determined focus on the role of personal agency in development can yield a better understanding of the complexities of the relationships between context and self and how they in¯uence successful development along the life course. The type of research that I have outlined here lies at the cusp between basic and applied paradigms. Therefore, the results of studies that focus on the various themes I have presented have real potential to impact the ®eld. From a basic research perspective, they allow strong tests of competing theoretical models and provide important integrative implications. From an applied perspective, the mass of coordinated knowledge that potentially can be gained would allow policy-oriented programmes the needed basis for designing interventions that are maximally effective. Such interventions would thereby re¯ect the complexities and multiple-pathways in the development of the agentic self. A larger concern is clearly the degree to which we can make statements about causality in the context of quasi-experimental designs. As with any scienti®c method, threats to validity are numerous and, therefore, great caution is required. On the other hand, solid designs, good theory, and appropriate use of the analytic machinery that is available has the potential to advance our knowledge base at an incredible rate. Multiple levels of analysis techniques and the judicious use of experimentation procedures can alleviate many of the concerns. In order to reap fully the bene®ts of contextual-organismic studies, however, training programmes will need

12. Agency in development

237

to focus greater emphasis on both the basics and advances in quasiexperimental analytic techniques and expand instruction on correspondingly sophisticated designs. By explicitly examining the role of agency in development, we can identify those features of both individuals and contexts that maximally contribute to the agentic self and successful development along the life course. Understanding the developmental avenues that lead to the emergence of agentic and ef®cacious individuals can reap great rewards in terms of the productivity and well-being of individuals across the lifespan. ACKNOWLEDGEMENT Parts of this work were supported by a grant from Yale College, Yale University. REFERENCES Baltes, P.B., & Mayer, K.U. (Eds.). (1999). The Berlin Aging Study: Aging from 70 to 100. New York: Cambridge University Press. Bandura, A. (1997). Self-ef®cacy: The exercise of control. New York: Freeman. Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173±1182. BrandstaÈdter, J. (1984). Personal and social control over development: Some implications of an action perspective in lifespan developmental psychology. In P.B. Baltes & O.G. Brim, Jr. (Eds.), Lifespan development and behavior (pp. 1± 32). New York: Academic Press. Bronfenbrenner, U. (1995). Developmental ecology through space and time: A future perspective. In P. Moen & G.H. Elder, Jr. (Eds.), Examining lives in context: Perspectives on the ecology of human development (pp. 619±647). Washington, DC: American Psychological Association. Bryk, A.S., & Raudenbush, S.W. (1992). Hierarchical linear models. Newbury Park, CA: Sage. Cairns, R.B. (1983). The emergence of developmental psychology. In P.H. Mussen (Series Ed.) & W. Kessen (Vol. Ed.), Handbook of child psychology: Vol. 1. History, theory, and methods (4th ed., pp. 41±102). New York: John Wiley. Cairns, R.B. (1998). The making of developmental psychology. In W. Damon (Series Ed.) & R.M. Lerner (Vol. Ed.), Handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed., pp. 25±105). New York: John Wiley. Chapman, M. (1984). Intentional action as a paradigm for developmental psychology: A symposium. Human Development, 27, 113±114. Chapman, M., & Skinner, E.A. (1985). Actions in development/development in action. In M. Frese & J. Sabini (Eds.), Goal-direct behavior: The concept of action in psychology (pp. 199±213). Hillsdale, NJ: Lawrence Erlbaum Associates Inc.

238

Little

Chapman, M., Skinner, E.A., & Baltes, P.B. (1990). Interpreting correlations between children's perceived control and cogitive performance: Control, agency, or means±ends beliefs? Developmental Psychology, 26, 246±253. Cicchetti, D., & Rogosch, F.A. (1999). Conceptual and methodological issues in developmental psychopathology research. In P.C. Kendall & J.N. Butcher (Eds.), Handbook of research methods in clinical psychology (2nd ed., pp. 433±465). New York: John Wiley. Dodge, K.A., & Coie, J.D. (1987). Social-information-processing factors in reactive and proactive aggression in children's peer groups. Journal of Personality and Social Psychology, 53, 1146±1158. GarieÂpy, J-L. (1996). The question of continuity and change in development. In R.B. Cairns, G.H. Elder, Jr., & E.J. Costello (Eds.), Developmental science (pp. 78±96). New York: Cambridge University Press. Gottlieb, G. (1997). Commentary: A systems view of psychobiological development. In D. Magnusson (Ed.), The lifespan development of individuals: Behavioral, neurobiological, and psychosocial perspectives: A synthesis (pp. 76±103). New York: Cambridge University Press. Graham, J.W., & Hofer, S.M. (2000). Multiple imputation in multivariate research. In T.D. Little, K.U. Schnabel, & J. Baumert (Eds.), Modeling longitudinal and multilevel data: Practical issues, applied approaches, and speci®c examples (pp. 201±218). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Graham, S., & Juvonen, J. (1998). Self-blame and peer victimization in middle school: An attributional analysis. Developmental Psychology, 34, 587±538 Grob, A., Little, T.D., Wanner, B., Wearing, A.J., & EURONET. (1996). Adolescents' well-being and perceived control across fourteen sociocultural contexts. Journal of Personality and Social Psychology, 71, 785±795. Hawley, P.H. (1999). The ontogenesis of social dominance: A strategy-based evolutionary perspective. Developmental Review, 19, 91±132. Hawley, P.H., & Little, T.D. (1999). On winning some and losing some: Social dominance in toddlers. Merrill Palmer Quarterly, 45, 185±214. Hawley, P.H., & Little, T.D. (2002). Evolutionary and developmental perspectives on the agentic self. In D. Cervone & W. Mischel (Eds.), Advances in personality science (pp. 177±195). New York: Guilford Press. Hawley, P.H., Little, T.D., & Pasupahti, M. (in press). Winning friends and in¯uencing peers: Strategies of peer in¯uence in late childhood. International Journal of Behavioral Development. Kenny, D.A. (1990). Design issues in dyadic research. In C. Hendrick & M.S. Clark (Eds.), Review of personality and social psychology: Research methods in personality and social psychology, Vol. 11 (pp. 164±184). Newbury Park, CA: Sage. Kreft, I., & De Leeuw, J. (1998). Introducing multilevel modeling. Thousand Oaks, CA: Sage. Kreppner, K. (1992). Development in a developing context: Rethinking the family's role for children's development. In L.T. Winegar & J. Valsiner (Eds.), Children's development within social context: Research and methodology, Vol. 2 (pp. 161± 182). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Lerner, R.M. (1995). Developing individuals within changing contexts: Implications of developmental contextualism for human development research, policy, and programs. In T.A. Kindermann & J. Valsiner (Eds.), Development of personcontext relations (pp. 13±37). Hillsdale, NJ: Lawrence Erlbaum Associates Inc.

12. Agency in development

239

Lerner, R.M. (1996). Relative plasticity, integration, temporality, and diversity in human development: A developmental contextual perspective about theory, process, and method. Developmental Psychology, 32, 781±786. Lewis, M.D. (1997). Personality self-organization: Cascading constraints on cognition±emotion interaction. In A. Fogel, M.C.D.P. Lyra, & J. Valsiner (Eds.), Dynamics and indeterminism in developmental and social processes (pp. 193±216). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Little, T.D. (1997). Mean and covariance structures (MACS) analyses of crosscultural data: Practical and theoretical issues. Multivariate Behavioral Research, 32, 53±76. Little, T.D. (1998). Sociocultural in¯uences on the development of children's actioncontrol beliefs. In J. Heckhausen & C.S. Dweck (Eds.), Motivation and selfregulation across the life span (pp. 281±315). New York: Cambridge University Press. Little, T.D., Hawley, P.H., Henrich, C.C., & Marsland, K. (2002). Three views of the agentic self: A developmental synthesis. In R.M. Ryan & E.L. Deci (Eds.), Handbook of self-determination research (pp. 389±404). Rochester, NY: University of Rochester Press. Little, T.D., Jones, S.M., Henrich, C.C., & Hawley, P.H. (2001). Toward a uni®ed model of aggression: Differentiating form from function. Manuscript submitted for publication. Little, T.D., Lopez, D.F., & Wanner, B. (2001). Children's action-control behaviors (coping): A longitudinal validation of the behavioral inventory of strategic control. Anxiety, Stress, and Coping, 14, 315±336. Little, T.D., Oettingen, G., Stetsenko, A., & Baltes, P.B. (1995). Children's actioncontrol beliefs and school performance: How do American children compare with German and Russian children? Journal of Personality and Social Psychology, 69, 686±700. Little, T.D., Schnabel, K.U., & Baumert, J. (Eds.). (2000). Modeling longitudinal and multilevel data: Practical issues, applied approaches, and speci®c examples. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Little, T.D., & Wanner, B. (1997). The Multi-CAM: A multidimensional instrument to assess children's action-control motives, beliefs, and behaviors (Materialen aus der Bildungsforchung, No. 59, ISBN 3-87985-064-x). Berlin: Max Planck Institute for Human Development. Lopez, D.F., & Little, T.D. (1996). Children's action-control beliefs and emotional regulation in the social domain. Developmental Psychology, 32, 299±312. Magnusson, D., & Cairns, R.B. (1996) Developmental science: Toward a uni®ed framework. In R.B. Cairns, G.H. Elder, Jr., & J.E. Costello (Eds.), Developmental science (pp. 7±30). New York: Cambridge University Press. Marcoulides, G.A., & Schumacker, R.E. (Eds.). (1996). Advanced structural equation modeling: Issues and techniques. Mahwah, NJ: Lawrence Erlbaum Associates Inc. McArdle, J.J. (1994). Structural factor analysis experiments with incomplete data. Multivariate Behavioral Research, 29, 409±454. Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of personality: Reconceptualizing situations, dispositions, dynamics, and invariance in personality structure. Psychological Review, 102, 246±268. Neale, M.C. (2000). Individual ®t, heterogeneity, and missing data in multigroup structural equation modeling. In T.D. Little, K.U. Schnabel, & J. Baumert (Eds.),

240

Little

Modeling longitudinal and multilevel data: Practical issues, applied approaches, and speci®c examples (pp. 249±267). Mahwah, NJ: Lwarence Erlbaum Associates Inc. Oettingen, G., Little, T.D., Lindenberger, U., & Baltes, P.B. (1994). Causality, agency, and control beliefs in East versus West Berlin children: A natural experiment on the role of context. Journal of Personality and Social Psychology, 66, 579±595. Poulin, F., & Boivin, M. (1999). Proactive and reactive aggression and boys' friendship quality in mainstream classrooms. Journal of Emotional and Behavioral Disorders, 7, 168±177. Price, J.M., & Dodge, K.A. (1989). Reactive and proactive aggression in childhood: Relations to peer status and social context dimensions. Journal of Abnormal Child Psychology, 17, 455±471. Ryan, R.M. (1993) Agency and organization: Intrinsic motivation, autonomy, and the self in psychological development. In J.E. Jacobs (Ed.), Nebraska symposium on motivation, 1992: Developmental perspectives on motivation (Vol. 40, pp. 1±56). Lincoln, NE: University of Nebraska Press. Ryan, R.M., Sheldon, K.M., Kasser, T., & Deci, E.L. (1996). All goals are not created equal: An organismic perspective on the nature of goals and their regulation. In P.M. Gollwitzer & J.A. Bargh (Eds.), The psychology of action: Linking cognition and motivation to behavior (pp. 7±26). New York: Guilford Press. Shanahan, M., Valsiner, J., & Gottlieb, G. (1997). Developmental concepts across disciplines. In J. Tudge, M. Shanahan, & J. Valsiner (Eds.), Comparisons in human development: Understanding time and context (pp. 13±71). New York: Cambridge University Press. Skinner, E.A. (1995). Perceived control, motivation, and coping. Thousand Oaks, CA: Sage. Skinner, E.A. (1996). A guide to constructs of control. Journal of Personality and Social Psychology, 71, 549±570. Skinner, E.A., & Chapman, M. (1987). Resolution of a developmental paradox: How can perceived internality increase, decrease, and remain the same across middle childhood? Developmental Psychology, 23, 44±48. Skinner, E.A., Chapman, M., & Baltes, P.B. (1988). Children's beliefs about control, means-ends, and agency: Developmental differences during middle childhood. International Journal of Behavioral Development, 11, 369±388. Smith, J.S., & Baltes, P.B. (1999). Trends and pro®les of psychological functioning in very old age. In P.B. Baltes & K.U. Maier (Eds.), The Berlin Aging Study: Aging from 70 to 100 (pp. 197±226). New York: Cambridge University Press. Stetsenko, A., Little, T.D., Gordeeva, T.O., Grasshof, M., & Oettingen, G. (2000). Gender effects in children's beliefs about school performance: A cross-cultural study. Child Development, 71, 273±287. Widaman, K.F. (2000). Testing cross-group and cross-time constraints on parameters using the general linear model. In T.D. Little, K.U. Schnabel, & J. Baumert (Eds.), Modeling longitudinal and multilevel data: Practical issues, applied approaches, and speci®c examples (pp. 163±185). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Weisz, J.R. (1990). Development of control-related beliefs, goals, and styles in childhood and adolescence: A clinical perspective. In J. Rodin, C. Schooler, & W. Schaie (Eds.), Self-directedness: Cause and effects throughout the life course (pp. 19±49). Hillsdale, NJ: Lwarence Erlbaum Associates Inc.

13 The development of aggressive behaviour during childhood Richard E. Tremblay

University of Montreal, Canada

To the memory of Robert B. Cairns (1933±1999) The major issues of the present appear to be, in large measure, the same ones that thoughtful contributors to the science (of behavioural development) had addressed in the past. Cairns, 1983 (p. 90)

Paradoxically, the more human beings become ``civilised'', the more they appear to be preoccupied by violence. Research on the development and prevention of violent behaviour is a booming industry as we enter the new millennium. Aggressive behaviour is central to most theories of human behaviour in such varied disciplines as zoology, economics, and public health. In this chapter I describe what I believe we know about the development of aggression in childhood as well as what we do not know. WHAT DO WE MEAN BY THE DEVELOPMENT OF AGGRESSIVE BEHAVIOUR?

Aggressive or antisocial behaviour? Judging from the lack of attention to the de®nition of aggressive behaviour in numerous research papers, it may come as a surprise that the major problem with the ®eld could be a problem of de®nition. We have been putting our ®nger on this problem over and over again (e.g., Berkowitz, 1962; Burt, 1925; Buss, 1961; Cairns, 1979; Coie & Dodge, 1998; Hartup & deWit, 1974; Parke & Slaby, 1983; PitkaÈnen, 1969), but it regularly comes back to haunt us. For example, in the last twentieth-century edition of Carmichael's Handbook of child psychology, the reviewers of research on aggressive behaviour chose to break with the tradition of previous reviewers who had limited their discussion to research on aggression. Coie and Dodge

242

Tremblay

(1998) decided to broaden the review to antisocial behaviour and conduct disorder because ``the comorbidity of aggression with other antisocial behaviours suggests that an understanding of the etiology and developmental course of aggression might be enhanced by including aggression into the broader class of antisocial behaviour'' (p. 781). The aggregation of different types of aggressive behaviours, and the aggregation of aggressive behaviours with different forms of antisocial behaviour creates an important problem for a developmental science aiming to understand the origin and development of these behaviours. This problem can be observed by examining the content of the ``aggression'' scales that have been used over the past decades. The content of the scales de®nes what we are measuring and most of the popular scales contain a mix of behaviours that range from physical aggression to attention-seeking and disobedience. Consider the following items from one of the most frequently used ``aggressive'' rating scales for parents (Achenbach & Edelbrock, 1983): ``argues, brags, demands attention, disobeys, poor peer relations, jealous, lies, shows off, stubborn, moody, sulks, loud''. The common denominator of these items appears to be that they can be annoying. To the extent that an annoying or irritating person causes discomfort to others, and to the extent that ``aggressive'' means someone who causes discomfort to others, we could conclude that that person is ``aggressive''. But, should we classify such children in the same ``aggressive'' category as those who physically attack? The Achenbach and Edelbrock (1983) ``aggressive'' scale has 23 items with only 2 that refer clearly to physical aggression, and 2 others that could be interpreted as physical aggression. Such scales are regularly used to identify either ``aggressive'', ``externalising'', ``conduct problem'', or ``antisocial'' individuals in clinical practice. They have also been used to identify genetic in¯uences on phenotypes which are alternatively referred to as ``aggressive'', ``externalising'', ``conduct disordered'', ``delinquent'', or ``antisocial'' (e.g., Eley, Lichenstein, & Stevenson, 1999; O'Connor, McGuire, Reiss, Hetherington, & Plomin, 1998; Slutske et al., 1997; Taylor, Iacono, & McGue, 2000). Peer rating ``aggression'' scales were also created with the procedure used to create the parent rating scale described earlier, i.e., factor analyses of large pools of items. They led to similar results. The ``aggression'' scale for one of the few large longitudinal studies speci®cally aimed at understanding the development of aggressive behaviour (Huesmann, Eron, Lefkowitz, & Walder, 1984; Lefkowitz, Eron, Walder, & Huesmann, 1977) includes the following items: disobeys teacher, gives dirty looks, makes up stories and lies, does things that bother others, gets in trouble. Only 2 of the 10 items could be interpreted as physical aggression: starts ®ghts, pushes and shoves. Is aggressive behaviour always antisocial? Is the aim of research on aggressive behaviour to prevent the development of aggressive behaviour in the human species? These apparently naive questions highlight the

13. Development of aggressive behaviour 243 importance of clearly de®ning what we mean by aggressive behaviour and aggression. Studies of aggressive behaviour in animals other than humans have tended to use de®nitions based on the characteristics of the behaviour of the aggressive individual, i.e., topographical qualities of the behaviour rather than effect of the behaviour on the ``victim'' (Hartup & deWit, 1974). On the other hand, the most frequently used de®nition for research on the development of human aggressive behaviour over the last half of the twentieth century is a ``moral'' judgement approach where an observer decides that the behaviour he observed was or was not intended to be ``harmful to another person''.

Intent to harm Two recent reviews of research on aggressive behaviour in children and youth chose de®nitions that refer to the negative effect of the action on the ``victim''. Loeber and Stouthamer-Loeber (1998) chose a de®nition that refers to the negative impact of the act on the victim, without reference to intent: ``aggression is de®ned as those acts that in¯ict bodily or mental harm on others'' (p. 242). Coie and Dodge (1998) endorsed the Parke and Slaby (1983) de®nition, which refers to the possible negative consequence of the act, but stresses the intent: ``behaviour that is aimed at harming or injuring another person or persons'' (p. 550). Attribution of intent is particularly important when human aggression is considered socially de®ned. The methodological consequence to this decision is that research on human aggression must rely on the social judgement of an observer for ``intent to harm'' (e.g., Bandura, 1973; Walters & Parke, 1964). The ``intent'' criterion is especially problematic for studying the development of aggressive behaviour during early childhood. For example, Kagan (1974) argued ``that a young child cannot be aggressive until he has some psychic intention of injuring another'' (p. 109). He concluded that aggressive behaviour de®ned in this way starts ``well into the second year'' after birth, when a child can ``put himself into the psychic state of another''. One can easily imagine what would happen to research on aggressive behaviour in nonhuman animals if such a criterion were used. Does a dog intend to hurt when it bites another dog that tries to take its food away? Does a 12-month-old boy intend to hurt when he hits the peer who grabbed the toy in his hand? The age of ``onset'' of intent may have been clear to Kagan in 1974, but it is still a subject of debate (Diamond, Werker, & Lalonde, 1994; Flavell & Miller, 1998; Zelazo, Helwig, & Lau, 1996). The intent criterion is not only a problem for infants and nonhuman animals. Anger and fear lead to reactions that are clearly not under the control of one's will. It could be argued in fact that many, if not most, of the aggressive behaviours following intense frustration are impulsive behaviours that were not intended. Gray (Gray, 1971, 1982) proposed that a ``®ght±¯ight system'' controlled the behavioural reactions to unconditioned

244

Tremblay

punishment and nonreward. Lewis, Alessandri, and Sullivan (1990) showed that 4-month-old babies clearly expressed facial anger reactions to frustrations. These same reactions are expressed more clearly by the limbs of children a few months later when motor maturation enables the child to hit and kick. The careful description of the development of aggressive behaviours during early childhood should help us understand the social and moral value of these behaviours. If we a priori decide that aggressive behaviours cannot exist before a given age we of course prevent the falsi®cation of the hypothesis. Similarly, if we start by taking a moral stance to de®ne aggressive behaviour, we bias our observations.

Conclusion More than a quarter of a century ago Hartup and deWit (1974) concluded their discussion on the de®nitions of aggression by recommending that investigators adjust de®nitions to the problem they want to address. This recommendation was made after they had observed that investigators' de®nitions of aggression were too general. Bandura (1973) made the same recommendation: ``A high degree of speci®city is required at the investigatory level because there is little reason to believe that the diverse activities subsumed under the omnibus label `aggression', though sharing some ingredients in common, have the same determinants'' (p. 11). It is dif®cult to understand why research on the development of aggressive behaviour went in the opposite direction in the quarter century that followed these recommendations. Investigators not only aggregate largely different forms of behaviours that have traditionally been considered ``aggressive'', but they also tend to aggregate aggressive behaviours with socially disruptive behaviours such as opposition and hyperactivity, with illegal activities such as smoking marijuana, and with health-related risky behaviours such as not using a condom. There were of course some exceptions to the general rule. For example, Cairns and Cairns (1984) focused on physical aggression. Pulkkinen's (PitkaÈnen, 1969; Pulkkinen, 1983, 1987) pioneering work on offensive and defensive aggression was followed by Day, Bream, and Pal (1992), Dodge and Coie (1987), Dodge, Lochman, Harnish, Bates, and Pettit (1997), and Vitaro, Gendreau, Tremblay, and Oligny (1998). Olweus (1978, 1993) initiated an important tradition of research on bullying in schools (for a review see Smith & Brain, 2000). The importance of differentiating types of aggression can be seen very clearly in the comparisons of female and male aggression. Most studies of sex differences in human aggression up to the late 1980s concluded that males became much more aggressive than females after early childhood (e.g., Feshbach, 1970; Hartup, 1974; Maccoby & Jacklin, 1974, 1980). Following the work on indirect aggression among nonhuman female primates (see Hrdy, 1981; Hrdy & Williams, 1983), Lagerspetz, Bjorkqvist, and Peltonen

13. Development of aggressive behaviour 245 Intensity

Offensive Indirect

Direct

Defensive

Vectors of the modes of aggression (physical, verbal, mimic)

Figure 13.1 A descriptive model of aggression (PitkaÈnen, 1969).

(1988) showed that 11- and 12-year-old girls had higher levels of indirect aggression than boys. These results were con®rmed by a number of studies with younger and older subjects in different cultures: e.g., Bjorkqvist, È sterman, and Kaukiainen (1992), Crick, Casas, and Ku (1999), and O Tremblay et al. (1996a) (for a recent review of sex differences in aggression see Archer, 2000). Figure 13.1 is reproduced from the doctoral thesis of Lea Pulkkinen (PitkaÈnen, 1969). It clearly illustrates the different dimensions that some of the later work on aggression has attempted to assess, mainly during the school years.

ONSET OF AGGRESSIVE BEHAVIOUR AND INCREASE IN SERIOUSNESS

Onset The recent use of the word ``onset'' in studies of antisocial behaviour illustrates the strong impact the developmental perspective has had in this area of research during the last decades of the twentieth century. ``Onset'' has become one of the most fashionable words in the ®elds of developmental psychopathology and developmental criminology (e.g., American

246

Tremblay

Psychiatric Association, 1994; Farrington & Hawkins, 1991; Lahey et al., 1998; Loeber, Stouthamer-Loeber, & Green, 1991; Mof®tt, Caspi, Dickson, Silva, & Stanton, 1996; Tremblay, Pihl, Vitaro, & Dobkin, 1994). It can be used to describe the start of a phenomenon that has a relatively short duration, e.g., onset of a cold, and onset of a panic attack. From a developmental perspective the word onset generally refers to the age at which an individual ®rst starts to engage in a type of behaviour that will endure for a relatively long period of time. For example, we can talk of age of onset of standing on two feet without support, age of onset of using words, age of onset of walking and running, age of onset of self-awareness, etc. In criminology, the developmental perspective has led to the concepts of onset, participation, escalation, desistance, termination, and duration to describe criminal careers (see Blumstein, Cohen, Roth, & Visher, 1986; Farrington, Loeber, & Van Kammen, 1990; Loeber & Hay, 1997; Nagin & Land, 1993). The American Psychiatric Association (1994) introduced the concept of early and late onset conduct disorder for children in its latest version of the Diagnostic and Statistical Manual of Mental Disorders. The early onset form, i.e., before 10 years of age, appears to be characterised by physical aggression (Lahey et al., 1998). In their review of misconceptions concerning the development of aggression, Loeber and Stouthamer-Loeber (1998) conclude that ``the age of onset1 of aggression in male populations is not concentrated only in the preschool years'' (p. 246). They present data from their longitudinal study of boys from Pittsburgh which indicate that less than 5% of the boys had ``onset'' of minor aggression before age 5 years, whereas close to 40% had onset of minor aggression by 13 years of age (Figure 13.2). These results con®rm the general impression that as children grow older they become more physically aggressive. In his presidential address to the American Society of Criminology, Elliott (1994) presented similar results from a longitudinal study of a random sample of USA adolescents showing an increase in violent offending for boys and girls from 12 to 17 years of age. The apparent rapid increase in deviant behaviour during adolescence, which is followed by an equivalently rapid decrease, has been labelled the age±crime curve (e.g. Farrington, 1986). It was described by the Belgian mathematician-astronomer-biosocial scientist Adolphe Quetelet in his 1833 Research on the propensity for crime at different ages (Figure 13.3). Quetelet (1833, p. 65) concluded that ``Age is without contradiction the cause which acts with the most energy to develop or moderate the propensity for crime. This fatal propensity seems to develop in proportion to the intensity of physical strength and passions in man''. Late twentieth-century scientists have suggested that the rise in testosterone levels during adolescence explained both the increase in strength and in physically aggressive passions 1 I underline.

13. Development of aggressive behaviour 247 100

Cumulative percentage

80 Aggression Fighting Violence

60

40

20

0 3

4

5

6

7

8

9

10

11

12

13

14

15

16

Age of onset

Figure 13.2 Cumulative age of onset of different forms of aggression in the oldest sample of the Pittsburgh Youth Study (Loeber & Stouthamer-Loeber, 1998).

(e.g., Ellis & Coontz, 1990; Eysenck & Gudjonsson, 1989). The alternative explanation has of course been environmental. As children become older they are more and more in¯uenced by their environment, and are thus more likely to learn to aggress from such bad environmental in¯uences as deviant families, deviant peers, and the media (e.g. Dishion, Patterson, Stoolmiller, & Skinner, 1991; Eron & Huesmann, 1986; Farrington, 1998; Huesmann & Miller, 1994; McCord, 1991; Patterson, 1982; Vitaro, Tremblay, Kerr, Pagani-Kurtz, & Bukowski, 1997). The social learning approach to the development of aggressive behaviour has had a strong in¯uence since Bandura (1973) rallied the troops in reaction to the classic ethological perspective published by Konrad Lorenz a decade earlier. The English translation of the book On aggression (Lorenz, 1966) was a bestseller in North America. In this book Lorenz relied heavily on a phylogenetic perspective to convince readers that man had inherited an instinct for aggression, which he needed to learn to control to prevent self-destruction. Surprisingly, he did not pay much attention to the ontogenetic perspective. His prescriptions to prevent the perversions of the aggressive instinct appear to target mainly the behaviour of adults. While Lorenz argued that humans were naturally aggressive and had to learn ways to control these aggressive tendencies, Bandura argued that

248

Tremblay

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0–16 16–21 21–25 25–30 30–35 35–40 40–45 45–50 50–55 55–60 60–65 65–70 70–80

80+

Age

Figure 13.3 Male age±crime curve (Quetelet, 1833).

humans learned to aggress. He started his book by prescribing the developmentalist approach: ``A complete theory of aggression, whatever its orientation, must explain how aggressive patterns of behaviour are developed'' (p. 43). In the next chapter, logically titled ``Origins of aggression'' , he described experiments with nursery school children showing that they imitated adults who physically and verbally aggressed in¯ated dolls. These experiments appeared to satisfy Bandura's requirement to ``explain how aggressive patterns of behaviour are developed''. He concluded that the origin of aggression in man was social learning. Surprisingly, there is no reference in the book to the work on aggressive behaviour in infants and toddlers by child development pioneers like Bridges (1931), Dawe (1934), and Goodenough (1931), who are still cited at the end of the twentieth century. This debate has far-reaching consequences, not only for child development investigators and educators, but also for political scientists, philosophers, and law makers. Because the underlying debate is clearly our view of human nature, it is not surprising that investigators are likely to prefer the ``origin of aggressive behaviour'' that best ®ts their view of human nature, and their political commitment. But we still do not clearly understand the development of aggressive behaviour in the early years. This was a conclusion reached by Hartup and deWit (1974) a quarter of a century ago, and the progress made can be measured by the fact that Coie and Dodge (1998) had to refer to a cross-sectional study of 11 infants and

13. Development of aggressive behaviour 249 13 toddlers published almost 70 years ago (Goodenough, 1931) to brie¯y allude to the development of physical aggression in early childhood.

Increase in seriousness Systematic and unsystematic observations of single cases had long led philosophers (e.g., Hobbes, 1651/1958; Locke, 1693/1996) and child specialists (e.g., Sully, 1895) to the conclusion that young children react with anger, physically aggress, and need to be socialised. Part of the confusion concerning the concept of ``onset'' appears to come from the concept of ``seriousness''. Physical aggression is usually considered the most serious form of aggression. Courts will generally sanction more severely a physical aggression than a verbal one. Teachers and parents are more likely to do the same when they have to sanction a child who hit another and one who gave ``dirty looks''. Both observational data from small samples of children (Bridges, 1931; Dawe, 1934; Goodenough, 1931; Hartup, 1974; Murphy, 1937; Restoin et al., 1985; Strayer & Strayer, 1976; Strayer & Trudel, 1984) and recent parent reports, and self-reports of large samples (Choquet & Ledoux, 1994; Keenan & Wakschlag, 2000; Tremblay et al., 1996a, 1999) indicate that children start by physically aggressing during infancy and go on to verbal aggression once they have learned to talk. Figure 13.4 represents mothers' ratings of age at onset of physical attacks for a sample of 112 children assessed at age 17 and 30 months in a population survey of families in QueÂbec. At both points in time the mothers were asked if their child physically attacked others. If the response was positive we asked the age of the child when he ®rst physically attacked someone. This is the same procedure that Loeber and Stouthamer-Loeber (1998) used to collect data on age of onset of physical aggression with 13year-old boys. Results show that 20% of mothers report that their 17month-old child physically attacks others. By 30 months of age 25% of the children are reported to physically attack. This retrospective information indicates that some children initiate these behaviours towards the end of the ®rst year after birth, whereas others start during the second and the third year. The cumulative age of onset of physical aggression curves mirror those in Figure 13.2, but with major age of onset differences. The Pittsburgh data collected at 13 years of age indicated that less than 5% of 3year-olds had initiated minor physical aggressions. The QueÂbec data indicate that 80% of the children are reported by their mothers to use some form of physical aggression by 17 months of age (Tremblay, Japel et al., 1999). These major differences in the identi®cation of age of onset of physical aggression are probably due largely to the use of a retrospective procedure to collect the data on age of onset. Figure 13.4 shows that although 20% of the mothers report physical attacks at 17 months of age, when they are asked the same question at 30 months, less than 5% recall that their child had started to physically attack by 17 months. The recall

250

Tremblay

Cumulative percentage

30 25

17 months 30 months

20 15 10 5 0 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Age in months

Figure 13.4 Age at onset of physical aggression: attacks physically.

problem will clearly be greater when these mothers or their child will be asked to recall age of onset during adolescence. Parent reports in a sample of more than 22,000 children representative of the population of 0- to 11-year-old children in Canada in 1994 indicate that while the frequency of physical aggression decreases from the third to the eleventh year of life, indirect aggression increases from 4 to 8 years of age (Tremblay et al., 1996a). We will of course need longitudinal studies to con®rm that these results represent intra-individual change. However, at the group level, the form of aggressive behaviour that is generally considered more ``serious'', or ``socially unacceptable'' (physical aggression) is clearly ontogenetically antecedent to less ``serious'' forms of aggressive behaviour, such as verbal aggression or indirect aggression. The exciting work that needs to be done is to study the developmental trajectories of these aggressive behaviours at the intra-individual level during infancy and toddlerhood, and their relationship with later development. This work should be as exciting as the work on brain development during the ®rst years of life (see Cynader & Frost, 1999). Clearly, we cannot identify the origins of aggressive behaviour by focusing on the consequences of the acts. The focus on the severity of the consequences of aggressive behaviours is a legitimate area of study, but different from the study of the origins of aggressive behaviour. These two areas should lead to different measuring instruments. HOW STABLE IS AGGRESSIVE BEHAVIOUR? A strong interest in the stability of aggressive behaviour in the last half of the twentieth century preceded the search for the age of onset. Olweus

13. Development of aggressive behaviour 251 (1979) published one of the most often-cited reviews of research on the stability of aggressive behaviour. Based on longitudinal data (time 1 to time 2) from 16 samples of males he concluded that aggressive behaviour was as stable as intelligence. A similar review of the possibly hundreds of time 1 to time 2 correlations of aggressive behaviour scores that have been published since the Olweus review would most probably lead to the same conclusion. A few studies starting towards the end of infancy indicate that stability of physical aggression is already high in the pre-school years (Cummings, Iannotti, & Zahn-Waxler, 1989; Hay, Castle, & Davies, 2000; Keenan & Shaw, 1994). The concept of stability measured by relative position to others on an aggression scale is very different from the concept of stability measured by comparing the absolute level of aggression an individual has at two or more points in time. For example, different methodologies will be needed to answer the following two questions: (a) Did Paul increase the frequency of his physical aggressions from 2 to 15 years of age? (b) Was Paul more aggressive than Luc at 2 and 15 years of age? Much of the research on aggressive behaviour in the latter half of the twentieth century used a methodology that could address only the latter question. Although many studies were trying to ®nd the age of onset of aggression, very few attempted to describe the absolute changes in the development of speci®c forms of aggression. One notable exception was a longitudinal study by Cairns and colleagues (Cairns, Cairns, Neckerman, Fergusson, & GarieÂpy, 1989) in which annual assessments of physical aggression were used to describe the changes in the frequency of physical aggression for boys and girls from grade 4 to grade 12. Their results clearly showed that contrary to the age±crime curve discussed earlier, the frequency of physical aggression decreased systematically with age. Unfortunately, there do not appear to be similar studies during the preschool period that would enable us to understand the absolute stability of different forms of aggression during these years of rapid change. Similar studies on forms of aggressive behaviour other than physical aggression would also be useful to understand how physical aggression is transformed into more socially acceptable aggressive behaviours. We in fact have almost no knowledge of the development of socially positive aggressive behaviour (Pulkkinen, 1996). The description of mean levels of a given type of aggressive behaviour for a given population over time is only a ®rst step in understanding the ontogeny of aggressive behaviour. Most of the developmental work on aggressive behaviour has been based on the ``variable approach'' where the variable is the main unit of analysis, e.g. the mean scores of a variable measured at two or more points in time are compared or correlated. An alternative strategy is the ``person approach'' where the person is the main unit of analysis (Block, 1971; Cattell, 1965; Magnusson & Bergman, 1988). The aim of the analysis is to go beyond the mean of the group and identify

252

Tremblay 4

Physical aggression

3.5 3 2.5 2 1.5 1 0.5 0 6

10

11

12

13

14

15

Age

Chronic 4% High desister 28% Moderate desister 52% Low 17%

Figure 13.5 Trajectories of physical aggression for boys age 6 to 15.

categories of developmental trajectories. A number of suggestions have been made over the past decade concerning different trajectories or pathways leading to deviant forms of social development (e.g., Loeber, 1988; Mof®tt, 1993; Patterson, DeBaryshe, & Ramsey, 1989), but appropriate statistical procedures to test these hypotheses have been lacking. Nagin and Land (1993) devised a statistical procedure to study different types of criminal careers that appears to be a promising tool for identifying developmental trajectories. Applied to repeated measures of physical aggression on a large sample of boys followed from 6 to 15 years of age, the procedure enabled Nagin and Tremblay (1999) to identify four developmental trajectories of physical aggression. The four groups had four different levels of physical aggression in kindergarten (Figure 13.5). The group with the highest level was the smallest (4%) and maintained that level up to age 15. Similar results were obtained in several other instances (see Broidy et al., in press; Tremblay, 2000). This technique can be used to study trajectories of different types of behaviour simultaneously (Brame, Nagin, & Tremblay, 2001; Nagin & Tremblay, 2001). This will enable investigators to describe how two or more forms of aggressive behaviour develop differently over time in different groups of subjects. For example, one would expect to observe that some of the most physically aggressive infants would substantially reduce their use of physical aggression but increase

13. Development of aggressive behaviour 253 their use of verbal aggression, others would reduce physical aggression and increase indirect aggression, while still others would reduce physical aggression and not go on to engage in much verbal or indirect aggression. Understanding these different developmental trajectories will be extremely important to plan interventions that would help individuals follow trajectories leading to the prosocial use of their proneness to aggressive behaviour. The major problem in pushing the ®eld forward may not be one of sophisticated statistical procedure, but more a problem with having the appropriate data. There are very few longitudinal studies that have collected data on aggressive behaviour from infancy onwards, and those that have may not have collected the appropriate data. Most large-scale longitudinal studies ask a small set of questions on numerous variables and the data collection points are generally not closer than a year apart. We may need longitudinal studies that speci®cally target the early development of aggressive behaviour. UNDERSTANDING AND CHANGING THE COURSE OF DEVELOPMENT Although aggressive behaviour during the ®rst 5 years of life has not been studied enough to understand the early developmental course of different forms of aggression, the processes that lead to the different developmental courses, and their impact on the later developmental courses, much work in the last 25 years has attempted to understand the developmental processes leading to aggressive behaviour during the elementary school years and adolescence. Most of this work was inspired by the social learning model elaborated by Bandura (1973). Investigators focused on interactions within the family (e.g., Dumas & Wahler, 1985; Patterson, 1982; Wahler & Dumas, 1986), on interactions with peers (e.g., Boivin & Vitaro, 1995; Coie, Dodge, Terry, & Wright, 1991; Dishion, 1990; Dodge & Coie, 1987), and on the impact of the media (e.g., Eron, 1982; Huesmann & Eron, 1986; Huesmann & Miller, 1994). The work of Patterson and his colleagues at the Oregon Social Learning Center (e.g., Patterson, 1982; Patterson, Dishion & Bank, 1984) is a good example of the short-term longitudinal studies using detailed direct observations in attempts to understand how family interactions in¯uence the development of aggression. The investigators made repeated observations of the interactions among family members in families of pre-adolescents and adolescents referred for behaviour problems. The behaviours of family members, such as ``disapproves, negative commands, dependent, destroys, humiliates, ignores, physically negative, and noncompliant'' were recorded continuously during observation sessions by trained observers in the clinic or in the families' homes. Their conclusion from the analyses of thousands of

254

Tremblay

hours of observations was that parents' management of children's behaviour reinforces aggressive behaviour (see Patterson, 1982; Patterson, Capaldi, & Bank, 1991). These studies made important methodological contributions to the observation of social interactions that trigger aggressive behaviours at a proximal level, but they also showed the enormous effort that would be needed to apply these methodological innovations to understand the longterm processes that lead to different developmental trajectories of aggressive behaviour. In the 1980s, the social learning theory and cognitive theories were merged to create social-information-processing models of aggressive behaviour (Dodge, 1986; Huesmann, 1988; Rubin & Krasnor, 1986). These models, like the family interaction studies, were attempts to understand the proximal processes leading to an aggression. However, unlike the family interaction studies, the focus was not on behaviours that can be observed by the naked eye, but on the activity of the neurons in the subject's brain. Thus, subjects assessed as ``aggressive'' and ``not aggressive'' by adults or peers were asked to reveal how they would respond to hypothetical situations. Results have generally indicated that children labelled aggressive, conduct disordered, antisocial, and delinquent are less attentive to most social cues, they are more attentive to aggressive social cues, they make more frequent hostile interpretations of others' intentions, they generate less alternative responses to a speci®c situation, their alternative responses are qualitatively different, they respond more impulsively, and they tend to give a more positive value to an aggressive response (Coie & Dodge, 1998; Dodge, in press). Although the aim of these studies is meant to understand proximal causes of aggressive behaviours, the results are often used as support of the social learning theory concerning the origins of aggressive behaviour (Huesmann, 1998). The best way to address causal questions is through experiments. Longitudinal studies are the alternative when experiments are impossible, but causal questions never get satisfying answers from longitudinal studies. It is sometimes argued that longitudinal studies must be used with children because experiments are not ethical. We often forget that children are constantly the objects of interventions that are guided more by instinct than by sound knowledge of child development. This is especially the case with children and youth who tend to be disruptive. Preventive and corrective interventions are generally given to these children, and ethical considerations would suggest that the choice of these interventions should be based on evidence of their effectiveness. The process by which we establish the effectiveness of these interventions should be experimental, and thus offer an excellent opportunity to test causal hypotheses. Interventions to help violent individuals or prevent the development of chronic trajectories of physical aggression offer the opportunity to test hypotheses concerning the causes of aggression while searching for effective ways to help at-risk individuals.

13. Development of aggressive behaviour 255 Most interventions to prevent or reduce aggression are not done within an experimental procedure, but a number of well-designed randomised experiments with children to prevent the development of antisocial behaviour have been reported (e.g., Loeber & Farrington, 2001; McCord & Tremblay, 1992; Tremblay & Craig, 1995; Tremblay, LeMarquand, & Vitaro, 1999b; Wasserman & Miller, 1998). These studies have a long tradition. One of the ®rst, and one of the best designed, the Cambridge± Somerville study (McCord, 1978, 1992; Powers & Witmer, 1951), started in the 1930s. Although some of these studies have shown that interventions can have iatrogenic effects (Dishion, McCord, & Poulin, 1999; McCord, 1978), some have shown long-term positive effects (e.g., Hawkins, Farrington, & Catalano, 1998; Olds et al., 1998; Tremblay, Kurtz, MaÃsse, Vitaro, & Pihl, 1995; Vitaro, Brendgen, & Tremblay, 2001; Weikart & Schweinhart, 1992). However, to my knowledge, none have shown a signi®cant long-term impact on the most feared form of aggression. This may be because most have not used physical aggression as an outcome and, in keeping with the tradition of trying to address aggression in general, most of these intervention experiments did not speci®cally target physical aggression. Interventions with school-age children have targeted globally de®ned aggression, anger control, and disruptive behaviour (e.g., Hawkins et al., 1998; Kellam, Rebok, Ialongo, & Mayer, 1994; Lochman, 1992; Olweus, 1993; Tremblay, MaÃsse, Pagani, & Vitaro, 1996b; Vitaro, Brendgen, & Tremblay, 1999), whereas interventions with pre-school children tend to target parenting skills and cognitive development (e.g., Olds et al., 1998; Schweinhart, Barnes, & Weikart, 1993; Webster-Stratton, 1998). Parent training addresses discipline issues and the use of physical punishment, but could focus more speci®cally on the regulation of physical aggression during the pre-school years. If the regulation of physical aggression is learned during the pre-school years, one would expect that interventions speci®cally targeting this regulation during this sensitive period would prevent chronic trajectories of physical aggression and increase the likelihood of positive aggression. By targeting different hypothesised causal mechanisms, such as emotional regulation, executive functions, information processing, impulsivity, parental discipline, and peer in¯uence, we could test these hypotheses more directly than we can with longitudinal studies, and simultaneously ®nd the best preventive interventions. CONCLUSION Robert Cairns, an outstanding scientist who studied aggression and its development, summarised the state of knowledge in child development at the beginning of the twentieth century in the following way: ``For whatever reasons, the information available about longitudinal development by the

256

Tremblay

end of the ®rst period (1912) of the era's history was either sketchy (e.g., Binet's study of his two daughters) or subjective and retrospective (e.g., psychoanalytic interviews)'' (Cairns, 1983, p. 68). Longitudinal studies started in the following period (1913±1946) and increased linearly over the next half-century. Unfortunately, except for a few longitudinal studies like the one conducted by Cairns and Cairns (1994), the progress made in understanding the developmental course of aggressive behaviour was much less than it could have been because of problems with the de®nition of aggressive behaviour and because the ®rst years of development were not studied. From that perspective there has been less progress made in understanding the development of aggressive behaviour than Mills (1899) expected a century ago when he recommended longitudinal studies of individuals from birth to maturity. However, the lack of adequate data has not prevented a proliferation of general and speci®c theories concerning aggressive behaviour. Cairns's comment on the theories at the beginning of the century can be applied to the theories at the end of the century: ``it was on this fragmentary information that the most in¯uential psychoanalytic and behaviouristic theories of cognitive and personality development were formulated, and few data were available to assess their implications or correct their shortcomings'' (1983, p. 68). What should developmentalists do in the twenty-®rst century to understand the development of aggressive behaviour? Well, my answer to this question will not come as a surprise to readers who have had the time and patience to read this chapter from beginning to end. I hope that in 2100, we can publish a review of the past hundred years of research on aggression in which the author will report that hundreds of studies, in dozens of cultures all over the world, will have followed the developmental course of different types of aggressive behaviours from the ®rst year after birth to late adulthood for hundreds of thousands of subjects. The author of that essay should be in a position to describe numerous life-course developmental trajectories of aggressive behaviour replicated across cultures. He or she will probably remind the reader that developmentalists completed the basic descriptive work needed to understand the origin, development, and expression of human aggression almost a hundred years after geneticists completed the basic descriptive work of the human genome. The 2100 reviewer will certainly note that a complete description of the human genome was an easy task once the technology was available, whereas a lifespan description of developmental trajectories of aggression was less a matter of having the technology than the collective will, perseverance, and humility of the community of behaviour developmentalists to pool their resources and create large international longitudinal projects on the development of the key behaviour problems that have major impacts on health and well-being, and thus on the human, social, and ®nancial capital of human societies. The author of that essay will also remind the readers

13. Development of aggressive behaviour 257 that the study of genetic and environmental effects on aggressive behaviour could not be adequately studied until the phenotypes were well described. He will most certainly recall that at the end of the twentieth century and beginning of the twenty-®rst there was a strong movement to study the genetic-environment correlations and interactions for social, cognitive, and emotional development (see Plomin & Rutter, 1998; Rutter, 1996), which, concerning the development of aggression, was handicapped by an equally strong movement to aggregate all forms of socially obnoxious behaviours into fuzzy labels such as ``antisocial'', ``externalising'', and ``disruptive''. I also hope that the author of the essay will be able to report that important progress has been made in the prevention of chronic physical violence. I expect that he will remind readers that the major advances in preventing chronic physical violence were made through alliances of the large-scale longitudinal studies, described earlier, with service delivery agencies to experiment different types of prevention programmes at different periods of development. The author of that essay will certainly stress the fact that these experiments led not only to the identi®cation of effective prevention programmes, but also to the identi®cation of important causal factors of chronic aggression for certain types of individuals. My strongest conviction is that it will be reported that the most effective interventions were very intensive, started prior to birth (still better if before conception) and, in some cases included permanent support, similar to support for diabetics. Finally, I hope that research on aggression in the twenty-®rst century will address and resolve the issue of prosocial aggressiveness. The latin origin of the verb ``aggress'' (ad gradi) means ``go towards''. Some dictionaries (e.g., American Heritage Dictionary, 1985; Rodale, 1978) give two different meanings to the adjective ``aggressive'': 1. hostile; 2. assertive, bold, enterprising. Other dictionaries (e.g. Oxford English Dictionary, 1981) limit the meaning to a disposition to attack. The challenge for the socialisation process is to foster assertive, bold, and enterprising behaviour with respect for the rights of others, rather than hostile behaviour. The challenge for investigators in the area of aggressive behaviour is ®rst, to discover whether hostile behaviour has the same developmental origin as assertive, bold, and enterprising behaviour, and second, to identify how children can be helped to learn to be assertive without being hostile. ACKNOWLEDGEMENTS The following institutions have supported my work on the development of social behaviour: the Canadian Donner Foundation, the Canadian Institute for Advanced Research, le Centre de Recherche Fernand-Seguin, le Conseil QueÂbeÂcois pour la Recherche Sociale, les Fonds pour la Formation des Chercheurs et l'Aide aÁ la Recherche du QueÂbec, le Fonds pour la Recherche

258

Tremblay

en Sante du QueÂbec, the MacArthur Foundation, the Medical Research Council of Canada, the Molson Foundation, the National Consortium for Violence Research, the National Health Research Development Program of Canada, the Social Sciences and Humanities Research Council of Canada, the University of Montreal. REFERENCES Achenbach, T.M., & Edelbrock, C. (1983). Manual for the Child Behavior Checklist and Revised Child Behavior Pro®le. Burlington, VT: University of Vermont, Department of Psychiatry. American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Archer, J. (2000). Sex differences in aggression between heterosexual partners: A meta-analytic review. Psychological Bulletin, 126, 651±680. Bandura, A. (1973). Aggression: A social learning analysis. New York: Holt. Berkowitz, L. (1962). Aggression: A social psychological analysis. New York: McGraw-Hill. È sterman, K., & Kaukiainen, A. (1992). The development of direct Bjorkqvist, K., O and indirect aggressive strategies in males and females. In K. BjoÈrkqvist & P. Niemelae (Eds.), Of mice and women: Aspects of female aggression (pp. 51±64). San Diego, CA: Academic Press. Block, J. (1971). Lives through time. Berkeley, CA: Bancroft Books. Blumstein, A., Cohen, J., Roth, J.A., & Visher, C.A. (Eds.). (1986). Criminal careers and ``career criminals''. Washington, DC: National Academy Press. Boivin, M., & Vitaro, F. (1995). The impact of peer relationships on aggression in childhood: Inhibition through coercion or promotion through peer support. In J. McCord (Ed.), Coercion and punishment in long-term perspectives (pp. 183±197). New York: Cambridge University Press. Brame, B., Nagin, D.S., & Tremblay, R.E. (2001). Developmental trajectories of physical aggression from school entry to late adolescence. The Journal of Child Psychology and Psychiatry, 58, 389±394. Bridges, K.M.B. (1931). The social and emotional development of the pre-school child. London: Kegan Paul. Broidy, L.M., Nagin, D.S., Tremblay, R.E., Brame, B., Dodge, K., Fergusson, D., Horwood, J., Loeber, R., Laird, R., Lynam, D., Mof®tt, T., Bates, J.E., Pettit, G.S., & Vitaro, F. (in press). Developmental trajectories of childhood disruptive behaviors and adolescent delinquency: A six site, cross national study. Developmental Psychology. Burt, C.L. (1925). The young delinquent. London: University of London Press. Buss, A.H. (1961). The psychology of aggression. New York: John Wiley. Cairns, R.B. (1979). The analysis of social interactions: Method, results and illustrations. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Cairns, R.B. (1983). The emergence of developmental psychology. In P.H. Mussen (Ed.), Handbook of child psychology, Vol. 1 (pp. 41±102). Toronto: John Wiley. Cairns, R.B., & Cairns, B.D. (1984). Predicting aggressive patterns in girls and boys: A developmental study. Aggressive Behavior, 10, 277±242.

13. Development of aggressive behaviour 259 Cairns, R.B., & Cairns, B.D. (1994). Life lines and risks: Pathways of youth in our time. New York: Cambridge University Press. Cairns, R.B., Cairns, B.D., Neckerman, H.J., Fergusson, L.L., & GarieÂpy, J.L. (1989). Growth and aggression: 1. Childhood to early adolescence. Developmental Psychology, 25, 320±330. Cattell, R.B. (1965). The scienti®c analysis of personality. Chicago: Aldine. Choquet, M., & Ledoux, S. (1994). Adolescents: EnqueÃte nationale. Paris: Les  ditions INSERM. E Coie, J.D., & Dodge, K.A. (1998). Aggression and antisocial behavior. In W. Damon & N. Eisenberg (Eds.), Handbook of child psychology: Vol. 3. Social, emotional, and personality development (pp. 779±862). New York: John Wiley. Coie, J.D., Dodge, K.A., Terry, R., & Wright, V. (1991). The role of aggression in peer relations: An analysis of aggression episodes in boys' play groups. Child Development, 62, 812±826. Crick, N.R., Casas, J.F., & Ku, H.-C. (1999). Relational and physical forms of peer victimization in preschool. Developmental Psychology, 35, 376±385. Cummings, E.M., Iannotti, R.J., & Zahn-Waxler, C. (1989). Aggression between peers in early childhood: Individual continuity and developmental change. Child Development, 60, 887±895. Cynader, M., & Frost, B. (1999). Mechanisms of brain development: Neuronal sculpting by the physical and social environment. In D. Keating & C. Hertzman (Eds.), Developmental health and the wealth of nations: Social, biological, and educational dynamics (pp. 153±184). New York: Guilford Press. Dawe, H.C. (1934). An analysis of 200 quarrels of preschool children. Child Development, 5, 139±157. Day, D.M., Bream, L.A., & Pal, A. (1992). Proactive and reactive aggression: An analysis of subtypes based on teacher perceptions. Journal of Clinical Child Psychology, 21, 210±217. Diamond, A., Werker, J.F., & Lalonde, C. (1994). Toward understanding commonalities in the development of object search, detour navigation, categorization, and speech perception. In G. Dawson & K.W. Fischer (Eds.), Human behavior and the developing brain (pp. 380±426). New York: Guilford Press. Dishion, T.J. (1990). Peer context of troublesome behavior in children and adolescents. In P.E. Leone (Ed.), Understanding troubled and troublesome youth (pp. 128±153). Beverly Hills, CA: Sage. Dishion, T.J., McCord, J., & Poulin, F. (1999). Iatrogenic effects in early adolescent interventions that aggregate peers. American Psychologist, 54, 755±764. Dishion, T.J., Patterson, G.R., Stoolmiller, M., & Skinner, M. L. (1991). Family, school, and behavioral antecedents to early adolescent involvement with antisocial peers. Developmental Psychology, 27, 172±180. Dodge, K.A. (1986). A social-information-processing model of social competence in children. In M. Perlmutter (Ed.), Minnesota symposium on child psychology, Vol. 18 (pp. 77±125). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Dodge, K.A. (in press). Conduct disorder. In M. Lewis & A.J. Sameroff (Eds.), Handbook of developmental psychopathology (2nd ed.). New York: Plenum Press. Dodge, K.A., & Coie, J.D. (1987). Social information processing factors in reactive and proactive aggression in children's peer groups. Journal of Personality and Social Psychology, 53, 1146±1158. Dodge, K., Lochman, J.E., Harnish, J.D., Bates, J.E., & Pettit, G.S. (1997).

260

Tremblay

Reactive and proactive aggression in school children and psychiatrically impaired chronically assaultive youth. Journal of Abnormal Psychology, 106, 37±51. Dumas, J.E., & Wahler, R.G. (1985). Indiscriminate mothering as a contextual factor in aggressive-oppositional child behavior: ``Damned if you do and damned if you don't''. Journal of Abnormal Child Psychology, 13, 1±17. Eley, T.C., Lichenstein, P., & Stevenson, J. (1999). Sex differences in the etiology of aggressive and nonaggressive antisocial behavior: Results from two twin studies. Child Development, 70, 155±168. Elliott, D.S. (1994). Serious violent offenders: Onset, developmental course and termination: The American Society of Criminology 1993 Presidential Address. Criminology, 32, 1±21. Ellis, L., & Coontz, P.D. (1990). Androgens, brain functioning, and criminality: The neurohormonal foundations of antisociality. In L. Ellis & H. Hoffman (Eds.), Crime in biological, social and moral contexts (pp. 162±193). New York: Praeger. Eron, L.D. (1982). Parent±child interaction, television violence, and aggression of children. American Psychologist, 37, 197±211. Eron, L.D., & Huesmann, L.R. (1986). The role of television in the development of prosocial and antisocial behavior. In D. Olweus, J. Block, & M. Ratkie-Iarrow (Eds.), The development of prosocial and antisocial behavior (pp. 285±314). New York: Academic Press. Eysenck, H.J., & Gudjonsson, G.H. (1989). The causes and cures of criminality. New York: Plenum Press. Farrington, D.P. (1986). Age and crime. In M. Tonry & N. Morris (Eds.), Crime and justice: An annual review of research, Vol. 7 (pp. 189±250). Chicago: University of Chicago Press. Farrington, D.P. (1998). Individual differences and offending. In M.H. Tonry (Ed.), The handbook of crime and punishment (pp. 241±268). New York: Oxford University Press. Farrington, D.P., & Hawkins, J.D. (1991). Predicting participation, early onset and later persistence in of®cially recorded offending. Criminal Behaviour and Mental Health, 1, 1±33. Farrington, D.P., Loeber, R., & Van Kammen, W.B. (1990). The long-term criminal outcomes of conduct problem boys with or without impulsive-inattentive behavior. In L.N. Robins & M. Rutter (Eds.), Straight and devious pathways from childhood to adulthood (pp. 62±81). New York: Cambridge University Press. Feshbach, S. (1970). Aggression. In P.H. Mussen (Ed.), Carmichael 's manual of child psychology, Vol. 2 (3rd ed.). New York: John Wiley. Flavell, J.H., & Miller, P.H. (1998). Social cognition. In D. Kuhn & R.S. Siegler (Eds.), Handbook of child psychology, Vol. 2: Cognition, perception, and language (pp. 851±898). New York: John Wiley. Goodenough, F.L. (1931). Anger in young children. Westport, CT: Greenwood Press. Gray, J.A. (1971). The psychology of fear and stress. London: Weidenfeld & Nicolson. Gray, J.A. (1982). The neuropsychology of anxiety. New York: Oxford University Press. Hartup, W.W. (1974). Aggression in childhood. American Psychologist, 29, 336±341. Hartup, W.W., & deWit, J. (1974). The development of aggression: Problems and

13. Development of aggressive behaviour 261 perspectives. In J. deWit & W.W. Hartup (Eds.), Determinants and origins of aggressive behavior (pp. 595±620). The Hague, Netherlands: Mouton. Hawkins, J.D., Farrington, D.P., & Catalano, R.P. (1998). Reducing violence through the schools. In D.S. Elliott, B.A. Hamburg, et al. (Eds.), Violence in American schools: A new perspective (pp. 188±216). New York: Cambridge University Press. Hay, D.F., Castle, J., & Davies, L. (2000). Toddlers' use of force against familiar peers: A precursor of serious aggression? Child Development, 71, 457±467. Hobbes, T. (1651/1958). Leviathan. Indianapolis, IN: Liberal Arts Press. Hrdy, S.B. (1981). The woman that never evolved. Cambridge, MA: Harvard University Press. Hrdy, S.B., & Williams, G.C. (1983). Behavioral biology and the double standard. In S.B. Hrdy & G.C. Williams (Eds.), Social behavior of female vertebrates. New York: Academic Press. Huesmann, L.R. (1988). An information processing model for the development of aggression. Aggressive Behavior, 14, 13±24. Huesmann, L.R. (1998). An information processing theory for understanding the interaction of emotions and cognitions in the development and instigation of aggressive behavior. Presidential Address, International Society for Research on Aggression. Ramapo College, New Jersey, July 1998. Huesmann, L.R., & Eron, L.D. (1986). Television and the aggressive child: A crossnational comparison. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Huesmann, L.R., Eron, L.D., Lefkowitz, M.M., & Walder, L.O. (1984). Stability of aggression over time and generations. Developmental Psychology, 20, 1120±1134. Huesmann, L.R., & Miller, L.S. (1994). Long-term effects of repeated exposure to media violence in childhood. In L.R. Huesmann (Eds.), Aggressive behavior: Current perspectives (pp. 153±186). New York: Plenum Press. Kagan, J. (1974). Development and methodological considerations in the study of aggression. In J. deWit & W.W. Hartup (Eds.), Determinants and origins of aggressive behavior (pp. 107±114). The Hague, Netherlands: Mouton. Keenan, K., & Shaw, D.S. (1994). The development of aggression in toddlers: A study of low-income families. Journal of Abnormal Child Psychology, 22, 53±77. Keenan, K., & Wakschlag, L.S. (2000). More than the terrible twos: The nature and severity of behavior problems in clinic-referred preschool children. Journal of Abnormal Child Psychology, 28, 33±46. Kellam, S.G., Rebok, G.W., Ialongo, N., & Mayer, L.S. (1994). The course and malleability of aggressive behavior from early ®rst grade into middle school: Results of a developmental epidemiologically based preventive trial. Journal of Child Psychology and Psychiatry, 35, 259±281. Lagerspetz, K.M., Bjorkqvist, K., & Peltonen, T. (1988). Is indirect aggression typical of females: Gender differences in aggressiveness in 11- to 12-year-old children. Aggressive Behavior, 14, 403±414. Lahey, B.B., Loeber, R., Quay, H.C., Applegate, B., Shaffer, D., Waldman, I., Hart, E.L., McBurnett, K., Frick, P.J., Jensen, P.S., Dulcan, M.K., Canino, G., & Bird, H.R. (1998). Validity of DSM-IV subtypes of conduct disorder based on age of onset. Journal of the American Academy of Child and Adolescent Psychiatry, 37, 435±442. Lefkowitz, M.M., Eron, L.D., Walder, L.O., & Huesmann, L.R. (1977). Growing up

262

Tremblay

to be violent. A longitudinal study of the development of aggression. New York: Pergamon Press. Lewis, M., Alessandri, S.M., & Sullivan, M.W. (1990). Violation of expectancy, loss of control, and anger expressions in young infants. Developmental Psychology, 26, 745±751. Lochman, J.E. (1992). Cognitive-behavioral intervention with aggressive boys: Three-year follow-up and preventive effects. Journal of Consulting and Clinical Psychology, 60, 426±432. Locke, J. (1693/1996). Some thoughts concerning education. Indianapolis, IN: Hackett. Loeber, R. (1988). Natural histories of conduct problems, delinquency, and associated substance use: Evidence for developmental progressions. In B.B. Lahey & A.E. Kazdin (Eds.), Advances in clinical child psychology, Vol. 11 (pp. 73±124). New York: Plenum Press. Loeber, R., & Farrington, D.P. (Eds.). (2001). Child delinquents: Development, interventions and service needs. Thousand Oaks, CA: Sage. Loeber, R., & Hay, D.F. (1997). Key issues in the development of aggression and violence from childhood to early adulthood. Annual Review of Psychology, 48, 371±410. Loeber, R., & Stouthamer-Loeber, M. (1998). Development of juvenile aggression and violence. Some common misconceptions and controversies. American Psychologist, 53, 242±259. Loeber, R., Stouthamer-Loeber, M., & Green, S.M. (1991). Age at onset of problem behaviour in boys, and later disruptive and delinquent behaviours. Criminal Behaviour and Mental Health, 1, 229±246. Lorenz, K. (1966). On aggression. New York: Harcourt, Brace & World. Maccoby, E.E., & Jacklin, C.N. (1974). The psychology of sex differences. Stanford, CA: Stanford University Press. Maccoby, E.E., & Jacklin, C.N. (1980). Sex differences in aggression: A rejoinder and reprise. Child Development, 51, 964±980. Magnusson, D., & Bergman, L.R. (1988). Individual and variable-based approaches to longitudinal research in early risk factors. In M. Rutter (Ed.), Studies of psychosocial risk: The power of longitudinal data (pp. 45±61). New York: Cambridge University Press. McCord, J. (1978). A thirty-year follow-up of treatment effects. American Psychologist, 33, 284±289. McCord, J. (1991). Family relationships, juvenile delinquency, and adult criminality. Criminology, 29, 397±417. McCord, J. (1992). The Cambridge±Somerville Study: A pioneering longitudinal experimental study of delinquency prevention. In J. McCord & R.E. Tremblay (Eds.), Preventing antisocial behavior: Interventions from birth through adolescence (pp. 196±206). New York: Guilford Press. McCord, J., & Tremblay, R.E., (Eds.). (1992). Preventing antisocial behavior: Interventions from birth through adolescence. New York: Guilford Press. Mills, W. (1899). The nature of animal intelligence and the methods of investigating it. Psychological Review, 6, 262±274. Mof®tt, T.E. (1993). Adolescence-limited and life-course persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674±701. Mof®tt, T.E., Caspi, A., Dickson, N., Silva, P.S., & Stanton, W. (1996). Childhood-

13. Development of aggressive behaviour 263 onset versus adolescent-onset antisocial conduct problems in males: Natural history from ages 3 to 18 years. Development and Psychopathology, 8, 399±424. Murphy, L.B. (1937). Social behavior and child personality. New York: Columbia University Press. Nagin, D.S., & Land, K.C. (1993). Age, criminal careers, and population heterogeneity: Speci®cation and estimation of a nonparametric, mixed Poisson model. Criminology, 31, 327±362. Nagin, D.S., & Tremblay, R.E. (1999). Trajectories of boys' physical aggression, opposition, and hyperactivity on the path to physically violent and nonviolent juvenile delinquency. Child Development, 70, 1181±1196. Nagin, D.S., & Tremblay, R.E. (2001). Analyzing developmental trajectories of distinct but related behaviors: A group-based method. Psychological Methods, 6, 18±34. O'Connor, T.G., McGuire, S., Reiss, D., Hetherington, M.E., & Plomin, R. (1998). Co-occurrence of depressive symptoms and antisocial behavior in adolescence: A common genetic liability. Journal of Abnormal Psychology, 107, 27±37. Olds, D., Henderson, C.R., Cole, R., Eckenrode, J., Kitzman, H., Luckey, D., Pettitt, L., Sidora, K., Morris, P., & Powers, J. (1998). Long-term effects of nurse home visitation on children's criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial. Journal of the American Medical Association, 280, 1238±1244. Olweus, D. (1978). Aggression in the schools. Washington, DC: Hemisphere. Olweus, D. (1979). Stability of aggressive reaction patterns in males: A review. Psychological Bulletin, 85, 852±875. Olweus, D. (1993). Bullying at school: What we know and what we can do. Oxford: Blackwell. Parke, R.D., & Slaby, R.G. (1983). The development of aggression. In P.H. Mussen (Eds.), Handbook of child psychology: Socialization, personality and social development, Vol. 4 (pp. 547±641). New York: John Wiley. Patterson, G.R. (1982). A social learning approach to family intervention: III. Coercive family process. Eugene, OR: Castalia. Patterson, G.R., Capaldi, D., & Bank, L. (1991). An early starter model for predicting delinquency. In D.J. Pepler & K.H. Rubin (Eds.), The development and treatment of childhood aggression (pp. 139±168). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Patterson, G.R., DeBaryshe, B.D., & Ramsey, E. (1989). A developmental perspective on antisocial behavior. American Psychologist, 44, 329±335. Patterson, G.R., Dishion, T.J., & Bank, L. (1984). Family interaction: A process model of deviancy training. Aggressive Behavior, 10, 253±267. PitkaÈnen, L. (1969). A descriptive model of aggression and nonaggression with applications to children's behaviour. JyvaÈskylaÈ Studies in Education, Psychology and Social Research, No. 19. JyvaÈskylaÈ: University of JyvaÈskylaÈ. Plomin, R., & Rutter, M. (1998). Child development, molecular genetics, and what to do with genes once they are found. Child Development, 69, 1223±1242. Powers, E., & Witmer, H. (1951). An experiment in the prevention of delinquency: The Cambridge±Somerville Youth Study. Montclair, NJ: Columbia University Press. Pulkkinen, L. (1983). Finland: The search for alternatives to aggression. In A.P.

264

Tremblay

Goldstein & M.H. Segal (Eds.), Aggression in global perspective, Vol. 1 (pp. 104± 144). New York: Pergamon Press. Pulkkinen, L. (1987). Offensive and defensive aggression in humans: A longitudinal perspective. Aggressive Behavior, 13, 197±212. Pulkkinen, L. (1996). Proactive and reactive aggression in early adolescence as precursors to anti- and prosocial behavior in young adults. Aggressive Behavior, 22, 241±257. Quetelet, A. (1833). Research on the propensity for crime at different ages. Brussels: M. Hayez, Printer to the Royal Academy. Restoin, A., Montagner, H., Rodriguez, D., Girardot, J.J., Laurent, D., Kontar, F., Ullmann, V., Casagrande, C., & Talpain, B. (1985). Chronologie des comportements de communication et pro®ls de comportement chez le jeune enfant. In R.E. Tremblay, M.A. Provost, & F.F. Strayer (Eds.), Ethologie et deÂveloppement de l 'enfant (pp. 93±130). Paris: Editions Stock/Laurence Pernoud. Rubin, K.H., & Krasnor, L.R. (1986). Social-cognitive and social-behavior perspectives on problem-solving. In M. Perlmutter (Ed.), The Minnesota Symposium on Child Psychology, Vol. 18 (pp. 1±68). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Rutter, M. (Ed.). (1996). Genetics of criminal and antisocial behaviour. Chichester, UK: John Wiley. Schweinhart, L.L., Barnes, H.V., & Weikart, D.P. (1993). Signi®cant bene®ts. The High/Scope Perry School Study through age 27. Ypsilanti, MI: High/Scope Press. Slutske, W.S., Heath, A.C., Dunne, M.P., Statham, D.J., Dinwiddie, S.H., Madden, P.A.F., Martin, N.G., & Bucholz, K.K. (1997). Modeling genetic and environmental in¯uences in the etiology of conduct disorder: A study of 2,682 adult twin pairs. Journal of Abnormal Psychology, 106, 266±279. Smith, P.K., & Brain, P. (2000). Bullying in schools: Lessons from two decades of research. Aggressive Behavior, 26, 1±9. Strayer, F.F., & Strayer, J. (1976). An ethological analysis of social agonism and dominance relations among preschool children. Child Development, 47, 980±988. Strayer, F.F., & Trudel, M. (1984). Developmental changes in the nature and function of social dominance among young children. Ethology and Sociobiology, 5, 279±295. Sully, J. (1895). Studies of childhood. London: Longmans, Green. Taylor, J., Iacono, W.G., & McGue, M. (2000). Evidence for a genetic etiology of early-onset delinquency. Journal of Abnormal Psychology, 109, 634±643. Tremblay, R.E. (2000). The development of aggressive behaviour during childhood: What have we learned in the past century? International Journal of Behavioral Development, 24, 129±141. Tremblay, R.E., Boulerice, B., Harden, P.W., McDuff, P., PeÂrusse, D., Pihl, R.O., & Zoccolillo, M. (1996a). Do children in Canada become more aggressive as they approach adolescence? In Human Resources Development Canada & Statistics Canada (Eds.), Growing up in Canada: National Longitudinal Survey of Children and Youth (pp. 127±137). Ottawa: Statistics Canada. Tremblay, R.E., & Craig, W. (1995). Developmental crime prevention. In M. Tonry & D.P. Farrington (Eds.), Building a safer society: Strategic approaches to crime prevention, Vol. 19 (pp. 151±236). Chicago: The University of Chicago Press. Tremblay, R.E., Japel, C., PeÂrusse, D., Boivin, M., Zoccolillo, M., Montplaisir, J.,

13. Development of aggressive behaviour 265 & McDuff, P. (1999a). The search for the age of ``onset'' of physical aggression: Rousseau and Bandura revisited. Criminal Behavior and Mental Health, 9, 8±23. Tremblay, R.E., Kurtz, L., MaÃsse, L.C., Vitaro, F., & Pihl, R.O. (1995). A bimodal preventive intervention for disruptive kindergarten boys: Its impact through midadolescence. Journal of Consulting and Clinical Psychology, 63, 560±568. Tremblay, R.E., LeMarquand, D., & Vitaro, F. (1999b). The prevention of ODD and CD. In H.C. Quay & A.E. Hogan (Eds.), Handbook of disruptive behavior disorders (pp. 525±555). New York: Kluwer Academic/Plenum. Tremblay, R.E., MaÃsse, L.C., Pagani, L., & Vitaro, F. (1996b). From childhood physical aggression to adolescent maladjustment: The MontreÂal Prevention Experiment. In R.D. Peters & R.J. McMahon (Eds.), Preventing childhood disorders, substance abuse and delinquency (pp. 268±298). Thousand Oaks, CA: Sage. Tremblay, R.E., Pihl, R.O., Vitaro, F., & Dobkin, P.L. (1994). Predicting early onset of male antisocial behavior from preschool behavior. Archives of General Psychiatry, 51, 732±738. Vitaro, F., Brendgen, M., & Tremblay, R.E. (1999). Prevention of school dropout through the reduction of disruptive behaviors and school failure in elementary school. Journal of School Psychology, 37, 205±226. Vitaro, F., Brendgen, M., & Tremblay, R.E. (2001). Preventive intervention: Assessing its effects on the trajectories of delinquency and testing for mediational processes. Applied Developmental Science, 5, 201±213. Vitaro, F., Gendreau, P.L., Tremblay, R.E., & Oligny, P. (1998). Reactive and proactive aggression differentially predict later conduct problems. Journal of Child Psychology and Psychiatry, 39, 1±9. Vitaro, F., Tremblay, R.E., Kerr, M., Pagani-Kurtz, L., & Bukowski, W.M. (1997). Disruptiveness, friends' characteristics, and delinquency in early adolescence: A test of two competing models of development. Child Development, 68, 676±689. Wahler, R.G., & Dumas, J.E. (1986). Maintenance factors in coercive mother-child interactions: The compliance and predictability hypothesis. Journal of Applied Behavior Analysis, 19, 13±22. Walters, R.H., & Parke, R.D. (1964). Social motivation, dependency, and susceptibility to social in¯uence. In L. Berkowitz (Ed.), Advances in experimental social psychology, Vol. 1. New York: Academic Press. Wasserman, G.A., & Miller, L.S. (1998). The prevention of serious and violent juvenile offending. In R. Loeber & D.P. Farrington (Eds.), Serious and violent juvenile offenders: Risk factors and successful interventions (pp. 197±247). Thousand Oaks, CA: Sage. Webster-Stratton, C. (1998). Preventing conduct problems in Head Start children. Journal of Consulting and Clinical Psychology, 66, 715±730. Weikart, D.P., & Schweinhart, L.J. (1992). High/Scope preschool program outcomes. In J. McCord & R.E. Tremblay (Eds.), Preventing antisocial behavior: Interventions from birth to adolescence (pp. 67±86). New York: Guilford Press. Zelazo, P.D., Helwig, C.C., & Lau, A. (1996). Intention, act, and outcome in behavioral prediction and moral judgement. Child Development, 67, 2478±2492.

14 Cross-cultural views on human development in the third millennium Pierre R. Dasen

University of Geneva, Switzerland and

Ramesh C. Mishra

Banaras Hindu University, Varanasi, India INTRODUCTION In the 1986 special issue of IJBD (International Journal of Behavioral Development) on cross-cultural human development, the co-editors, Gustav Jahoda and Pierre Dasen, mentioned ``taking culture seriously'' as the main characteristic of a cross-cultural perspective. They complained that ``mainstream'' developmentalists showed little sign of interest in culture. In the meantime, the situation has changed considerably. At ISSBD biennial meetings, a large proportion of presentations are now culturally oriented, in various ways, and the same is true of articles in IJBD. Whether one is satis®ed with the situation or not is of course a matter of personal interests. In the four volumes of the Handbook of child development (Damon, 1998), out of 71 chapters, there are 3 that we would consider cultural, none crosscultural. The other chapters make an occasional reference to culture, but very rarely are cross-cultural studies mentioned. To take but one example, the chapter on aggression and antisocial behaviour (Coie & Dodge, 1998) makes no reference to the important work of Schlegel and Barry (1991), which includes a cross-cultural study of the antecedents of antisocial behaviour in 168 societies world-wide. There is one page on issues of gender, ethnicity, and cultural speci®city, in which the authors state: ``However, the importance of examining developmental models separately in various gender, ethnic, and cultural groups has become recognised and is likely to characterise research over the next decade'' (p. 785). In other words, there has been, from our perspective, considerable progress in the last decade or two, but there is still a lot to be done. Indeed, the study of human development is still heavily dominated by a Western, or eurocentric (and hence ethnocentric) perspective. Ethnocentrism is all the more dif®cult to track in that it is mainly involuntary and unconscious (Segall, Dasen, Berry, & Poortinga, 1999). For example, we are writing here about prospects in the ``third millennium''. Which third millennium? Only the Western, Christian oneÐwhich, although or because it is now practised in most of the world (in addition to local calendars that are still in use), in fact represents Western hegemony. In the year AD 2000,

14. Cross-cultural views 267 Hindus had already been in the third millennium for 57 years. But there are different calendars within the Indian subcontinent alone: it was the year 1921 in the Shak year, 1405 in the Bengali calendar, 1120 for the Newars of Nepal, and 2542 for some Buddhists. At the same time, Muslims started the year 1420, while the Hebraic calendar indicated the year 5761, the Alexandrian one (still in use for monks in Ethiopia), 1992, and 1716 for Copts in Egypt. Everyone knows that the calendar is a social convention, yet somehow it is taken as an objective reality. While the belief that we have entered a third millennium is thus rather culture-speci®c, it is irrelevant, of course, to the project of taking stock and re¯ecting on future developments of our discipline. We will do so with a speci®cally cross-cultural point of view. The other side of the coin, beyond the scope of this chapter, would be to show that there is also a need for taking the developmental perspective seriously in cross-cultural psychology. A CHANGE IN PARADIGMS In all of the social sciences, the last decades have brought about a shift in paradigms, from the positivism inspired by nineteenth-century natural sciences to the constructivism of post-modernist, post-industrial societies (Guba, 1990). For the moment, this shift is still seen as con¯ictual, or at least problematic. In the domain of (cross-)cultural studies, the debate has taken the form of an opposition between quantitative, antecedent±consequenceoriented comparative studies, close to the traditional mainstream of psychology, and the qualitative, hermeneutic, ``thick descriptions'' of so-called ``cultural psychology'', inspired to a large extent by the relativism of anthropology. The debate is very well illustrated by the various chapters in Volume 1 of the second edition of the Handbook of cross-cultural psychology (Berry, Poortinga, & Pandey, 1997). There is no doubt that the debate has substance, if only in the shift of the de®nitions of the key concept of ``culture''. Rather than being a relatively static and stable property of societies, seen as different entities sharing a set of norms and symbols, the more recent de®nitions of culture insist on dynamic processes. Culture is no longer shared by all, but is being coconstructed by each individual (Boesch, 1991; Bruner, 1996; Cole, 1996; Shweder, 1990). Intentionality has replaced causality as an explanatory principle. Similarly, socialisation is no longer conceived of as an external in¯uence of socialisation agents who control the behaviour of the upcoming generation, but as a process of negotiation between the developing individual and other actors on the scene (Camilleri & Malewska-Peyre, 1997). Although we welcome such a substantive change towards more dynamic process-oriented theories, there are a few teething problems that need to be highlighted. The proponents of the incoming paradigms, no doubt because they still identify as a minority threatened by the powerful establishment of

268

Dasen and Mishra

positivistic science, tend to be defensive, if not aggressive. They try to establish their credo not so much by demonstrating how useful it is, but by bashing the retarded oldies who still believe in sampling, inferential statistics, the assessment of reliability, and other such outdated symptoms of a model based on physics. The problem with this type of language is that it is immediately rejected by the other group as unacceptable; instead of a scienti®c debate, one gets the tightening of opposed social groups that need to reinforce the positive identity of their in-group by putting down the outgroup. (Cross-)cultural psychologists should be able to do better in terms of con¯ict avoidance and mediation! To our minds, there does not have to be a con¯ict, but two interesting and complementary points of view. A problem occurs, however, when the paradigmatic shift is too extreme: The strong emphasis on individual intentionality may in fact be an ethnocentric view based on Western individualism (KagitcËibasi, 1997). And we do not believe that all societies will evolve towards a world-wide, post-modern individualism on the model of Western capitalistic hegemony; at least, we hope that is not the fate of the so-called ``global village''. Linked to an extreme theoretical shift are also problems in methodology. Considering that there is no reality that is to be described, that there are no facts to be collected, no data to be analysed, that social scientists can do no more than writing literary texts, leads to a sort of suicide of the social sciences. Because of the dif®culties of going about an empirical study, the tendency is to stay at home and deconstruct concepts; armchair social science was not only characteristic of the past. We think that this is extremely harmful and discouraging, particularly for young scientists, who ®nd it dif®cult to design a feasible study. The bias against ``facts'' sometimes even leads to inventing instead of collecting data. For example, in a paper on parental reasoning, Valsiner and Litvinovic (1996) establish a typology of parents' discourse with children, and give ®ctional examples of each. Other researchers who still go around and do interviews, in the belief that this is the way to re¯ect what people have to say, are likely to be taxed as positivists. The concluding chapter in Volume 1 of the Handbook of cross-cultural psychology (Poortinga, 1997) is entitled ``Towards convergence?''. Indeed, we believe that this seemingly unreconcilable opposition of paradigms is a swing of the pendulum, a necessarily extreme revolution, which will subside into a more moderate, median position. Note that we do not predict a return to positivism, but to what Guba (1990) has termed post-positivism. It is also the approach advocated by Berry, Poortinga, Segall, and Dasen (1992) and Segall et al. (1999) for the broad ®eld of cross-cultural psychology. We see two reasons for such a development: First, quantitative and qualitative methodologies, or reliability and validity, are not necessarily opposed but complementary. For each research project, depending on its goal, one or the other may be more adequate, if not a combination of

14. Cross-cultural views 269 both. The second reason is that, as Segall (1993) has pointed out, the socalled new paradigms are not necessarily completely recent inventions. Cultural psychology has been advocated for a long time within anthropology and cross-cultural psychology as the ``emic'' approach (Jahoda, 1992; Jahoda & Krewer, 1997). As far back as 1974, Berry and Dasen de®ned three goals of comparative cognitive psychology (Berry & Dasen, 1974, p. 14): 1

To transport our present hypotheses and laws to other cultural settings [in order] to test their applicability or generalisability; 2 To explore new cultural systems to discover cognitive variations and differences we have not experienced within our own cultural context; 3 To compare our prior understanding with our newer knowledge within diverse cultures to generate more universal descriptions, hypotheses and laws of human cognitive functioning.

The second goal was even then recognised to be more important than the ®rst (Berry & Dasen, 1974, p. 14): The basic aims of the comparative method cannot be limited to a mere testing of extant hypotheses; we must be concerned with expanding and integrating our knowledge of human cognitive variation. Thus one of the more recent attempts to de®ne the method for psychologists (Whiting, 1968), which limits the method to hypothesis testing, is too narrow for the needs of cognitive research. Indeed, as Strodbeck (1964) has argued, the discovery aspect (aim no. 2) of the method, has more to contribute to our overall aims than the initial testing aspect (aim no. 1). Hence, the goals of cross-cultural psychology are both understanding local phenomena and, at the same time, attempting to develop panhuman generalisations. The development of ``cultural psychology'' with its emphasis on culture as a phenomenon ``co-constructed'' by individuals through interactions rather than as something given from outside has added new dimensions to the ongoing debate between cultural universalists and cultural relativists. Although the search of culture-speci®c variations in psychological processes is a common goal of both cultural and cross-cultural psychologists, and although most cultural psychologists do believe in the notion of psychic unity (Green®eld, 1997, p. 309; Shweder & Sullivan, 1993, p. 517), the former question the theoretical relevance of this universal. They believe that panhuman characteristics (even if identi®ed) are likely to be so abstract that they would have very limited explanatory power to account for speci®c psychological phenomena. In spite of this stand, evidence for universal explanations in cultural psychology is not altogether absent. Constructs like ``scaffolding'' (Green®eld, 1984) or ``guided participation'' (Rogoff, 1990) are good examples of processes useful for a general explanation.

270

Dasen and Mishra

Increasingly, attempts at explanations in the cross-cultural approach are formulated in ways such that the perspective of different cultural groups is not distorted, whereas in the cultural approach, explanations are formulated in ways that characterise psychological functioning of individuals within a particular cultural group at a particular point of time in a particular social-historical context. A rapprochement between these positions is necessary for a proper understanding of the cultural aspects of human development. METHODOLOGY As mentioned, the shift in paradigms also has important methodological implications. For a long time, psychological research, including child development research, has been characterised by laboratory studies, use of sophisticated experimental designs, statistical analyses, and an epistemology rooted in the natural sciences. The emphasis was on the study of the effect of immediate context on behaviour, with a general disregard for variables that were distal in nature. Methodological rigour was considered as the most important criterion to make the study of behaviour part of a truly scienti®c enterprise. In spite of some departures from mainstream psychology in the last two decades, a central feature of studies of human development has remained generally unchanged: The quantitative approach still dictates decisions regarding the forms of data collection and their analyses. Consequently, data obtained by any method regarding any psychological quality of individuals do involve categorisation of responses in ways that allow some sort of statistical comparison between groups. The reductionism implicit in the manipulation of variables and the dominance of quanti®cation have been a major aspect of the critique of psychological studies for a long time (Harre & Secord, 1972). It is argued that the methods and the measurement techniques used by academic psychology re¯ect the dominance of a mechanistic model in the analysis and understanding of human behaviour (Eckensberger, 1979; Eckensberger & Burgard, 1983). The paradigm shift mentioned earlier has already led to a considerable change during the last decades. A number of psychologists have expressed serious doubts on the applicability of laboratory research to real-life settings. As a result, a number of post-positivistic nonexperimental methodologies have surfaced, and these are being used in an increasingly wider range of empirical studies of human development. A number of questions are being asked in new terms and answered in new ways. The development of alternative strategies in order to ensure greater representation of qualitative techniques of analysis in cross-cultural studies of human development is an important agenda for psychologists to address in the forthcoming years.

14. Cross-cultural views 271 There are efforts to develop an ecologically valid psychology based on studies grounded in the ``real world'' of people. This move towards naturalness has made psychology of human development more open to research on a range of topics that were previously neglected (e.g., studies on the development of self, and on ``everyday cognition''). With the inclusion of these new topics in the research agenda, the importance of different types of data collection strategies (e.g., diary studies, self-reports, conversation and discourse analyses) is recognised. Although these techniques are considered to be full of problems of objectivity, the pursuit of objectivity itself has been replaced by taking the insider's perspective, and much more attention has been paid to interpretive, ecological, and theoretical validity (Green®eld, 1997, pp. 316±317). Qualitative procedures to ensure reliability, replicability, and theoretical generalisability are being devised (Miles & Huberman, 1994; Pourtois & Desmet, 1988), and much more attention is needed to these issues. To search for convergence is also important in the methodological domain. The introduction of more qualitative methods does not preclude crosscultural comparisons. Indeed, we may hope that in the forthcoming years, researchers will include more diverse groups and draw more appropriate samples in their studies. Sampling should occur on the basis of serious theoretical considerations, not in the casual ways of so-called ``safari-style'' cross-cultural research (Segall et al., 1999). A caution is to be observed at this point. As Miller (1997) argues, employing interpretive techniques that make use of indigenous language is as biased a practice as complete reliance on experimental or comparative approaches. All methodological strategies have their own strengths and weaknesses; the decision about their ef®cacy can be made only in the context of the questions that one is trying to answer. NEEDED FURTHER DEVELOPMENTS 1: TAKING NEW THEORIES TO THE CROSS-CULTURAL TEST MORE QUICKLY Psychology, be it developmental or not, still produces new theories on the basis of research with very limited samples, often of a single sociocultural background (such as middle-class Anglo Psych-1 students) in the setting of a compulsory and largely arti®cial laboratory experiment. Yet, the intention is to develop theories that say something about human behaviour more generally. Even in the newer constructivist paradigms, which take context into account, the purpose is not to make a statement only about Psych-1 students. Yet, there is very little cross-cultural testing of theories going on, in fact less than there was in the 1970s and 1980s. Cognitive science, for example, has become an important domain of research, yet culture is almost absent

272

Dasen and Mishra

either in the machine and arti®cial intelligence, or in the human counterpart. Internal processes are treated as if they were universal by de®nition, or rather, the question of universality is not even asked (Wassmann, 1993). However, one paradox is that it is not just the hard-nosed, experimental theories that are deemed universal without an empirical test: The theories that potentially take cultural variables into account do not fare much better. A few examples will illustrate our point. Another paradox is that theory developers should, in fact, not have to fear the outcome of the cross-cultural test. Theories are hardly ever completely rejected; the most common pattern is to ®nd that some parts hold up whereas others have to be modi®ed. The end result is a more complex theory able to take more situations into account. Instead of being afraid that their theories will not hold up, the senior scientists that devise them should systematically encourage their students to take them to the ®eld. There is scope for many PhDs.

Neo-Piagetian theories Piaget's theory was perhaps particularly ill-suited to taking culture seriously, yet it has led, over a couple of decades, to an enormous amount of cross-cultural research (Dasen & Heron, 1981; Mishra, 1997). Neo-Piagetian theories (see Dasen & de Ribaupierre, 1987, for an overview) are much more complex in terms of taking social and cultural variables into account. Yet, except for a very small number of studies limited to one or the other detail of such theories, none of them has led to a large-scale, complete replication in at least two cultural contexts other than that in which they were designed. Theories of mind Another outgrowth of genetic epistemology is the study of ``theories of mind'', i.e., the tendency to impute mental states to oneself and to others. One draws on theories of mind to understand other peoples' behaviours and psychological states, and by projecting oneself onto others. There are indications that, whereas chimpanzees have little mentalistic insight (Tomasello, 1999), the basic processes of theories of mind are universal in human infants (Lillard, 1998). Studies with young children give mixed results. Initial cross-cultural results seemed to point towards universality (Flavell, Zhang, Zou, Dong, & Qi, 1983, in China; Avis & Harris, 1991, with Baka pygmy children), whereas more recent research has pointed to cultural differences in the rate of development of theories of mind (Wahi & Johri, 1994, in India), or even structural differences linked to language. For example, Vinden (1996) used the standard ``surprising objects'' (e.g., a sponge that looks like a rock) and ``deceptive container'' (e.g., a matchbox full of pebbles) tasks with Junmn Quechua children in Peru. These children

14. Cross-cultural views 273 demonstrated an understanding of the appearance±reality distinction, but performed poorly on questions regarding representational change (``Before you touched the object, what did you think it was?) and false belief (``If another child just looks at the object, what would he think it is?''). The author explains these differences by the absence of explicit mentalistic vocabulary in the Quechua language. In another study involving Mofu children in Cameroon, Tolai and Tainae children of Papua New Guinea, and Western children attending a school in that country, Vinden (1999) found that almost all children from non-Western cultures had dif®culty predicting an emotion based on a false belief.

Folk psychologies Whether the basic processes of theories of mind turn out to be universal or not in children, there is the paradox that, according to ethnographic records, adults' theories of mind, or what Lillard (1998) calls ``ethnopsychologies'', vary a great deal from culture to culture. The author takes as typical examples the differential attraction to magic, different conceptual distinctions regarding thoughts and feelings, sensory inputs, and the links between body and mind, denying negative emotions, and different values regarding rational thought versus feelings, and science versus spirituality. As children become adults, they come to accept these different ethnopsychologies, yet little is known about the cognitive processes that this entails. In other words, there is a big gap between the psychological studies that attempt to assess speci®c thinking processes in children, and the ethnographic descriptions of ethnopsychologies such as reviewed by Lillard, reminding us of the important distinction between individual processes and collective representations pointed out by Harris and Heelas (1979). Here again, we seem to come up against a struggle between paradigms. Cultural psychologists often advance explanations of behaviour in terms that are often used in folk psychology. Cognitive science, which may be regarded as a rival of folk psychology (Haselager, 1997), uses models as a primary means of understanding and explaining behaviour. The ease and speed with which common-sense reasoning and understanding takes place is unexpectedly dif®cult to capture in these models. This source of uneasiness in cognitive science has become known as ``the frame problem''. Folk psychology has a fair chance of being vindicated if cognitive science is not capable of answering the frame problem in a way that is compatible with at least the most central characteristics of folk psychology. The question is whether cognitive science models or the folk psychological approach will dominate the study of cross-cultural human development. It is true that the status of folk psychology has to be decided on empirical grounds. Hence, ®nding an area in which folk psychological assumptions can be tested is of considerable importance, especially from the point of view of the analysis of human development. Parental belief systems

274

Dasen and Mishra

and their effect on the development of cognitive or social competence among children (Harkness & Super, 1996) represent one area in which some work has already been accomplished, and support for the role of parental ethno-theories in human development has been obtained (Bril, Dasen, Sabatier, & Krewer, 1999). We expect that many more new areas will emerge in the forthcoming decades to look for the application of folk psychological principles to issues of human development. At a later stage, however, an overarching theoretical framework may be required to put various pieces of folk psychological information together to develop a universal psychology of human development (Dasen, 1993).

Metacognition Another area of research on cognitive development has become important in recent years, and is starting to be studied from a cross-cultural perspective. This is metacognition, or ``knowledge about, and awareness of, one's own capabilities and cognitive plans vis-aÁ-vis the task'' (Davidson & Freebody, 1988). Metacognition includes knowledge about cognition, about oneself as a learner and the factors that in¯uence one's performance (declarative knowledge), about the execution of skills (procedural knowledge) and about when and why to apply cognitive actions (conditional knowledge) (Schraw, 1989; Schraw & Moshmann, 1995), as well as control processes such as planning, monitoring, and evaluation. These authors distinguish three levels of representation: 1 2 3

Cognitive level: tacit (without any explicit awareness), domain-speci®c, with limited transfer. Metacognitive level: informal, fragmentary, some across-domain transfer. Conceptual level: formal (explicit and integrated) mental models and theories, broad transfer.

According to Schraw (1989), metacognition is acquired through autonomous learning, peer-regulated learning, and direct learning, the latter two mainly in schools. Hence, Western-type schooling at secondary level may be necessary but not suf®cient for the attainment of at least the conceptual if not the metacognitive level, mirroring the conclusion Dasen and Heron (1981) have put forward for Piaget's formal operations. Whether the metacognitive level is also dependent on schooling is an interesting research question, open to cross-cultural testing. There certainly are metacognitive skills developed outside of school. For example, Australian Aboriginal teacher trainees were found to possess metacognitive knowledge about indigenous story telling that seemed to be situationspeci®c, however, and dif®cult to transfer to school tasks (Davidson & Freebody, 1988). The same authors found socioeconomic and ethnic

14. Cross-cultural views 275 differences in the metacognitive knowledge children bring to school (linked, for example, to the degree of reading in the home), and this was linked to success in school. Cultural differences in metacognition have been reviewed by Davidson (1994). Research on metacognition is linked to the topic of everyday cognition (Segall et al., 1999; Schliemann, Carraher, & Ceci, 1997), in which there is some cross-cultural information, for example on planning skills (Gauvin, 2001; Strohschneider & GuÈss, 1998; Tanon, 1994). We may hopefully expect much more research on the topic of metacognition, both in schools and in everyday contexts, in the next decade.

French social psychology and migrant studies Over the past two decades, there has accumulated a considerable amount of research in French-speaking countries and regions of Europe and North America on the processes linked to the contact between people of diverse cultural origins, what the French call ``interculturation'' (a term they prefer to the more common ``acculturation'' in English). This research has been spurred by the increasing migrations from further a®eld than was the case with the traditional ``guest workers'' in Europe, and the belated understanding that these people (not just working hands but also their families) had come to stay. We cannot summarise here even a small part of this research1, but will take only one example: Camilleri (Camilleri & Malewska-Peyre, 1997) has established a most interesting typology of what he called ``identity strategies'', in other words a rich set of acculturation processes. UnfortunatelyÐfrom our point of view (see also Dasen & Ogay, 2000)Ðthe research is based almost exclusively on migrants from North Africa in France. Several aspects of the theory seem to be linked directly to this particular social and political situation, for example, fundamental differences in values linked to religion (Islam vs. Christianity), and the fact that the migrants are being despised and discriminated against by the host society. How would the theory stand up to a test with a different group of migrants in a different political situation? Sapru (1999), for example, has studied identity processes of Indian adolescents in Geneva, a group that suffers no racism or discrimination. Yet the parents use one of the strategies described by Camilleri, that of exaggerating some of the traits of the culture of origin (such as religious practices and the arts), mainly in order to keep their children busy and away from what they see as nefarious effects of the Western surroundings. In this example, we ®nd the same strategy, but for different reasons.

1 A French-speaking international organisation, the Association pour la Recherche Interculturelle (ARIC), has provided a network for researchers in this ®eld, and has contributed to publish much of the work, for example as conference proceedings.

276

Dasen and Mishra

Lifespan developmental psychology Lifespan developmental psychology is another example, close to the interests of IJBD readers. Major developments of great importance have taken place in this ®eld, yet little cross-cultural testing has taken place. When Valsiner and Lawrence (1997) were asked to review the literature for Volume 2 of the Handbook of cross-cultural psychology, they could come up with only a handful of serious studies, and were therefore led to provide their own theoretical models. Of course, many studies exist that document various periods of the lifespan, such as infancy, adolescence, or old age, but these are always limited to that particular age range.

Neuropsychology The tremendous advancement in neuropsychological studies in some speci®c areas has posed new challenges to developmental psychologists, including those who study patterns of development in a variety of cultural contexts. The neuropsychological approach attempts to discover the functional organisation of the brain as it relates to a variety of cognitive skills and abilities. Researchers in this ®eld generally believe that cognitive functions associated with brain pathology have similar manifestations across members of the human species. However, neuropsychological tests tend to measure abilities that seem to be culturally learned2. The crosscultural evidence suggests that the performance on psychological tests is moderated by factors such as culture, ecological demands, language, and educational level of individuals or groups (Irvine & Berry, 1988). Spatial ability may be taken as an example. Studies indicate that in comparison to agricultural populations, the hunting-gathering populations demonstrate highly developed spatial ability (Berry, 1976; K. Mishra, 1998; Mishra, Sinha, & Berry, 1996). This is evidenced by ®eld-independent (highly differentiated) functioning of individuals on several perceptual, cognitive, and motor tasks. Silverman (1981) suggests that ®eld dependence (low differentiation) on cognitive tasks is related to relative left-hemispheric dominance. Witelson and Swallow (1988) indicate that right-hemisphere lateralisation is responsible for several aspects of spatial perception among children and adults. Whether ®eld-independence among hunter-gatherers is associated with right-hemispheric lateralisation or with long-term ecological adaptations (which in turn result in hemispheric lateralisation) is not known. Our understanding of cultural differences in neuropsychological test performance is extremely limited so far. However, neuropsychological testing in different parts of the world has been undertaken with a universalistic

2 For example, using a spoon or scissors, keying a lock, tooth-brushing, screwing a nut, or complex activities such as writing alphabets or reproducing words.

14. Cross-cultural views 277 assumption of human brain functioning. The validity of this assumption has been questioned in recent years. Matthews (1992, p. 421) has observed that ``a very limited kind of neuropsychology, appropriate to only a fraction of world's population, is presented to the rest of the world as if there could be no other kind of neuropsychology, and as if the educational and cultural assumptions [. . .] on which neuropsychology is based were obviously universals that applied everywhere in the world''. One may perceive it as a ``cultural construction'' of neuropsychology. The possibility that different patterns of brain organisation may appear in the course of time due to different ecological demands cannot completely be ruled out. In some research, an attempt has been made towards direct or indirect analyses of the impact of eco-cultural variables in neuropsychological test performance. Ardila (1995) has reviewed neuropsychological studies of perceptual and cognitive functions, bilingualism, illiteracy, and handedness. There is considerable evidence to suggest that the outcomes of neuropsychological assessments are in¯uenced signi®cantly by ecological and cultural factors. On the one hand, these ®ndings point to a close linkage between biology and culture. On the other hand, they suggest that taking culture seriously does not exclude the possibility of taking biology into account (Keller, Poortinga & Schoelmerich, 2002). It seems that ecological conditions with which people negotiate their day-to-day life generate processes of adaptation at the biological and cultural levels simultaneously. The eco-cultural framework proposed by Berry et al. (1992) provides room for biological level adaptations. We may hope that researchers of human development would study seriously the processes of biological adaptation generated by long-lasting eco-cultural features of the populations. APPROPRIATE THEORETICAL FRAMEWORKS The concern for the search of appropriate methodology in the study of human developmental processes and their corresponding representations in the human brain raises the question of an appropriate theoretical framework that may be used to guide research in this ®eld, and that allows for a meaningful integration of many diverse sets of data. Dasen (1998, in press) is providing a description and an attempt at integration of these frameworks. Research in developmental psychology has widely used the ecological model of Bronfenbrenner (1979, 1993). This model suggests a strategy through which one can look for variables in the proximal or distal environment of children in order to analyse their in¯uences on development. Although within-culture variations in the development of children have been analysed often in terms of the variable in¯uences involved at different levels, the model has not been much used in cross-cultural research. However, two introductory textbooks of cross-cultural psychol-

278

Dasen and Mishra

ogy, those of Gardiner, Mutter, and Kosmitzki (1997) and Segall et al. (1999), give it some importance. Harkness and Super (1983, 1992; Super & Harkness, 1986, 1997) have proposed another framework, called ``developmental niche'', in which the developing child is situated in a system formed by the physical and social contexts, the caregivers' educational practices, and their ``ethnotheories'' or social representations concerning child development. This framework makes it clear that the unit of analysis should be neither the child alone nor the contexts alone, but the child in context, thus combining the perspectives of psychology and anthropology. The niche is an open system, itself set in the wider macro-social components that are taken care of by another useful theoretical framework, the ``eco-cultural framework'' developed over the years by Berry (1976, 1980; Berry et al., 1992; Segall et al., 1999). This itself has historical antecedents (cf. Berry, 1995; Jahoda, 1995), particularly Whiting's (1977) psycho-cultural model, in considering that culture is basically an adaptation to ecological and socio-historical contexts. The accommodation of cultural change as an important variable capable of changing cultural features and behavioural characteristics of individuals or groups extends the scope of this framework in comparison to other models (Mishra et al., 1996). NEEDED FURTHER DEVELOPMENTS 2: APPLICATIONS OF CROSS-CULTURAL PSYCHOLOGY TO HUMAN AND SOCIAL DEVELOPMENT Cross-cultural psychologists generally conduct research in settings where people experience a variety of problems. These range from developmental issues related to basic human needs (e.g., health, education) to social issues that are of vital concern in contemporary societies (e.g., environmental and cultural crises). Given this state of affairs, cross-cultural research on human development cannot be pursued only as a scienti®c enterprise; instead, such research always involves some inherent social implications. The growth of applied cross-cultural psychology (Brislin, 1990; Dasen, Berry, & Sartorius, 1988) can be linked to some of these concerns. Issues related to poverty, schooling, physical and mental health, environment, acculturation, and group relations have drawn considerable attention. Over the years, other problems (e.g., community intervention) have been added to the list as research topics. Berry, Mishra, and Tripathi (2002) indicate that issues of human and social development cannot be grasped and solved without a thorough understanding of the cultural context in which they are rooted. The study of human development processes in different cultural settings can be mutually bene®ting in the sense of adopting a comprehensive approach to their solution. The goals of ``applied developmental psychology'' (Serpell, 1992) can be met only by analysing human development in the context of

14. Cross-cultural views 279 child development programmes and policies pursued in different parts of the world. The third millennium looks forward to vigorous efforts in this direction.

WORKING CONDITIONS IN DIFFERENT CONTEXTS Overcoming eurocentrism is one of the goals of a cross-cultural approach. However, this will only be possible by also reducing the Western hegemony in the means of producing research. It is absolutely normal that theories should be grounded in the socio-historical context in which they are devised, and it is this embeddedness that increases the likelihood of their being useful. However, this also means that, as long as more than 90% of the researchers reside in the North America and Europe, the majority of the research will show Western bias. PeÂrez and Dasen (1999) have edited a special issue of the UNESCO journal, Prospects, to review the situation of educational research in various parts of the world (see also Akkari, Sultana, & Gurtner, 2001). There are enormous regional disparities in the number of researchers, and their means (such as funding and access to publications). In the past few decades, the situation has improved in many parts of Asia (Mishra, 1999), while it has deteriorated in Africa. A common problem is to bring interesting research that is being produced despite limited means to the attention of an international audience. Most of the research is being done for university degrees, and remains in the form of unpublished MA and PhD theses. This is an enormous waste of resources, and one may hope that the increasing availability of the Internet will provide some solution to this problem.

INDIGENISATION AND FOLK PSYCHOLOGY One reaction to the Western bias in research and in textbooks has been the development of so-called ``indigenous psychologies'', i.e., the production by local psychologists of locally relevant theories based on local culture. Sinha (1997) has reviewed this trend; most of this type of research is done in Asia, although it is only weak in Latin America, and almost absent in Africa (but see Nsamenang, 1992). Here again, our own feeling is that we are witnessing a necessary, but temporary, swing of the pendulum. It is absolutely essential to acknowledge and give status to this local knowledge. In the long term, however, it would be a great disadvantage to have only an accumulation of idiosyncratic and incomparable pieces (Dasen, 1993); at some stage, a return to overarching theories will have to occur, ®tting the pieces of the puzzle so as to create a truly ``global'' picture.

280

Dasen and Mishra

CONCLUSION The relationship between culture and behaviour has been of an ``on-again'' and ``off-again'' kind. In the past several decades, culture was believed to suggest an inordinate degree of boundedness, homogeneity, coherence, and stability (Brumann, 1999). This conceptualisation of culture is not presently in wide acceptance. On the other hand, there has developed what Sahlins (1994) calls the ``. . . fashionable idea that there is nothing usefully called a culture'' (p. 386) in the past decade or so. There has also been a movement towards ``writing against culture'' (Abu-Lughod, 1991; Fernandez, 1994). The argument is that social reality is characterised by variability, inconsistencies, con¯ict, change, and individual agency, which are not subject to cultural explanations. In the midst of these controversies, however, culture has exerted an in¯uence beyond the borders of the discipline of anthropology. For example, Hannerz (1996, p. 30) observes that people now seem to agree that: ``. . . culture is everywhere. Immigrants have it, business corporations have it, young people have it, women have it, even ordinary middle-aged men have it, all in their own versions . . . We see advertising where products are extolled for ``bed culture'' and ``ice cream culture'', and something called ``the cultural defence plea'' is under debate in jurisprudence.'' If such a state of affairs continues, we may hope that the analysis of cultural features will certainly ®nd an inevitable place in any description and understanding of human behaviour. At the same time, a debate about the nature of behavioural explanations is also likely to continue. The main issue is whether explanations should be modelled on scienti®c theory-building, or on humanistic interpretation. This debate is very much alive today in the competing ®elds of cognitive science and folk psychology (Haselager, 1997), where the main question is about the right frame of mind. Can computational or neurocomputational models provide the right frame of mind, or can the folk systems of cognition (e.g., beliefs, hopes) do it? A similar debate persists in anthropology as well (Moore, 1997). Fischer (1996, p. 11) argues for ``the indissolvibility of scienti®c and humanistic approaches to the study of human societies''. D'Andrade (1996, p. 10) suggests that ``the con¯ict is not between humanists and scientists, but between those who think that anthropology should be primarily a kind of political action group and those who think it should be primarily a kind of scholarly study''. It is dif®cult to predict the future of these debates. At the same time, the knowledge we have gained about the role of culture in human behaviour is primarily due to such controversies. In the ®eld of cognitive development, such theoretical controversies have generated a series of insightful studies, and there have been serious efforts for rapprochement among them (Berry et al., 1992; Mishra, 1997). It may be noted that human sciences are prone to generate controversies, because they use human beings as subjects. Sciences where there is clear division between the

14. Cross-cultural views 281 observer and the observed are less likely to provoke such controversies. As Moore (1997, p. 275) observes, ``At a more fundamental level, we humans are simultaneously phenomena governed by laws of nature and chaotic selfcreations with endlessly complex ideas about the meanings of existence. Any ®eld that studies an organic actor like a human is bound to experience con¯icts.'' It is expected that a multifaceted, multidimensional, and ¯exible vision will emerge towards an adequate understanding of culture and human development in the new millennium. Cross-cultural (and cultural) psychology has long been a somewhat marginal, sometimes even exotic, enterprise, criticised for the absence of strong theory and for its weak methodology. But the situation has improved considerably in the last two decades, and we are now seeing many signs that ``culture is taken seriously''. The cross-cultural approach in psychology, just like the developmental approach, will have become completely successful when they both have disappeared as separate branches, i.e., when mainstream psychology considers that both a developmental and a cross-cultural perspective are absolutely essential.

REFERENCES Abu-Lughod, L. (1991). Writing against culture. In R.G. Fox (Ed.), Capturing anthropology: Working in the present (pp. 137±162). Santa Fe, NM: School of American Research Press. Akkari, A., Sultana, R., & Gurtner, J.-L. (Eds.). (2001). Politiques et strateÂgies eÂducatives. Termes de l 'eÂchange et nouveaux enjeux Nord-Sud. Bern: Peter Lang. Ardila, A. (1995). Directions of research in cross-cultural neuropsychology. Journal of Clinical and Experimental Neuropsychology, 17, 143±150. Avis, J., & Harris, P.L. (1991). Belief-desire reasoning among Baka children: Evidence for a universal conception of mind. Child Development, 62, 460±467. Berry, J.W. (1976). Human ecology and cognitive style: Comparative studies in cultural and psychological adaptation. New York: Sage/Halsted. Berry, J.W. (1980). Ecological analyses for cross-cultural psychology. In N. Warren (Ed.), Studies in cross-cultural psychology, Vol. 2 (pp. 157±189). London: Academic Press. Berry, J.W. (1995). The descendants of a model. Culture and Psychology, 1, 373± 380. Berry, J.W., & Dasen, P.R. (Eds.). (1974). Culture and cognition: Readings in crosscultural psychology. London: Methuen. Berry, J.W., Mishra, R.C., & Tripathi, R.C. (Eds.). (2002). Psychology in human and social development: Lessons from diverse cultures. New Delhi, India: Sage. Berry, J.W., Poortinga, Y.H., & Pandey, J. (Eds.). (1997). Handbook of crosscultural psychology: Vol. 1. Theory and method (2nd ed.). Boston: Allyn & Bacon. Berry, J.W., Poortinga, Y.H., Segall, M.H., & Dasen, P.R. (1992). Cross-cultural psychology: Research and applications. Cambridge: Cambridge University Press. Boesch, E.E. (1991). Symbolic action theory for cultural psychology. Berlin: Springer.

282

Dasen and Mishra

Bril, B., Dasen, P.R., Sabatier, C., & Krewer, B. (Eds.). (1999). Propos sur l 'enfant et l 'adolescent: quels enfants pour quelles cultures? Paris: L'Harmattan. Brislin, R.W. (Ed.) (1990). Applied cross-cultural psychology. Newbury Park, CA: Sage. Bronfenbrenner, U. (1979). The ecology of human development. Cambridge, MA: Harvard University Press. Bronfenbrenner, U. (1993).The ecology of cognitive development: Research models and fugitive ®ndings. In R. Wozniak & K.W. Fischer (Eds.), Development in context (pp. 3±44). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Brumann, C. (1999). Writing for culture: Why a successful concept should not be discarded. Current Anthropology, 40, 1±27. Bruner, J. (1996). The culture of education. Cambridge, MA: Harvard University Press. Camilleri, C., & Malewska-Peyre, H. (1997). Socialization and identity strategies. In J.W. Berry, P.R. Dasen, & T.S. Saraswathi (Eds.), Handbook of cross-cultural psychology: Vol. 2. Basic processes and human development (2nd ed., pp. 41±68). Boston: Allyn & Bacon. Carr, M., Kurtz, B.E., Schneider, W., Turner, L., & Borkowski, J.G. (1989). Strategy acquisition and transfer among American and German children: Environmental in¯uences on metacognitive development. Developmental Psychology, 25, 765±771. Chen, M.J., & Lin, Z.X. (1994). Chinese preschoolers' dif®culty with theory-of-mind tests. Bulletin of the Hong Kong Psychological Society, (32/33), 34±46. Coie, J.D., & Dodge, K.A. (1998). Aggression and antisocial behavior. In W. Damon & N. Eisenberg (Eds.), Handbook of child psychology: Vol. 3. Social, emotional and personality development (5th ed., pp. 779±862). New York: John Wiley. Cole, M. (1996). Cultural psychology: A once and future discipline. Cambridge, MA: Harvard University Press. Damon, W. (Ed.). (1998). Handbook of child psychology (5th ed.). (4 vols.). New York: John Wiley. D'Andrade, G. (1996). Science in anthropology. Anthropology Newsletter, 36, 10. Dasen, P.R. (1993). Theoretical/conceptual issues in developmental research in Africa. Journal of Psychology in Africa, 1, 151±158. Dasen, P.R. (1998). Cadres theÂoriques en psychologie interculturelle. In J.G. Adair, D. BeÂlanger, & K.L. Dion (Eds.), Advances in psychological science/ReÂcents deÂveloppements en psychologie scienti®que: Vol. 1. Social, personal, and cultural aspects/Aspects sociaux, personnels et culturels (pp. 205±227). Hove, UK: Psychology Press. Dasen, P.R. (in press). Theoretical frameworks in cross-cultural development psychology: An attempt at integration. In T.S. Saraswathi (Ed.), Cross-cultural perspectives in human development: Theory, research, and applications. New Delhi: Sage India. Dasen, P.R., Berry, J.W., & Sartorius, N. (Eds.). (1988). Health and cross-cultural psychology: Towards applications. Beverly Hills, CA: Sage. Dasen, P.R., & de Ribaupierre, A. (1987). Neo-Piagetian theories: Cross-cultural and differential perspectives. International Journal of Psychology, 22, 793±832. Dasen, P.R., & Heron, A. (1981). Cross-cultural tests of Piaget's theory. In H.C.

14. Cross-cultural views 283 Triandis & A. Heron (Eds.), Handbook of cross-cultural psychology: Vol. 4. Developmental psychology (pp. 295±342). Boston: Allyn & Bacon. Dasen, P.R., & Ogay, T. (2000). Pertinence d'une approche comparative pour la theÂorie des strateÂgies identitaires. In J. Costa-Lascoux, M.-A. Hily & G. VermeÁs (Eds.), Pluralite des cultures et dynamiques identitaires. Hommage aÁ Carmel Camilleri (pp. 55±80). Paris: L'Harmattan. Davidson, G.R. (1994). Metacognition, cognition and learning: Old dubitations and new directions. South Paci®c Journal of Psychology, 7, 81±31. Davidson, G.R., & Freebody, P. (1988). Cross-cultural perspectives on the development of metacognitive thinking. Hiroshima Forum for Psychology, 13, 21±31. Eckensberger, L.H. (1979). A metamethodological evaluation of psychological theories from a cross-cultural perspective. In L.H. Eckensberger, W.J. Lonner, & Y.H. Poortinga (Eds.), Cross-cultural contributions to psychology (pp. 255±275). Lisse, Netherlands: Swets & Zeitlinger. Eckensberger, L., & Burgard, P. (1983). The cross-cultural assessment of normative concepts: Some considerations on the af®nity between methodological approaches and preferred theories. In S.H. Irvine & J.W. Berry (Eds.), Human assessment and cultural factors (pp. 459±480). New York: Plenum Press. Fernandez, J.W. (1994). Culture and transcendent humanization: On the ``dynamics of the categorical''. Ethos, 59, 143±167. Fischer, M.M.J. (1996). Commentary: Science in anthropology. Anthropological Newsletter, 36 (4), 11. Flavell, J.H., Zhang, X.D., Zou, H., Dong, Q., & Qi, S. (1983). A comparison between the development of the appearance-reality distinction in the People's Republic of China and the United States. Cognitive Psychology, 15, 459±466. Gardiner, H., Mutter, J., & Kosmitzki, C. (1997). Lives across cultures: Crosscultural human development. Boston: Allyn & Bacon. Gauvin, M. (2001). The social context of cognitive development. New York: Guilford Press. Green®eld, P.M. (1984). A theory of the teacher in the learning activities of everyday life. In B. Rogoff & J. Lave (Eds.), Everyday cognition (pp. 117±138). Cambridge, MA: Harvard University Press. Green®eld, P.M. (1997). Culture as process: Empirical methods for cultural psychology. In J.W. Berry, Y.H. Poortinga, & J. Pandey (Eds.), Handbook of cross-cultural psychology: Vol. 1. Theory and method (2nd ed., pp. 301±346). Boston: Allyn & Bacon. Guba, E.G. (Ed.). (1990). The paradigm dialog. Newbury Park, CA: Sage. Hannerz, U. (1996). Transactional connections: Cultures, people, places. London: Routledge. Harkness, S., & Super, C.M. (1983). The cultural construction of child development. Ethos, 11, 221±231. Harkness, S., & Super, C.M. (1992). The developmental niche: A theoretical framework for analysing the household production of health. Social Science and Medicine, 38, 217±226. Harkness, S., & Super, C.M. (Eds.). (1996). Parents' cultural belief systems. Their origins, expressions, and consequences. New York: Guilford Press. Harre, R., & Secord, P.F. (1972). The explanation of social behaviour. Oxford: Blackwell.

284

Dasen and Mishra

Harris, P., & Heelas, P. (1979). Cognitive processes and collective representations. Archives EuropeÂennes de Sociologie, 20, 211±241. Haselagar, W.S.G. (1997). Cognitive science and folk psychology: The right frame of mind. London: Sage. Irvine, S.H., & Berry, J.W. (Eds.). (1988). Human abilities in cultural context. Cambridge: Cambridge University Press. Jahoda, G. (1992). Crossroads between culture and mind. New York: Harvester/ Wheatsheaf. Jahoda, G. (1995). The ancestry of a model. Culture and Psychology, 1, 11±24. Jahoda, G., & Krewer, B. (1997). History of cross-cultural and cultural psychology. In J.W. Berry, Y.H. Poortinga, & J. Pandey (Eds.), Handbook of cross-cultural psychology: Vol. 1. Theory and method (2nd ed., pp. 1±42). Boston: Allyn & Bacon. KagitcËibasi, C. (1997). Individualism and collectivism. In J.W. Berry, M.H. Segall, & C. KagitcËibasi (Eds.), Handbook of cross-cultural psychology: Vol. 3. Social psychology (2nd ed., pp. 1±49). Boston: Allyn & Bacon. Keller, H., Poortinga, Y.H., & Schoelmerich, A. (Eds.). (2002). Between biology and culture: Perspectives on ontogenetic development. Cambridge: Cambridge University Press. Lillard, A. (1998). Ethnopsychologies: Cultural variations in theories of mind. Psychological Bulletin, 123, 3±32. Matthews, C.G. (1992). Truth in labelling: Are we really an international society? Journal of Clinical and Experimental Neuropsychology, 14, 418±426. Miles, M., & Huberman, A.M. (1994). Qualitative data analysis: An expanded sourcebook. Thousand Oaks, CA: Sage. Miller, J.G. (1997). Theoretical issues in cultural psychology. In J.W. Berry, Y.H. Poortinga, & J. Pandey (Eds.), Handbook of cross-cultural psychology: Vol. 2. Basic processes and human development (pp. 85±128). Boston: Allyn & Bacon. Mishra, K. (1998). Cognitive style of Tharu children in relation to daily life activities and experiences of schooling. Unpublished doctoral dissertation, Banaras Hindu University, India. Mishra, R.C. (1997). Cognition and cognitive development. In J.W. Berry, P.R. Dasen, & T.S. Saraswathi (Eds.), Handbook of cross-cultural psychology: Vol. 2. Basic processes and human development (2nd ed., pp. 143±175). Boston: Allyn & Bacon. Mishra, R.C. (1999). Research on education in India. In S. PeÂrez & P.R. Dasen (Eds.), Open ®le: Educational research. Special issue of Prospects, Quarterly Review of Comparative Education, XXIX, 3 (pp. 335±348). Geneva: IBE/ UNESCO. Mishra, R.C., Sinha, D., & Berry, J.W. (1996). Ecology, acculturation and psychological adaptation: A study of Adivasis in Bihar. New Delhi, India: Sage. Moore, J.D. (1997). Visions of culture. London: Sage. Nsamenang, B. (1992). Human development in cultural context. Beverly Hills, CA: Sage. PeÂrez, S., & Dasen, P.R. (1999). Introduction. In S. PeÂrez & P.R. Dasen (Eds.), Open ®le: Educational research. Special issue of Prospects, Quarterly Review of Comparative Education, XXIX, 3 (pp. 327±334). Geneva: IBE/UNESCO. Poortinga, Y.H. (1997). Towards convergence? In J.W. Berry, Y.H. Poortinga, & J.

14. Cross-cultural views 285 Pandey (Eds.), Handbook of cross-cultural psychology: Vol. 1. Theory and method (2nd ed., pp. 347±387). Boston: Allyn & Bacon. Pourtois, J.-P., & Desmet, H. (1988). EpisteÂmologie et instrumentation en sciences humaines. LieÁge, Belgium: Mardaga. Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. New York: Oxford University Press. Sahlins, M. (1994). Goodbye to Tristes Tropes: Ethnography in the context of modern world history. In R. Borofsky (Ed.), Assessing cultural anthropology (pp. 377±394). New York: McGraw-Hill. Sapru, S. (1999). Parental practices and the identity development of adolescents. A study of Indian families in Delhi and Geneva. Unpublished PhD thesis, University of Geneva. Schlegel, A., & Barry, H. (1991). Adolescence: An anthropological enquiry. New York: Free Press. Schliemann, A., Carraher, D., & Ceci, S. (1997). Everyday cognition. In J.W. Berry, P.R. Dasen, & T.S. Saraswathi (Eds.), Handbook of cross-cultural psychology: Vol. 2. Basic processes and human development (2nd ed., pp. 177±216). Boston: Allyn & Bacon. Schraw, G. (1989). On the development of adult metacognition. In C. Smith & T. Pourchot (Eds.), Adult learning and development. Perspectives from educational psychology (pp. 89±1067). Mahwah, NJ: Lawrence Erlbaum Associates Inc. Schraw, G., & Moshmann, D. (1995). Metacognitive theories. Educational Psychology Review, 7, 351±371. Segall, M.H. (1993). La psychologie culturelle: un nouveau territoire? In F. Tanon & G. Vermes (Eds.), L'individu et ses cultures (pp. 91±97). Paris: L'Harmattan. Segall, M.H., Dasen, P.R., Berry, J.W., & Poortinga, Y.H. (1999). Human behavior in global perspective: An introduction to cross-cultural psychology (rev. 2nd ed.). Boston: Allyn & Bacon. Serpell, R. (1992). Applied developmental psychology. In M.H. Bornstein & M.E. Lamb (Eds.), Developmental psychology: An advanced textbook (pp. 677±729). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Shweder, R.A. (1990). Cultural psychology±what is it? In J.W. Stigler, R.A. Shweder, & G. Herdt (Eds.), Cultural psychology. Essays on comparative human development (pp. 1±42). Cambridge: Cambridge University Press. Shweder, R.A., & Sullivan, M.A. (1993). Cultural psychology: Who needs it? Annual Review of Psychology, 44, 497±523. Silverman, A.J. (1981). Psychophysiological and brain lateralization studies in ®eld dependence±independence. In S. Wapner & J. Demick (Eds.), Field dependenceindependence: Cognitive style across the life span (pp. 61±77). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Sinha, D. (1997). Indigenizing psychology. In J.W. Berry, Y.H. Poortinga, & J. Pandey (Eds.), Handbook of cross-cultural psychology: Vol. 1. Theory and method (2nd ed., pp. 129±170). Boston: Allyn & Bacon. Strohschneider, S., & GuÈss, D. (1998). Planning and problem solving. Differences between Brazilian and German students. Journal of Cross-Cultural Psychology, 29, 695±716. Super, C., & Harkness, S. (1986). The developmental niche: A conceptualisation at the interface of child and culture. International Journal of Behavioral Development, 9, 545±570.

286

Dasen and Mishra

Super, C.M., & Harkness, S. (1997). The cultural structuring of child development. In J.W. Berry, P.R. Dasen, & T.S. Saraswathi (Eds.), Handbook of cross-cultural psychology: Vol. 2. Basic processes and human development (2nd ed., pp. 1±39). Boston: Allyn & Bacon. Tanon, F. (1994). A cultural view on planning: The case of weaving in Ivory Coast. Tilburg, Netherlands: Tilburg University Press. Tomasello, M. (1999). The cultural origins of human cognition. Cambridge, MA: Harvard University Press. Valsiner, J., & Lawrence, J. (1997). Human development in culture across the life span. In J.W. Berry, P.R. Dasen, & T.S. Saraswathi (Eds.), Handbook of crosscultural psychology: Vol. 2. Basic processes and human development (2nd ed., pp. 69±106). Boston: Allyn & Bacon. Valsiner, J., & Litvinovic, G. (1996). Processes of generalization in parental reasoning. In S. Harkness & C.M. Super (Eds.), Parents' cultural belief systems. Their origins, expressions, and consequences (pp. 56±82). New York: Guilford Press. Vinden, P.G. (1996). Junin Quechua children's understanding of mind. Child Development, 67, 1707±1716. Vinden, P.G. (1999). Children's understanding of mind and emotion: A multiculture study. Cognition and Emotion, 13, 19±48. Wahi, S., & Johri, R. (1994). Questioning a universal theory of mind: Mental-real distinctions made by Indian children. Journal of Genetic Psychology, 155, 503± 508. Wassmann, J. (1993). Der kognitive Aufbruch in der Ethnologie. In J. Wassmann & P.R. Dasen (Eds.), Alltagswissen/Les savoirs quotidiens/Everyday cognition (pp. 95±133). Fribourg, Switzerland: Editions Universitaires. Whiting, J.W.M. (1977). A model for psychocultural research. In P.H. Leiderman, S.R. Tulkin, & A. Rosenfeld (Eds.), Culture and infancy: Variations in the human experience (pp. 29±48). New York: Academic Press. Witelson, S.F., & Swallow, J.A. (1988). Neuropsychological study of the development of spatial cognition. In J.S. Davis, M. Kritchevsky, & U. Bellugi (Eds.), Spatial cognition: Brain bases and development (pp. 373±409). Hillsdale, NJ: Lawrence Erlbaum Associates Inc.

15 Transitions in adolescent research Nancy L. Galambos and Bonnie J. Leadbeater University of Victoria, Canada

Like adolescence as a stage of life, research on adolescence has undergone many transitions in the century that has passed since the publication of G.S. Hall's (1904) two volumes entitled: Adolescence: Its psychology and its relation to physiology, anthropology, sociology, sex, crime, religion, and education (New York: Appleton Press). Historians speculated that adolescence, as a life stage, was an invention of industrialised cultures. It was a life stage created by increased needs for technical skills and education before entering the labour force and by the risks to moral and physical health that unsupervised, unemployed youth could encounter as they drifted into cities (Kett, 1977). Decade by decade, researchers have not only continued to study the list of topics in Hall's title, but have also reinvented adolescence by modifying and adding to this listÐfocusing research on puberty, formal operational thinking, identity development, generation gaps, family and peer in¯uences, school transitions, and contexts of development. Research speci®c to this stage of life now annually ®lls several journals, textbooks, and special topics books on issues such as teenage parenting, puberty, substance use, and girls' development. This research commands the attention of several disciplines of study as well as several professions such as health, education, religion, and justice. Today's challenges to youth continue to be thought of in terms of opportunities and risks. Challenges emanate not only from adolescents' engagement in risky behaviours (delinquency, alcohol and drug use, interpersonal violence, unprotected sexual activity, early parenting, and school dropout), but also from dramatic social changes that in¯uence their lives. Increases in poverty, homelessness, and unemployment as well as cutbacks in education, health, and prevention services have disproportionately negative impacts on young people and their families. The major causes of mortality in adolescence (including injuries and suicides) are founded in a complex array of social, economic, psychological, and biological factors. At the same time, most young people encounter adolescence better nourished and in more optimal physical health than ever before and are increasingly challenged by a diversity of opportunities to support their developing identities and talents. Many youth, at least in the Western

288

Galambos and Leadbeater

world, have access to the ®nancial resources necessary to participate in pursuits as wide-ranging as sports, music, art, travel, shopping, and computer-based activities. In this chapter we identify current and emerging directions in research on adolescence, wondering what social changes drive these new concerns and what the list will look like at the turn of the next century. We refer speci®cally to the need for research on adolescent cognition, positive outcomes, and resilience in contexts of adversity. We also point to changes in our research tools that allow for both more comprehensive statistical pictures of the patterns and processes of developmental changes as well as a more in-depth qualitative picture of the diversity of adolescent experiences, both within and across different cultures and ethnic or racial groups. Finally, we observe the need for increasing collaboration across disciplines and professions for creating social policies and contexts that can enhance young peoples' development as they face the rapidly changing and challenging contexts of the new millennium. INTER-INDIVIDUAL DIFFERENCES IN INTRA-INDIVIDUAL CHANGE Replacing the historical emphasis of research on universal, stage-like changes in adolescence, current and emerging trends in research on adolescent development naturally emphasise learning more about differences in how individuals change over time (i.e., their developmental trajectories) on many dimensions, including the physical, cognitive, emotional, and social. Moreover, research is increasingly devoted to understanding how changes in these dimensions are linked to one another. Our ability to examine and understand the interrelations among multiple dimensions of development is enhanced by new theoretical models that integrate domains of development and by advances in measurement (e.g., brain imaging techniques). The prominent focus on intra-individual change and interrelations among multiple aspects of development is evident in four trends in the study of adolescence.

Researchers increasingly recognise the need for more research on the cognitive, neurocognitive, and neurobiological changes of adolescence Although cognitive development has been cited as a core change of adolescence that accompanies other critical physical, emotional, and social transitions, following a burst of studies based on Piaget's theory of formal operational thinking, adolescent cognition became a largely neglected area of research (Keating & Sasse, 1996). A few notable exceptions include studies of age-related differences in information-processing capabilities

15. Transitions in adolescent research

289

(Kail, 1991) and in prospective memory (Ceci & Bronfenbrenner, 1985). These studies have demonstrated that by 14 years of age, adolescents are nearly as ef®cient as adults in speed of cognitive processing (Kail, 1991) and engage in effective time-monitoring strategies (Ceci & Bronfenbrenner, 1985). Another promising pocket of research pertains to adolescents' decision-making with respect to risk-taking behaviours and judicial competence. Research has shown that by middle adolescence, teens reason about such decisions in ways that are similar to adults' imperfect reasoning (Beyth-Marom & Fischhoff, 1997). As the gap in research on adolescents' cognitive development becomes more visible, it is imperative and likely that energetic and talented scholars will be attracted to the examination of how cognitive processes in adolescence are linked to other domains of adolescent development (e.g., peer relations) and behaviour. Similarly, although there is some documentation of neurological changes (e.g., decreases in synaptic connections; decreasing brain metabolism rates) prior to and during adolescence (Banich, 1997), how these developments map onto cognitive, emotional, and behavioural functioning remains a mystery. New methods of imaging the brain, such as functional magnetic resonance imaging (fMRI) and event-related potentials (ERP), offer exciting opportunities to learn more about the connections among the brain, cognition, and emotion in adolescence (Nelson & Bloom, 1997; Pine, 2001). fMRI has been used to demonstrate signi®cant reductions in volume of grey matter between childhood and adolescence (Sowell et al., 1999; Thompson et al., 2000). Links between these growth patterns and other physical, social, emotional, and behavioural changes await further research. Pine (2001) has proposed an integrative model that links brain states, as measured using fMRI, with emotion. The goal is to eventually understand the neural substrate of psychiatric disorders involving emotion dysregulation (i.e., mood and anxiety disorders such as depression and social phobia). Very few studies have linked brain functioning with emotion or emotional disorders, but those that have been done have been con®ned to samples of adults (Pine, 2001).

More research is likely to examine the role of emotion regulation in the adaptation of adolescents In the last decade, the issue of emotion regulation and its links to internalising and externalising behaviours in children has been an important topic of concern (Lemerise & Arsenio, 2000). Eisenberg and Fabes (1992), for example, proposed that both emotion regulation (control of the intensity of internal feeling states and physiological processes through attentional and cognitive processes) and emotion-related behavioural regulation (control of the behavioural concomitants of emotion) are key components in healthy development and the prevention of problem behaviour and psychopathology. Lower levels of emotion and behavioural regulation have been linked to

290

Galambos and Leadbeater

higher levels of externalising problems in elementary school-age children (Eisenberg et al., 2000). Few studies in adolescence, however, have been guided by an explicit emotion-regulation perspective (Galambos & Costigan, in press). Given concern with problems that ®rst appear or escalate in adolescence (e.g., depression, eating disorders, suicide, delinquency), it is likely that the use of an emotion-regulation perspective will advance our knowledge of the mechanisms that result in problems of adaptation. As an example, Mahady Wilton, Craig, and Pepler (2000) found that maladaptive emotional displays (e.g., aggression) were implicated in the victimisation experiences of children; observational data suggested that both bullies and victims had some dif®culties with emotional regulation. The adolescent's temperament is likely to interact with emotion regulation behaviours to affect his or her adjustment. Adolescents whose temperaments are naturally more emotionally intense in response to external events or those who are easily distressed might have greater challenges in managing emotions and behaving appropriately in distressing situations compared to adolescents who are endowed with easier temperamental characteristics. Moreover, parental socialisation strategies are likely to play a role in the development of emotion regulation (Galambos & Costigan, in press). For instance, a longitudinal study found that parental punitive reactions to children's negative emotions at ages 6 to 8 were linked to increases in these children's problem behaviours and negative emotions by early adolescence (ages 10 to 12) (Eisenberg et al., 1999). Recent work (Lengua, Wolchik, Sandler, & West, 2000; Stice & Gonzales, 1998) suggested that parenting behaviours such as rejection and inconsistent discipline can interact with temperamental characteristics such as positive emotionality and impulsivity in the prediction of young adolescents' adjustment (i.e., depression, antisocial behaviour). These results point to the value of considering adolescent adaptation from an emotion-regulation perspective that includes socialisation experiences.

More research is examining cultural, ethnic, and gender differences in adolescents' development Theories of adolescent moral, cognitive, and identity development that grew out of studies of male participants and illuminated the course of male adolescent development remain the foundation of much of our understanding of normative development in adolescence. The experiences of girls and adolescents from cultural minority groups have only begun to be described by adolescent researchers in the last decade of the past century (e.g., Leadbeater & Way, 1996). The research literature continues to be almost silent on the experiences of Native adolescents (see Mitchell & O'Nell, 1998, for an exception). Asian, African, and Hispanic youth have also been neglected in past research, although to be sure, there is a growing appreciation for investigating their development (Asakawa & Csikszentmihalyi,

15. Transitions in adolescent research

291

1998; Greenberger, Chen, Beam, Whang, & Dong, 2000; Quintana, CastanÄeda-English, & Ybarra, 1999; Steinberg, Dornbusch, & Brown, 1992). Adolescents with homosexual orientations (Savin-Williams, 1998a, 1998b), disabilities (Costigan, Floyd, Harter, & McClintock, 1997), and those in foster care (Rhodes, Haight, & Briggs, 1999) are deserving of and likely to be the focus of more intensive research in the new millennium. Methodologies that are sensitive to diversity (including qualitative analyses) are beginning to reveal what is unique and positive in the development of young people from non-European cultures and ethnic groups (McLoyd & Steinberg, 1998). Collaborations of researchers with youth and adult members of these groups that value their joint participation in all steps of the research (formulating questions, developing suitable methodologies, and interpreting and implementing ®ndings) are increasing (Walsh, MacMillan, & Parent, 2000). Combinations of anthropological, sociological, and psychological perspectives and methodologies may also reveal the nature and sources of diversity in adolescents' development.

Research on the transition of youth to young adulthood is rapidly expanding A recurring question in the literature on adolescence has been ``When does adolescence end?'' (Arnett & Taber, 1994). Embodied in this question are issues pertaining to the length of the transition, which is highly variable, the markers of adult status (e.g., chronological, social), and the processes by which the transition occurs (Shanahan, 2000). Arnett (2000) has proposed a concept called ``emerging adulthood,'' which is a period of role exploration, marked more by a lack of normative expectations than by clear paths and milestones, occurring in the late teens and twenties. Shanahan (2000) has presented some evidence to suggest that since the 1960s, there has been increased variability in the paths to adult roles. That is, the life course seems to be more and more individualised, as evidenced in the variable sequencing and overlap among transition markers such as school completion, marriage, entering the workforce, living independently, and becoming a parent. This variability may be connected both to short-term changes in the economy and labour force as well as to historical events such as wars. Moreover, the transition to adulthood is affected by socioeconomic status, race and ethnicity, and gender (Bynner, 2000; Shanahan, 2000). Until recently, our understanding of the transition to adulthood was based largely on the experiences of college students, a population easy for researchers to access. Now there are examples of longitudinal studies, with more diversity in sampling, that examine selected aspects of the transition to adulthood (Bachman, Wadsworth, O'Malley, Johnston, & Schulenberg, 1997; Jessor, Donovan, & Costa, 1991; Leadbeater & Way, 2001). Bachman et al. (1997) showed, for example, that marriage, pregnancy, and parenthood were associated with signi®cant decreases in substance use. On the

292

Galambos and Leadbeater

other hand, moving away from parents and getting divorced were transitions that were associated with increases in substance use. These shifts in the use of cigarettes, alcohol, marijuana, and cocaine are attributable, in part, to the impact of role transitions such as marriage on attitudes towards substance use and on recreational activities such as going out in the evening (Bachman et al., 2002). As other longitudinal studies of this nature appear, we will learn more about emerging adulthood and the processes that connect role transitions with changes in behaviour. An emerging trend in this area is the search for de®nitions and understandings of the transition to adulthood from the perspectives of young people experiencing it (Arnett, 1994; Shulman & Ben-Artzi, in press; TiltonWeaver, Vitunski, & Galambos, 2001). Arnett (1994), for instance, found that college students did not identify the attainment of adult status with role transitions such as completing education, marriage, and parenthood. Rather, the important transition markers were less tangible, including accepting responsibility for one's actions and forming one's own beliefs and values rather than relying on parents. In another study, poor teenage mothers based ``being grown up'' on their capacity for ®nancial and residential independence from their own parents rather than from their roles as parents (Leadbeater & Way, 2001). The value of understanding young people's conceptions of what it means to be an adult, or their ``implicit theories of adulthood,'' lies in the role that subjective markers might play in the life-planning process. Implicit theories of adulthood are presumed to help shape the life course, as individuals construct their futures to be in line with what they expect and hope to achieve (Ross, 1989; Shanahan, 2000). CONTEXTS OF DEVELOPMENT In the 1980s, there was a shift in research, such that contexts of adolescent development (e.g., family, school, and peer contexts) became important targets for examination (Steinberg & Shef®eld Morris, 2001). Although it is easy to consider contexts in terms of physical settings such as the school, a broad perspective on contextual in¯uences incorporates behavioural, psychological, social, and historical elementsÐexempli®ed in three current trends.

Emphasis on understanding the context and co-occurrence of adolescent problems is increasing There is a long history of research on the problems of youth, including risk behaviours (e.g., alcohol abuse), mental health outcomes such as depression, school-related troubles, delinquency, and violence. Typically, however, past research examined single problems (typi®ed by investigations of the correlates or sequelae of cigarette smoking, substance abuse, or

15. Transitions in adolescent research

293

sexual risk-taking, but not their interrelations) (Jessor, 1993; Ketterlinus & Lamb, 1994). Research examining the prevalence, antecedents, and consequences of multiple problems in adolescents has become important for two reasons. First, better predictions of the future developmental trajectories of adolescents can be made by understanding how they function on a variety of levels in multiple contexts. For instance, we know that adolescents experiencing multiple problems are most at risk for longer-term dif®culties; research examining comorbidity of problems will help to distinguish high-risk youth from other groups (Petersen, Richmond, & Leffert, 1993). Second, a more comprehensive understanding of how adolescents' problems and behaviours operate together will lead to more effective programmes designed to prevent or intervene into actions that can be potentially destructive to adolescents' mental and physical health (Lerner & Galambos, 1998). A research base is accumulating on the patterns of co-occurrence of problems in adolescence, including risk behaviours, mental health symptoms such as depression, and school-related problems (Ensminger, 1990; Fergusson, Horwood, & Lynskey, 1994; Galambos & Tilton-Weaver, 1998; Roeser, Eccles, & Freedman-Doan, 1999). Ensminger (1990), for example, examined familial and school antecedents to the co-occurrence of substance use, sexual risk, and violent behaviours in a sample of urban adolescents. Adolescents showing multiple problem behaviours (e.g., substance use, sex, and assault) were more frequently truant from school, had more permissive parental curfew rules, and were more aggressive in ®rst grade (at least among males) compared to adolescents with no problem behaviours. Interestingly, adolescent females who reported frequent sexual intercourse (but no other problems) were different from the no-problem and the multiproblem females on a number of variables. The origins, consequences, and co-occurrence of nonsexual and sex-related problem behaviours deserve further empirical consideration (Ketterlinus, Lamb, & Nitz, 1994; Roeser et al., 1999).

Research is accumulating on positive psychosocial outcomes and resilience in high-risk circumstances With the current focus on the problems of youth (in the media and in the ®eld of research on adolescence), it is easy to neglect consideration of their competencies, health-enhancing behaviours, and prosocial activities. Frequently, researchers have designated the absence of negative outcomes as the presence of healthy outcomes. Indeed, most measures of good health are simply low scores on measures of poor health, for example, not showing depressive symptoms. In order to successfully design policies and interventions that promote health, create protective environments, or prevent health-compromising behaviours, knowledge about adolescents' healthenhancing behaviours, protective contexts, and positive outcomes is required

294

Galambos and Leadbeater

(Lerner, Fisher, & Weinberg, 2000; Lerner & Galambos, 1998). Increasingly, research has de®ned positive psychosocial outcomes in terms of such assets or attributes as psychosocial maturity, self-esteem, the possession of positive coping strategies, school engagement, community involvement, church attendance, involvement in positive peer networks, and health-enhancing behaviours (e.g., healthy diet, regular exercise) (Allen, Kuperminc, Philliber, & Herre, 1994; Galambos & Ehrenberg, 1997; Jessor, Turbin, & Costa, 1998; Sandler, Ayers, Suter, Schultz, & Twohey, in press). Research on resilience in contexts of adversity has also begun to focus on the development of competence in multiple domains in childhood and adolescence (Luthar, Cicchetti, & Becker, 2000; Masten & Coatsworth, 1998). Since the late 1970s, researchers following this approach have investigated the mechanisms and processes that lead some individuals to thrive despite adverse life circumstances. For example, the focus of most past research on adolescent mothers has been on the predictors of, the risk factors associated with, or the negative outcomes resulting from teenage parenthood; but we now know that the majority of adolescent mothers experience positive outcomes (e.g., high school graduation, employment) (The Alan Guttmacher Institute, 1994; Furstenberg, Brooks-Gunn, & Morgan, 1987; Leadbeater, 1996; Way & Leadbeater, 1999). Research investigating pathways towards positive outcomes among adolescent mothers has begun to reveal the multiple individual, family, and community components of resiliency. Although early research operationalised resiliency as a unidimensional construct (e.g., educational achievement or social competence), recent research reveals the problems of considering only one or two domains in an assessment of resiliency. Luthar, Dorenberger, and Zigler's (1993) research on resiliency with inner-city adolescents found that those students who reported having good grades and being socially competent also reported having high levels of psychological distress. Other researchers have indicated that whereas individuals may succeed in one or two areas of their lives, their success in these areas may mask more internal struggles such as depression or low self-esteem (Masten & Coatsworth, 1998; Wright, Masten, Northwood, Hubbard, 1997). Resiliency is not a static, unidimensional process but a dynamic, multidimensional one that changes over time (Luthar et al., 2000; Masten & Coatsworth, 1998). The ways one is able to acquire, maintain, and express strength in the midst of adversity will vary depending upon developmental stage, external expectations, family relationships, and life circumstances (Masten & Coatsworth, 1998; McLoyd, 1998).

Greater attention is being paid to the impact of historical and social contexts and change on adolescent development The path-breaking work of Elder (1974) illustrated the signi®cance of historical and social changes such as the Great Depression in the lives and

15. Transitions in adolescent research

295

development of children and adolescents. The amount of research devoted to examining such issues is increasing, particularly in Europe. Ongoing research in Germany is a case in point. The dramatic reuni®cation of East and West Germany in 1990 was the starting place for a number of studies comparing, in formerly divided regions of Germany, young people's family and peer relations, externalising and internalising behaviours, physical health, self-esteem, identity, and vocational choices (see, e.g., Bynner & Silbereisen, 1999; Silbereisen & von Eye, 1999). Rather than providing consistent examples of striking regional differences among adolescents, such research is pointing to more subtle and complex interrelations among social change, region of residence, and adolescent development. Noack and Kracke (1997) reported, for instance, that West German adolescents' higher levels of connectedness to their families buffered them from the potentially negative effects of increased economic strain. In East German adolescents, however, higher family connectedness was associated with elevated adjustment problems among adolescents whose family ®nancial situation worsened. In another comparison of East and West German adolescents, sampled shortly after reuni®cation, the timing of some transitions in adolescence differed (Silbereisen, 2000). Adolescents in former East Germany accomplished care of oneself and taking over household responsibilities (e.g., making own lunch, doing household chores) earlier than did their West German counterparts. Moreover, there was no gender difference in the rates of these activities in former East Germany, as there was in adolescents from the West (with girls making these transitions earlier). There was little evidence of regional differences in romantic activities, but initial occupational preferences were established earlier for adolescent girls and boys in the East. Boys in the East engaged earlier in political discussions than did boys in the West. Such differences were attributed to institutional differences in the two political regions (Silbereisen, 2000). In former East Germany, for example, most women were in the labour force, perhaps resulting in adolescents' earlier immersion in household work. Schooling for adolescents in the East was also geared towards earlier vocational selections. The achievement of a better understanding of the in¯uences of historical and social contexts on adolescents is likely to be facilitated through international collaborations among researchers. One example of a massive, collaborative undertaking is EURONET, the European Network Project, which involved 11 research teams from different nations across Europe (Bulgaria, Czech Republic, Finland, France, Germany, Hungary, Norway, Poland, Romania, Russia, and Switzerland) and 1 in the United States. The members of this project shared knowledge, experience, methods, and instruments to explore cross-national differences in adolescents' leisure activities and contexts, future orientations, perceived control, psychological well-being, and other aspects of adolescents' lives (Alsaker & Flammer,

296

Galambos and Leadbeater

1999). Cross-national comparisons of time use showed some interesting differences. Adolescents in France, the United States, and Finland spent the most time in school, for example, whereas adolescents in Bulgaria, Poland, and Romania spent the least. Results on average duration of sleep per night showed that, in most of the 12 nations, older adolescents (15 to 17 years) slept signi®cantly less than did younger adolescents (13 to 14 years). Overall, adolescents in the United States, Russia, and Poland slept less than 8 hours a night whereas adolescents in Switzerland, France, and Romania slept close to 9 hours. Cross-national differences in working for money were striking, with adolescents in the United States spending more than twice as much time working than adolescents in any of the other nations. In some nations (France, Russia) adolescents did not work at all. By observing how psychosocial outcomes in adolescence differ across national, political, and ideological contexts, and linking these outcomes to differences in social structures and programmes, important insights can be gained. Such information can be used in the design of social policies, programmes, and interventions that work to ameliorate problems and facilitate healthy transitions in youth. For example, knowledge about the associations between schooling and occupational trajectories in different contexts can be used to enhance the school-to-work transition (Hamilton & Hamilton, 2000). METHODOLOGICAL APPROACHES Some of the most exciting trends in research on adolescence include new methodological approaches, or at least, a growing appreciation for methods and research tools that have been available but under-utilised. Here we mention the person-centred approach, which was central to Block's (1971) volume, Lives through time. Also of importance are qualitative approaches that have been in existence for a very long time, and have been employed by scholars from many disciplines. As interdisciplinary work proliferates, scholars of adolescent development increasingly see the value in collecting and analysing qualitative data.

Person-centred approaches are used more often to examine constellations of individual, family, and/or community characteristics that affect developmental trajectories in adolescence Past research on adolescence has followed a variable-centred approach, which examines interrelationships among single variables through, for example, regression analysis. This can be contrasted with person- (or pattern-)centred approaches, which consider individuals as consisting of multiple attributes (e.g., cognitive, biological, behavioural) that are integrated into an organised system (Magnusson & ToÈrestad, 1993). Advocates

15. Transitions in adolescent research

297

of the person-centred approach argue that the adolescent's pro®le across a number of indicators may carry more meaning for understanding his or her development than do single variables (Magnusson & Cairns, 1996). Methods such as cluster analysis or con®gurational frequency analysis, which group individuals on the basis of their pro®le of characteristics, exemplify the person-centred approach to understanding inter-individual differences. The groups may then be compared on other relevant variables to explore the antecedents, correlates, or consequences of a particular pro®le (e.g., Roeser et al., 1999). This approach is appealing, as researchers seek increasingly to understand within-group differences in the development of adolescents who differ in combinations of personal resources and challenges, access to protective factors in their lives, and exposure to risky environments. Recently, this approach has been used alone and in combination with variable approaches to provide us with important knowledge about adolescent development (Galambos & Tilton-Weaver, 2000; Salmivalli, 1998; Schulenberg, Wadsworth, O'Malley, Bachman, & Johnston, 1996; Stattin & Magnusson, 1996). Schulenberg et al. (1996), for instance, used person-centred analyses to classify and then predict longitudinal trajectories of binge drinking in adolescents from the ages of 18 to 24 years. Risk factors present in their senior year in high school (i.e., having less concrete future plans and a stronger desire to drink) distinguished senior-year binge drinkers who would continue to drink (i.e., the chronic binge drinkers) from the senioryear binge drinkers who would subsequently reduce their drinking (i.e., decreased binge drinking group). Moreover, individuals with a pattern of increased binge drinking were low on internal personal control and on conventionality. Such analyses lend a richness to the data that would be missed by following only a variable-centred approach.

Qualitative data analysis is being used to achieve a more in-depth understanding of individual differences in adolescent development To gain as much knowledge as possible about adolescent development, qualitative data analysis will become increasingly valued. At the same time that methodological designs are becoming more complicated and statistical strategies even more sophisticated (e.g., the use of hierarchical linear modelling; Maggs & Schulenberg, 1998), there is a growing appreciation of the potential uses for qualitative data. Some of the most powerful studies have incorporated and combined qualitative and quantitative components. For example, in examining the ethnic identity of African and Mexican American adolescents, Phinney and Devich-Navarro (1997) used interview data to identify three patterns of ethnic/American identi®cation (blended biculturalism; alternating biculturalism; separated from American culture). In quantitative analyses, scales typically used to measure ethnic identity validated the qualitative analysis and themes from the qualitative interviews

298

Galambos and Leadbeater

were used to provide rich descriptions of these young people's experiences. Qualitative analyses have also been used to illuminate ``voices'' not typically heard in the development of standardised measures (Way, 1998), and to challenge stereotypes of de®cits created by the past, almost exclusive, focus on risks in minority group adolescents (e.g., Leadbeater & Way, 2001). APPLICATION TO PROGRAMMES AND POLICY Ultimately, the goal of scholars in the ®eld of adolescence is to see their skills, research, and knowledge used to promote healthy adolescent development through the application of their work to the design of prevention and intervention programmes, as well as social policies. One emerging trend exempli®es this desire to reach out to adolescents in the real world.

University±community and interdisciplinary collaborations are increasing in number as a way to solve the problems of youth and improve the probability of their healthy futures The way in which most developmental research has been conducted has contributed to gaps in public awareness, youth programming, policy making, and research evidence. Academics from single disciplines who investigate issues relevant to youth programming and policy expend substantial effort to establish informal, often short-lived collaborations with schools or occasionally community programmes. At best, ®ndings are reported back to collaborators and are published in academic journals. Theory and the knowledge generated only gradually affect general beliefs and social change through the education of students in universities who become involved in efforts to promote youth well-being or through media efforts to inform the public following crises involving young people. Innovative programming often lacks access to evidence of successful strategies and knowledge about the changing contexts in which children and youth are developing. Such knowledge could not only increase the effectiveness of programmatic activities but could also improve the quality of programme evaluations and of their advocacy and programme funding decisions. Programming in several areas relevant to youth, cooperative education efforts, research collaborations, and policy decisions have suffered from inadequate access to research evidence. Examples of these problems include: the application of adult disease models of chemical dependency to the prevention and treatment of smoking, drug, and alcohol use in adolescence despite lack of outcome data to support their effectiveness with adolescents (Kellam & Anthony, 1998); the proliferation of youth programmes that offer services but do not promote youth competence, empowerment, and involvement (Fetterman, Kafterian, & Wandersman, 1996); and failure to address contextual factors that limit or enhance youth well-being (Leadbeater & Way, 1996).

15. Transitions in adolescent research

299

Gaps in knowledge can be addressed by cross-discipline and crossprofessional communication and collaboration. Models for action research have begun to emerge that specify the principles for collaborations among researchers and institutions or individuals who will use the results (Lerner et al., 2000; Small, 1996; Weinberg & Erickson, 1996). Such collaborations will in¯uence the relevance of questions asked, increase commitment to programme implementation, provide a more complete picture of how programmes work, and make available a context for the application of ®ndings. Collaborative research strategies with a wide community basis of support and interdisciplinary expertise also have the potential to enhance the scienti®c quality of research through the use of randomised and longitudinal designs to evaluate interventions, access to more representative samples, and the creation of possibilities for investigations of complex questions (e.g., at individual and neighbourhood levels) (Kellam, Ling, Merisca, Brown, & Ialongo, 1998). Researchers generate knowledge about a wide range of policy-relevant issues for adolescents, such as father involvement, after-school care, mental health concerns, teenage parenting, poverty, maltreatment and sexual abuse, and injury prevention. However, researchers have only rarely seen an advocacy role as being within their domain of activities and very few training programmes see social policy development as a concern for scholars of adolescence. Gaps in policy and research-based knowledge stem from differences in the settings, goals, priorities, orientations, methods, and time schedules that characterise the distinct worlds of research and policy (e.g., see Shonkoff, 2000). Although there may well be few absolute ``truths'' in social science research, existing research-based knowledge can and should guide social policy. Long-time advocate for children and policy maker, psychologist Ed Zigler, tells us that often enough is known for action (Zigler & Hall, 2000, p. 20): Some social conditions are so severe and pressing as to warrant policy construction or immediate pragmatic action even in the absence of consensus on the data. When this occurs, the efforts made need to be based on wise judgement, well informed by scholarship, with provisions for the evaluation and revision of programs as needed. We should not wait, for instance, for outcome data showing that children who have eaten at least one nutritious meal a day get higher test scores than hungry children in order to support policies that alleviate children's hunger. Some journals are now encouraging authors to include statements about the policy-relevance of their research. The Society for Research in Child Development's Social Policy Reports reviews social issues that are informed by psychological research and a new American Psychological Association journal entitled: Children's Services: Social Policy, Research, and Practice

300

Galambos and Leadbeater

(Lawrence Erlbaum Associates Inc.) directly solicits articles that relate research to policy concerns. One exemplar of collaborative work between researchers and policy makers has resulted in an edited volume entitled Investing in children, youth, families, and communities: Strengths-based research and policy (Maton, Schellenbach, Leadbeater, & Solarz, in press). Each of the chapters draws on knowledge that illuminates the individual competencies and environmental conditions that make a positive difference in the development of children and youth, families, and communities who are dealing with adverse circumstances. Each chapter was written by both academics and individuals directly involved in in¯uencing public policy at the local, state, or federal level. Research collaborations between researchers and policy makers are needed to set in motion efforts for give and take, clari®cation, and re-clari®cation of the complementary interests of both groups. CONCLUSION At the beginning of this new millennium it is possible to see the advances in knowledge that will be made by researchers in collaboration with those who have a stake in the healthy development of adolescents. We conclude with one observation that is all too often overlooked; that, by and large, youth have been seen as more the objects of research and intervention rather than as stakeholders in the research process. Bibby and Posterski (1992, p. 301) argued that ``young people in the 1990s do not see themselves as agents of change. They are like mirrors on the walls of society. Their attitudes and behaviour mainly re¯ect the status quo. They have not been able to carve out much of a world that is uniquely their own. Distinctiveness has eluded them''. If youth do not see themselves as agents of change, it may indeed be because the mirrors of society have not re¯ected or valued their agency as participants in it. We have been preoccupied with creating services for rather than with youth. Mechanisms for involving young people in setting priorities for research questions and in disseminating and applying research ®ndings can be created through youth advisory groups and youth community-action programmes. The bene®ts of viewing youth as a resource rather than merely the focus of research concern and programming have yet to be fully explored. By involving youth as collaborators in the processes of research, programme development, and social policy construction, piecemeal solutions to their problems may be avoided, and a focus on their potentials could be realised. REFERENCES The Alan Guttmacher Institute. (1994). Sex and America's teenagers. New York: Author.

15. Transitions in adolescent research

301

Allen, J.P., Kuperminc, G., Philliber, S., & Herre, K. (1994). Programmatic prevention of adolescent problem behaviors: The role of autonomy, relatedness, and volunteer service in Teen Outreach Program. American Journal of Community Psychology, 22, 617±638. Alsaker, F.D., & Flammer, A. (1999). The adolescent experience: European and American adolescents in the 1990s. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Arnett, J.J. (1994). Are college students adults? Their conceptions of the transition to adulthood. Journal of Adult Development, 1, 213±224. Arnett, J.J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist, 55, 469±480. Arnett, J.J., & Taber, S. (1994). Adolescence terminable and interminable: When does adolescence end? Journal of Youth and Adolescence, 23, 517±537. Asakawa, K., & Csikszentmihalyi, M. (1998). The quality of experience of Asian American adolescents in academic activities: An exploration of educational achievement. Journal of Research on Adolescence, 8, 241±262. Bachman, J.G., O'Malley, P.M., Schulenberg, J., Johnston, L.D., Bryant, A.L., & Merline, A.C. (2002). The decline of substance use in young adulthood: Changes in social activities, roles, and beliefs. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Bachman, J.G., Wadsworth, K.N., O'Malley, P.M., Johnston, L.D., & Schulenberg, J.E. (1997). Smoking, drinking, and drug use in young adulthood: The impacts of new freedoms and new responsibilities. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Banich, M.T. (1997). Neuropsychology: The neural bases of mental function. Boston: Houghton Mif¯in. Beyth-Marom, R., & Fischhoff, B. (1997). Adolescents' decisions about risks: A cognitive perspective. In J. Schulenberg, J.L. Maggs, & K. Hurrelmann (Eds.), Health risks and developmental transitions during adolescence (pp. 110±135). Cambridge: Cambridge University Press. Bibby, R.W., & Posterski, D.C. (1992). Teen trends: A nation in motion. Toronto: Stoddart. Bynner, J. (2000). Social change and the sequencing of developmental transitions. In L.J. Crockett & R.K. Silbereisen (Eds.), Negotiating adolescence in times of social change (pp. 89±103). Cambridge: Cambridge University Press. Bynner, J., & Silbereisen, R.K. (Eds.). (1999). Adversity and challenge in life in the New Germany and England. Baskingstoke, UK: Macmillan. Ceci, S.J., & Bronfenbrenner, U. (1985). ``Don't forget to take the cupcakes out of the oven'': Prospective memory, strategic time-monitoring, and context. Child Development, 56, 152±164. Costigan, C.L., Floyd, F.J., Harter, K.S.M., & McClintock, J. (1997). Family process and adaptation to children with mental retardation: Disruption and resilience in family problem-solving interactions. Journal of Family Psychology, 11, 515±529. Eisenberg, N., & Fabes, R.A. (1992). Emotion regulation and the development of social competence. In M.S. Clark (Ed.), Emotion and social behavior: Vol. 14. Review of personality and social psychology (pp. 119±150). Newbury Park, CA: Sage. Eisenberg, N., Fabes, R.A., Shepard, S.A., Guthrie, I.K., Murphy, B.C., & Reiser,

302

Galambos and Leadbeater

M. (1999). Parental reactions to children's negative emotions: Longitudinal relations to quality of children's social functioning. Child Development, 70, 513± 534. Eisenberg, N., Guthrie, I.K., Fabes, R.A., Shepard, S., Losoya, S., Murphy, B.C., Jones, S., Poulin, R., & Reiser, M. (2000). Prediction of elementary school children's externalizing problem behaviors from attentional and behavioral regulation and negative emotionality. Child Development, 71, 1367±1382. Elder, G.H. (1974). Children of the Great Depression. Chicago: University of Chicago Press. Ensminger, M.E. (1990). Sexual activity and problem behaviors among black, urban adolescents. Child Development, 61, 2032±2046. Fergusson, D.M., Horwood, L.J., & Lynskey, M.T. (1994). The comorbidities of adolescent problem behaviors: A latent class model. Journal of Abnormal Child Psychology, 22, 339±354. Fetterman, D.M., Kafterian, S.J., & Wandersman, A. (1996). Empowerment evaluations. Thousand Oaks, CA: Sage. Furstenberg, F.F. Jr., Brooks-Gunn, J., & Morgan, S.P. (1987). Adolescent mothers in later life. Cambridge: Cambridge University Press. Galambos, N.L., & Costigan, C.L. (in press). Emotional and personality development in adolescence. In R.M. Lerner, M.A. Easterbrooks, & J. Mistry (Eds.), Comprehensive handbook of psychology: Vol. 6. Developmental psychology. New York: John Wiley. Galambos, N.L., & Ehrenberg, M.F. (1997). The family as health risk and opportunity: A focus on divorce and working families. In J. Schulenberg, J.L. Maggs, & K. Hurrelmann (Eds.), Health risks and developmental transitions during adolescence (pp. 139±160). Cambridge: Cambridge University Press. Galambos, N.L., & Tilton-Weaver, L.C. (1998). Multiple-risk behaviour in adolescents and young adults. Health Reports, 10, 9±20. Galambos, N.L., & Tilton-Weaver, L.C. (2000). Adolescents' psychosocial maturity, problem behavior, and subjective age: In search of the adultoid. Applied Developmental Science, 4, 178±192. Greenberger, E., Chen, C., Beam, M., Whang, S.-M., & Dong, Q. (2000). The perceived social contexts of adolescents' misconduct: A comparative study of youths in three cultures. Journal of Research on Adolescence, 10, 365±388. Hamilton, S.F., & Hamilton, M.A. (2000). Research, intervention, and social change: Improving adolescents' career opportunities. In L.J. Crockett & R.K. Silbereisen (Eds.), Negotiating adolescence in times of social change (pp. 267±283). Cambridge: Cambridge University Press. Jessor, R. (1993). Successful adolescent development among youth in high-risk settings. American Psychologist, 48, 117±126. Jessor, R., Donovan, J.E., & Costa, F.M. (1991). Beyond adolescence: Problem behavior and young adult development. Cambridge: Cambridge University Press. Jessor, R., Turbin, M.S., & Costa, F.M. (1998). Protective factors in adolescent health behavior. Journal of Personality and Social Psychology, 75, 788±800. Kail, R. (1991). Processing time declines exponentially during childhood and adolescence. Developmental Psychology, 27, 259±266. Keating, D.P., & Sasse, D.K. (1996). Cognitive socialization in adolescence: Critical period for a critical habit of mind. In G.R. Adams, R. Montemayor, & T.P.

15. Transitions in adolescent research

303

Gullotta (Eds.), Psychosocial development during adolescence: Progress in developmental contextualism (pp. 232±258) Thousand Oaks, CA: Sage. Kellam, S.G., & Anthony, J.C. (1998). Targeting early antecedents to prevent tobacco smoking: Findings from an epidemiologically based randomized ®eld trial. American Journal of Public Health, 88, 1490±1495. Kellam, S.G., Ling, X., Merisca, R., Brown, C.H., & Ialongo, N. (1998). The effect of the level of aggression in the ®rst grade classroom on the course and malleability of aggressive behavior into middle school. Development and Psychopathology, 10, 169±170. Kett, J.F. (1977). Rites of passage: Adolescence in America, 1790 to the present. New York: Basic Books. Ketterlinus, R.D., & Lamb, M.E. (Eds.). (1994). Adolescent problem behaviors: Issues and research. Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Ketterlinus, R.D., Lamb, M.E., & Nitz, K.A. (1994). Adolescent nonsexual and sexrelated problem behaviors: Their prevalence, consequences, and co-occurrence. In R.D. Ketterlinus & M.E. Lamb (Eds.), Adolescent problem behaviors: Issues and research (pp. 17±40). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Leadbeater, B.J. (1996). School outcomes for minority group adolescent mothers at 28 to 36 months postpartum: A longitudinal follow-up. Journal of Research on Adolescence, 6, 629±648. Leadbeater, B.J.R., & Way, N. (Eds.). (1996). Urban adolescent girls: Resisting stereotypes, creating identities. New York: New York University Press. Leadbeater, B.J.R., & Way, N. (2001). Growing up fast: Transitions to early adulthood of inner-city adolescent mothers. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Lemerise, E.A., & Arsenio, W.F. (2000). An integrated model of emotion processes and cognition in social information processing. Child Development, 71, 107±118. Lengua, L.J., Wolchik, S.A., Sandler, I.N., & West, S.G. (2000). The additive and interactive effects of parenting and temperament in predicting adjustment problems of children of divorce. Journal of Clinical Child Psychology, 29, 232± 244. Lerner, R.M., Fisher, C.B., & Weinberg, R.A. (2000). Toward a science for and of the people: Promoting civil society through the application of developmental science. Child Development, 71, 11±20. Lerner, R.M., & Galambos, N.L. (1998). Adolescent development: Challenges and opportunities for research, programs, and policies. Annual Review of Psychology, 49, 413±446. Luthar, S., Cicchetti, D., & Becker, B. (2000). The construct of resilience: A critical evaluation and guidelines for future work. Child Development, 71, 543±562. Luthar, S.S., Dorenberger, C.H., & Zigler, E. (1993). Resilience is not a unidimensional construct: Insights from a prospective study of inner-city adolescents. Development and Psychopathology, 5, 703±717. Maggs, J.L., & Schulenberg, J. (1998). Reasons to drink and not to drink: Altering trajectories of drinking through an alcohol misuse prevention program. Applied Developmental Science, 2, 48±60. Magnusson, D., & Cairns, R.B. (1996). Developmental science: Toward a uni®ed framework. In R.B. Cairns, G.H. Elder, Jr., & E.J. Costello (Eds.), Developmental science (pp. 7±30). Cambridge: Cambridge University Press.

304

Galambos and Leadbeater

Magnusson, D., & ToÈrestad, B. (1993). A holistic view of personality: A model revisited. Annual Review of Psychology, 44, 427±452. Mahady Wilton, M.M., Craig, W.M., & Pepler, D.J. (2000). Emotional regulation and display in classroom victims of bullying: Characteristic expressions of affect, coping styles and relevant contextual factors. Social Development, 9, 226±245. Masten, A.S., & Coatsworth, J.D. (1998). The development of competence in favorable and unfavorable environments: Lessons from research on successful children. American Psychologist, 53, 205±220. Maton, K., Schellenbach, C., Leadbeater, B., & Solarz, A. (in press). Investing in children, youth, families, and communities: Strengths-based research and policy. Washington, DC: American Psychological Association. McLoyd, V. (1998). Socioeconomic disadvantage and child development. American Psychologist, 53, 185±202. McLoyd, V., & Steinberg, L. (1998). Studying minority adolescents: Conceptual, methodological and theoretical issues. Mahwah, NJ: Lawrence Erlbaum Associates Inc. Mitchell, C.M., & O'Nell, T.D. (1998). Problem and conventional behaviors among American Indian adolescents: Structure and validity. Journal of Research on Adolescence, 8, 97±122. Nelson, C.A., & Bloom, F.E. (1997). Child development and neuroscience. Child Development, 68, 970±987. Noack, P., & Kracke, B. (1997). Social change and adolescent well-being: Healthy country, healthy teens. In J. Schulenberg, J.L. Maggs, & K. Hurrelmann (Eds.), Health risks and developmental transitions during adolescence (pp. 54±84). Cambridge: Cambridge University Press. Petersen, A.C., Richmond, J.B., & Leffert, N. (1993). Social changes among youth: The United States experience. Journal of Adolescent Health, 14, 632±637. Phinney, J.S., & Devich-Navarro, M. (1997). Variations in bicultural identi®cation among African American and Mexican American adolescents. Journal of Research on Adolescence, 7, 3±32. Pine, D. (2001). Functional magnetic resonance imaging in children and adolescents: Implications for research on emotion. In J.M. Morihisha (Ed.), Review of Psychiatry: Vol. 20. Advances in brain imaging (pp. 71±105). Washington, DC: American Psychiatric Publications. Quintana, S.M., CastanÄeda-English, P., & Ybarra, V.C. (1999). Role of perspectivetaking abilities and ethnic socialization in development of adolescent ethnic identity. Journal of Research on Adolescence, 9, 161±184. Rhodes, J.E., Haight, W.L., & Briggs, E.C. (1999). The in¯uence of mentoring on the peer relationships of foster youth in relative and nonrelative care. Journal of Research on Adolescence, 9, 185±202. Roeser, R.W., Eccles, J.S., & Freedman-Doan, C. (1999). Academic functioning and mental health in adolescence: Patterns, progressions, and routes from childhood. Journal of Adolescent Research, 14, 135±174. Ross, M. (1989). Relation of implicit theories to the construction of personal histories. Psychological Review, 96, 341±357. Salmivalli, C. (1998). Intelligent, attractive, well-behaving, unhappy: The structure of adolescents' self-concept and its relations to their social behavior. Journal of Research on Adolescence, 8, 333±354. Sandler, I., Ayers, T., Suter, J., Schultz, A., & Twohey, J. (in press). Adversities,

15. Transitions in adolescent research

305

strengths and public policy. In K. Maton, C. Schellenbach, B. Leadbeater, & A. Solarz (Eds.), Investing in children, youth, families, and communities: Strengthsbased research and policy. Washington, DC: American Psychological Association. Savin-Williams, R.C. (1998a). The disclosure to families of same-sex attractions by lesbian, gay, and bisexual youths. Journal of Research on Adolescence, 8, 49±68. Savin-Williams, R.C. (1998b). Forgotten youth: Gay, lesbian, and bisexual teens. Society for Research on Adolescence Newsletter, pp. 1±2, 10. Schulenberg, J., Wadsworth, K.N., O'Malley, P.M., Bachman, J.G., & Johnston, L.D. (1996). Adolescent risk factors for binge drinking during the transition to young adulthood: Variable- and pattern-centered approaches to change. Developmental Psychology, 32, 659±674. Shanahan, M.J. (2000). Pathways to adulthood in changing societies: Variability and mechanisms in life course perspective. Annual Review of Sociology, 26, 667±692. Shonkoff, J.P. (2000). Science, policy, and practice: Three cultures in search of a shared mission. Child Development, 71, 181±187. Shulman, S., & Ben-Artzi, E. (in press). Age-related differences in the transition from adolescence to adulthood and links with family relationships. Journal of Adult Development. Silbereisen, R.K. (2000). German uni®cation and adolescents' developmental timetables: Continuities and discontinuities (pp. 104±122). Cambridge: Cambridge University Press. Silbereisen, R.K., & von Eye, A. (Eds.). (1999). Growing up in times of social change. New York: de Gruyter. Small, S.A. (1996). Collaborative community-based research on adolescents: Using research for community change. Journal of Research on Adolescence, 6, 9±22. Sowell, E.R., Thompson, P.M., Holmes, C.J., Batth, R., Jernigan, T.L., & Toga, A.W. (1999). Localizing age-related changes in brain structure between childhood and adolescence using statistical parametric mapping. Neuroimage, 9, 587±597. Stattin, H., & Magnusson, D. (1996). Antisocial development: A holistic approach. Development and Psychopathology, 8, 617±645. Steinberg, L., Dornbusch, S.M., & Brown, B.B. (1992). Ethnic differences in adolescent achievement: An ecological perspective. American Psychologist, 47, 723±729. Steinberg, L., & Shef®eld Morris, A. (2001). Adolescent development. Annual Review of Psychology, 52, 83±110. Stice, E., & Gonzales, N. (1998). Adolescent temperament moderates the relation of parenting to antisocial behavior and substance use. Journal of Adolescent Research, 13, 5±31. Thompson, P.M., Giedd, J.N., Woods, R.P., MacDonald, D., Evans, A.C., & Toga, A.W. (2000). Growth patterns in the developing brain detected by using continuum mechanical tensor maps. Nature, 404, 190±193. Tilton-Weaver, L.C., Vitunski, E.T., & Galambos, N.L. (2001). Five images of maturity in adolescence: What does ``grown up'' mean? Journal of Adolescence, 24, 143±158. Walsh, C.A., MacMillan, H.L., & Parent, J. (2000). Ontario First Nations Youth Action Initiative: Final Report. Manuscript submitted to the Laidlaw Foundation. Way, N. (1998). Everyday courage: The lives and stories of urban teenagers. New York: New York University Press.

306

Galambos and Leadbeater

Way, N., & Leadbeater, B.J. (1999). Pathways toward educational achievement among African American and Puerto Rican adolescent mothers: Reexamining the role of social support from families. Development and Psychopathology, 11, 349±364. Weinberg, R.A., & Erickson, M.F. (1996). Minnesota's children, youth, and family consortium: A university-community collaboration. Journal of Research on Adolescence, 6, 37±54. Wright, M. O'D., Masten, A.S., Northwood, A., & Hubbard, J.J. (1997). Long-term effects of massive trauma: Developmental and psychobiological perspectives. In D. Cicchetti & S.L. Toth (Eds.), Rochester Symposium on Developmental Psychopathology: Vol. 8, The effects of trauma on the developmental processes (pp. 181±225). Rochester, NY: University of Rochester Press. Zigler, E.F., & Hall, N.W. (2000). Child development and social policy. New York: McGraw-Hill.

16 The impact of longitudinal studies on understanding development from young adulthood to old age K. Warner Schaie

The Pennsylvania State University, USA

INTRODUCTION The purpose of this chapter is to review progress in our understanding of adult development in the study of behaviour over the past half-century, It became clear to me very quickly that my own perceptions of this progress have largely been shaped by my involvement in the pursuit of large-scale longitudinal studies. These studies were mostly designed to systematically identify the in¯uences that differentiate between the lucky individuals who age successfully (cf. Rowe & Kahn, 1987) and those who are exposed to a variety of hazards that can lead to early decline and inadequate functioning in the ®nal part of their lives (cf. Schaie, 1989, 1996b, c; Schaie & Willis, 2002). What follows, then, will heavily lean on what I have learned from longitudinal studies, primarily those of cognition and personality. Over the course of the last century many relevant lessons learned from the social and biological sciences have required behavioural scientists to become less parochial and more comfortable in considering the convergence of scienti®c ®ndings from adjacent disciplines. I will begin this assessment of the recent history of the developmental psychology of ageing by summarising relevant early developments of a substantive and methodological nature. I will then provide a theoretical framework that represents my understanding that behavioural change can only be understood (and predicted, for that matter) by being examined in the context of societal change (cf. Riley, Foner, & Riley, 1999). The model also gives due recognition to lasting heritable in¯uences (cf. Schaie, Plomin, Willis, GruberBaldini, & Dutta, 1992), as well as to the obvious age-related changes in the ef®ciency of the physiological infrastructure (Cristofalo, Tresini, Francis, & Volker, 1999). Then follows a more detailed discussion of what we have learned from longitudinal studies. I will therefore, by necessity, deal primarily with work that occurred in the second half of the century. More detailed accounts of the psychological aspects of adult development during the ®rst half of the century can be found in Birren (1961), Birren and Birren (1990), Birren and Schroots (2001), Jones (1959), and Riegel (1977). Finally, I will do

308

Schaie

some prognostication of what research directions we might expect in the proximal future. THE FIRST HALF OF THE CENTURY The study of adult development and ageing developed quite slowly during the ®rst half of the twentieth century. A comprehensive overview of early gerontology can be found in Cowdry's Problems of aging (1939). This volume was quite light on psychological content and even Birren's monumental Handbook of aging and the individual, published in 1959, still con®ned psychological ageing to no more than about half of this singlevolume account. An early seminal in¯uence for the psychology of ageing was a quasiautobiographical account of the last part of his life by one of the founders of American developmental psychology, G. Stanley Hall (1922). But attention to the psychological development of adults was also fostered by the need of the military during World War I to classify and assign large numbers of young and middle-aged draftees to emerging specialty roles. This need led to an extension of the mental testing movement begun by the work of Binet and Simon (1905) with children to the construction of tests suitable for adults (Yerkes, 1921). The availability of assessment instruments such as the Army Alpha test led to an interest in age-comparative studies (e.g., Jones & Conrad, 1933; Miles, 1931). Emergence of the ®eld of clinical psychology incurred further development of tests speci®cally tailored for work with adults (Wechsler, 1939) and continued the interest in studying adult age differences. An active research programme on human ageing was begun under the direction of Nathan Shock at the NIH with psychological research being initiated by James Birren in 1947. About the same time, the Nuf®eld Unit for Research on Ageing was established at the University of Cambridge (Welford, 1951). Research on ageing also began during the 1950s at the University of Bonn in Germany (Schmitz-Scherzer & Thomae, 1983). The Division on Adult Development and Aging of the American Psychological Association was founded in 1945. But formal recognition of ageing as a major research ®eld probably occurred only when the National Institute on Aging was established in 1975. THE INFLUENCE OF METHODOLOGICAL ADVANCES Attention needs to be called to the role of methodological advances in shaping our understanding of adult development (cf. Schaie, 1988. 1992). These advances have resulted in several paradigmatic shifts that have markedly changed the face of research on adult developments. The

16. Impact of longitudinal studies

309

advances to be discussed here include ®rst of all the shift from the dominance of cross-sectional data collections to the recognition of the importance of longitudinal designs. Second, there have been signi®cant advances in the measurement of change, so essential to studies of development and ageing. Third, there has been a shift from data-driven exploratory factor analysis to hypothesis-testing con®rmatory factor analysis that has been particularly relevant to the study of developmental phenomena. Fourth, there has been a conceptual shift toward treating age as the dependent rather than the independent variable, and ®nally there have been increasing efforts to study development by means of methods of latent growth curves. The paradigmatic shift from the predominance of cross-sectional studies of age differences in the study of human development to the emphasis on the need for longitudinal data was motivated by a number of critical issues facing developmental researchers. The major issue, of course, was the growing understanding of the fact that, if antecedent±consequent relationships are to be elicited in the study of development, then one must follow the same individuals over time (Mason, Mason, Winsborough, & Poole, 1973; Ryder, 1965; Schaie, 1965, 1977; Schaie & Caskie, 2001). This understanding was precipitated to some extent by the follow-up of members of longitudinal studies that began in early childhood or late adolescence when their participants reached early adulthood or middle age (e.g., the Berkeley Growth and Guidance studies; Bayley & Oden, 1955; Eichorn, Clausen, Haan, Honzik, & Mussen, 1981; or the Iowa State ROTC followup studies, Owens, 1953, 1966). Failure to replicate the decline in function from young adulthood into middle age inferred from cross-sectional data (e.g., Jones & Conrad, 1933) necessitated coming to grips with the contrasting inferences to be drawn from cross-sectional and longitudinal data. The second methodological advance that had major impact on the study of adult development actually occurred in the ®eld of measurementÐ speci®cally following some heated debates and increasing sophistication in our understanding of the measurement of change. This debate began early on with the recognition that measurement imperfections (i.e., deviations of observed scores from true scores) were likely to cumulate in gain (or loss) scores comparing multiple measurements of the same individuals (Thorndike, 1924). This issue was further developed by Lord (1956), and the debate became more heated in the 1960s when many developmentalists despaired at being able to assess change adequately (cf. Cronbach & Furby, 1970; Harris, 1963). Because studies of adult development require tests of hypotheses about directional change, difference scores have always played an important role. Fortunately, it was soon recognised that the major problem is usually con®ned to two-point studies, leading to the advocation of multiple occasion studies (Nesselroade, Stigler, & Baltes, 1980; Rogosa, Brandt, & Zimowsky, 1982; Willett, 1989) and of growth curve modelling (see later).

310

Schaie

A third paradigmatic shift was facilitated by the introduction of formal methods of con®rmatory factor analysis and structural equations modelling. Earlier methods of exploratory factor analysis were utilised in studies of adult development to determine latent constructs from observed variables. But an even more important use of factor analysis has been its application to the determination of change in psychological constructs across samples of different age or within the same sample over time (cf. Reinert, 1970). Early approaches to the study of multiple groups can be found in the writing of Thurstone (1947), Guttman (1952), and in work in the Cattell laboratory (e.g., Cattell & Cattell, 1955). However, it was not until the advent of high-speed computers that algorithms could be developed that were applicable to more than a minimal number of variables (JoÈreskog, 1971; SoÈrbom & JoÈreskog, 1978). Con®rmatory factor analysis makes it possible to assess systematically the invariance (stability) of the regression of the latent constructs, which are of primary interest to science, upon the observed variables. In studies of adult development such invariance is a singular prerequisite for the comparison of individuals and groups over long periods of time, or the comparison of groups of different individuals who differ in salient characteristics. Con®rmatory factor analysis, for example, can be used to test hypotheses about the differentiation and dedifferentiation of psychological domains across the adult lifespan (Baltes & Lindenberger, 1997; Maitland, Intrieri, Schaie, & Willis, 2000; Reinert, 1970; Schaie, Maitland, Willis, & Intrieri, 1998). A fourth important methodological development involved the paradigmatic shift from considering chronological age as a dependent rather than as an independent variable, First introduced conceptually by Wohlwill (1973), behavioural scientists soon began to realise that the study of age or duration time as a dependent variable could be operationalised via methods of survival or event-time analysis (Allison, 1984; Schaie, 1989; Singer & Willett, 1991). This approach is important in both cognitive and health psychology, because the prediction of morbidity and mortality by means of earlier behavioural characteristics requires not only the de®nition of endpoints but also the timing (respectively age) at which such endpoints are most likely to occur (e.g., Bosworth, Schaie, Willis, & Siegler, 1999). As longitudinal studies have been conducted for longer periods of time with multiple measurements, it is now possible to apply powerful methods of growth curve modelling (LGM) that allow separating patterns of individual change over time from the group averages that had previously represented the primary focus of enquiry. These methods also allow incorporating covariates and predictors of different forms of development. Multivariate growth curve methods were ®rst introduced by Tucker (1958). Again intensive development required the advent of more powerful computational resources, leading to currently popular approaches of multilevel modelling (cf. Bryk & Raudenbush, 1987; Rogosa & Willett, 1985; Rudinger & Rietz,

16. Impact of longitudinal studies

311

2001; Willett & Sayer, 1994). LGM models are of particular interest in the study of adult development, because differences in genetic predisposition and environmental exposure may result in ageing patterns that differ markedly for subsets of the population as well as for groups of individuals with either very favourable or unfavourable life experiences. A THEORETICAL FRAMEWORK FOR UNDERSTANDING ADULT COGNITIVE DEVELOPMENT An understanding of development from early adulthood to old age must include embedding what we know about development within the context of changing environmental in¯uences and changes in individuals' physiological infrastructure. Figure 16.1 displays a schematic that speci®es my view of how these in¯uences might operate over the adult life course in the special case of cognition. The schematic contains two endpoints: one important for applied issues, the other of greater concern to basic science. The ®rst endpoint is concerned with the lifelong in¯uences that affect the level of late-life cognitive functioning so essential for maintaining independence and a life characterised by high quality. The second endpoint represents the status of the cortex at life's end and is descriptive of infrastructure relevant to the maintenance of cognitive functioning that can only be determined at post mortem. In this conceptual path model, as is conventional, rectangles are used to identify those individual indicators that are observed directly whereas ovals are used to indicate the latent constructs that would be inferred from measurement models for sets of observed variables. It could, of course, be argued that the arrows in Figure 16.1, other than those directed towards the endpoints, which represent the interplay of the various causal in¯uences, may be overly simplistic. Indeed, I could have posited several reciprocal relationships. However, I have tried to keep the model as simple as possible, because it is intended primarily for its heuristic value. More complex models may indeed be needed if the model were to be ®tted to speci®c sets of data. However, I would like to assure the reader that all of the causal paths speci®ed in the model can be supported by the results of empirical investigations in my own laboratory or those of other cognitive developmentalists. Some more details may be in order with respect to the attributes of the heuristic model. The initial bases for adult cognitive functioning must, of course, be attributed to both heritable (genetic) in¯uences as well as early environmental in¯uences typically experienced within the home of the biological parents. Although some of the behaviour genetic literature suggests that much of the early environmental variance is nonshared (e.g., Plomin & Daniels, 1987), there is recent retrospective evidence that there may indeed be some early shared environmental in¯uences upon later cognitive performance (Schaie & Zuo, 2001). Both generic and early

Socioeconomic status

Current family environment

Apo-E status

Lifestyles

Chronic disease

Late-life cognitive functioning

Neuropathology at autopsy

Figure 16.1 A conceptual model for in¯uences affecting late-life psychological functioning in the sample case of cognition.

Early family environment

Heritable cognitive abilities

Midlife cognitive functioning

312 Schaie

16. Impact of longitudinal studies

313

environmental factors are thought to in¯uence midlife cognitive functioning. The early environmental in¯uences will, of course, also exert in¯uences on midlife social status (Nguyen, 2000). By contrast, virtually no correlations have been found between retrospective accounts of family environment in the family of origin, and that in the current family (Schaie & Willis, 1995). However, the current family environment does seem to in¯uence midlife cognitive performance. Genetic factors are also likely to be implicated in the rate of cognitive decline in adulthood. Thus far the beststudied gene in this context is the Apo-E gene, one of whose alleles is thought to be a risk factor for Alzheimer's disease. Apo-E status is therefore added as a factor; the expression of the gene is probably not at issue prior to midlife. Given these considerations, the causal in¯uences can now be speci®ed that determine level of cognitive function in late life as well as cortical status at autopsy. The direct in¯uences to be implicated, in addition to genes whose expression is turned on in late life, are most likely to originate in midlife. They include level of cognitive functioning in midlife, midlife lifestyles, and the incidence and severity of chronic disease. But there are indirect in¯uences attributable to the effects of midlife cognitive function and lifestyles upon chronic disease, as well as shared family in¯uences on midlife cognition and of social status upon midlife lifestyles (cf. Willis & Reid, 1999). Some of these paths represent concurrent observations that would allow alternative (respectively reciprocal) causal paths, but most of the paths speci®ed by the model represent antecedent±consequent relationships which require longitudinal data for their estimation and understanding. This leads me to turn next to a brief account of what has been learned about adult cognitive development through longitudinal studies. LONGITUDINAL STUDIES OF ADULT DEVELOPMENT It is interesting to note that from the very beginning of empirical enquiry on development beyond adolescence, substantive concerns were limited primarily to the areas of intellectual development and personality traits. Investigators interested in the age-related aspects of learning and memory largely adopted the paradigms popular in early experimental child psychology and thus limited themselves to age-comparative studies of young and old adults. Only recently have we seen an interest in this area in studies that would investigate the developmental mechanisms by use of longitudinal paradigms (see Salthouse, 1999). But for the areas of intellectual development, and cognition as well, cross-sectional studies predominated until the late 1930s and clouded our understanding of adult development due to the confusion of age-related development with secular changes expressed as cohort effects.

314

Schaie

For example, successive studies of adult age differences in intelligence reported asymptotic peak performance to occur at ever later ages. Thus, Terman (1916), in his standardisation of the Stanford±Binet, thought that intelligence peaked at age 16, and Yerkes's (1921) estimate of World War I soldiers was even lower at age 13. By 1939, Wechsler's standardisation sample peaked between 18 and 24 years, similar to the top ages found by Jones and Conrad (1937). In my ®rst cross-sectional study (Schaie, 1958), the peak ages had moved up to the late twenties and early thirties.

Types of longitudinal studies The initial longitudinal studies that have informed our understanding of adult development were of two types. First, there were studies that began with a focus on early childhood and child-rearing practices, but whose participants were followed into adulthood. A prime example of such a study is the follow-up of the Berkeley Growth and Guidance studies (Bayley & Oden, 1955; Eichorn et al., 1981). A second group of studies traced participants who had been assessed as young adults as part of their college experience and reassessed in midlife or later. An example of such studies is Owens' (1953, 1966) follow-up of persons in their ®fties who had ®rst been assessed as ROTC members during World War I.

Contrast between cross-sectional and longitudinal studies The earlier cross-sectional studies (e.g., Jones & Conrad, 1933) had placed peak performance in intelligence and other positive psychological attributes in late adolescence or early young adulthood. with linear decline thereafter. By contrast, the longitudinal follow-up studies suggested that psychological growth continued generally into early midlife and for some variables (notably the verbal abilities) at least into the ®fties. My own early work (Schaie, 1958) with cross-sectional data on mental abilities and rigidity± ¯exibility in adults over the age range from the twenties to the sixties suggested that peak performance now occurred in the twenties and thirties, but linear decline still prevailed thereafter.

The Seattle Longitudinal Study I soon became convinced that the cross-sectional versus longitudinal issue needed to be confronted directly by following a structured cross-sectional sample covering most of the adult lifespan over time. I therefore designed a study that converted my original cross-sectional study into a series of short-term longitudinal studies of mental abilities each extending over a simultaneous 7-year period (Schaie & Strother, 1968). My replicated crosssectional ®ndings were quite similar to the original ®ndings, but the longitudinal data showed later ages of peak performance, maintenance of

16. Impact of longitudinal studies

315

average function on most abilities until the sixties, and only modest decline through the seventies. Further extensions of these studies (with some longitudinal data over as long as 42 years) over the past several decades have consistently replicated these ®ndings, with dramatic declines not experienced until the eighties are reached (Schaie, 1983, 1996b; Schaie & Hertzog, 1986; Schaie & Labouvie-Vief, 1974). I will come back to other matters learned from this study later on, but ®rst must dwell a little on other longitudinal approaches to the study of ageing.

Lifespan-oriented studies The early work on adult development was pretty much oriented within the context of a lifespan development framework (cf. Baltes, 1987, 1997), but the burgeoning ®eld of geropsychology soon divided into at least two rather different orientations. Some of us remained committed to the notion that an understanding of the ageing process required the careful charting of human development at least across the entire adult lifespan, if not beginning our enquiries in childhood. This orientation, which I share, holds that what is of primary interest is the understanding of the mechanisms that contribute to the behavioural differences between youth and old age within a process that extends across the lifespan. Most of the studies mentioned earlier illustrate this type of approach.

Studies originating in late life The second orientation, sometimes labelled the ``clinker method'' (after the residue that remains when charcoal is produced), considers the characteristics of the elderly as their primary interest, and would investigate the ageing process only from that period of life when a categorical transformation, such as leaving the work of world, or family dissolution due to death of a spouse, has begun. Representative longitudinal studies of the second orientation therefore began at an advanced age. Several of these types of studies began in the age decades of the sixties during the 1960s and 1970s. Perhaps the most prominent of studies begun in late life has been the Duke Longitudinal Study (Palmore, Busse, Maddox, Nowlin, & Siegler, 1985). But many others can be found in the literature conducted in a variety of industrial societies (e.g., Canada: Hultsch, Hertzog, Dixon, & Small, 1998; Germany: Schmitz-Scherzer & Thomae, 1983; Rott, 1993; Israel: Shanan, 1993; Sweden: Svensson, Dehlin, Hagberg, & Samuelsson, 1993; United Kingdom: Rabbitt, 1993). Other more recent studies have focused on following those in very late life (e.g., Baltes & Mayer, 1999; Poon, Sweeney, Clayton, & Merriam, 1992). These studies generally ®nd less decrement than would be suggested by cross-sectional data, only small average decline in the sixties, with

316

Schaie

increasingly steep decrement for each successive age decade. There is also a strong suggestion that decline accelerates as a precursor of eventual death (Berg, 1996; Bosworth et al., 1999). But most importantly, all of the studies call attention to vast individual differences in rate of change occurring for individuals of all levels of early psychological functioning and socioeconomic status. Thus while the frequency of individuals who show some decline increases at a near logarithmic rate once the sixties are passed, there are still rare individuals to be found who function exceedingly well even in the mid-eighties. What many of these studies also suggest is that there may be an individualised pattern of developmental trajectories (cf. Magnusson, 1998). For example, in the case of mental abilities, by the time they reach the sixties most individuals will have experienced a signi®cant drop in one of their abilities, but that ability will be speci®c to the individual (Schaie, 1989). Indeed, it is only from the longitudinal study of adult development that it is possible to enquire into possible mechanisms and/or causes of these vast individual differences in developmental progressions through adulthood. For a more detailed review of longitudinal studies of adults now in progress, as well as classic studies now completed, see Schaie and Hofer (2001). STRUCTURAL INVARIANCE OF CONSTRUCTS ACROSS AGE A bothersome problem in studies of adult development has been the everpresent question of whether apples and oranges are being compared in agecomparative studies. The experimental ageing literature is replete with reports of studies in which college students are compared with senior volunteers. Fashionable also have been the so-called Brinley plots, in which mean performance of young adults and older persons across tasks of varying dif®culty have been charted in order to obtain ratios that de®ne the extent of the disadvantage of the older group (e.g., Cerella, 1990). All this work is based on the assumption that the constructs studied (and technically the regression of the constructs on the marker variables used to measure them) remain invariant across the different groups, or in longitudinal studies within groups (cf. Meredith, 1993). The advent of con®rmatory factor analysis has made formal tests of structural invariance across different groups and over time practicable, and there have been a number of recent studies that have applied these methodologies to problems in adult development. The limited literature thus far available provides some reassurance, but also gives grounds for caution. Although good invariance has been demonstrated across much of midlife, there are signi®cant structural differences between young and old adults, as well as between elderly men and women (e.g., Maitland et al., 2000; Schaie et al., 1998). However, there are selected marker variables that do seem to be remarkably stable across the entire age range, which

16. Impact of longitudinal studies

317

therefore deserve special attention. These marker variables include unspeeded measures of vocabulary (ETS Advanced Vocabulary), as well as selected markers of inductive reasoning (PMA Letter Series), spatial orientation (STAMAT Object Rotations), and numeric ability (ETS Subtraction and Multiplication).

SOURCES OF INDIVIDUAL DIFFERENCES Individual differences in trajectories and patterns of decline invite investigations of in¯uences that can be identi®ed as potential causes of unfavourable or successful ageing. Although some of these in¯uences can be represented by psychological constructs, others must clearly be sought in the physiological infrastructure underlying effective behaviour, as well as the sociodemographic factors that will either bene®t or constrain individual development. The developmental impact of all of these in¯uences can, of course, only be assessed when relevant longitudinal data are available. Not all individuals decline in lock-step. Although linear or quadratic forms of decline may be detectable for large groups, individual decline appears to occur far more frequently in a stair-step fashion. Individuals will have unfavourable experiences, to which they respond with a modest decline in cognitive functioning, and then tend to stabilise for some time, perhaps repeating this pattern several times prior to their demise.

Genetic influences Certainly, genetic endowment will account for a substantial portion of individual differences. For example, evidence of heritability of adult intelligence has been provided from both twin studies (e.g., Finkel, Pedersen, McGue, & McClearn, 1995) and family studies (e.g., Schaie et al., 1992). Similarly, genetic variance has been identi®ed for a variety of personality traits as well as their stability in adulthood (e.g., Pedersen & Reynolds, 1998). Nevertheless, in most populations sampled, heritability explains on average at most 25% of cognitive abilities and less in the realm of personality. Hence, there are many other important sources of individual differences in psychological ageing that have been implicated.

Chronic disease The onset of intellectual decline seems to be markedly affected by the presence or absence of several chronic diseases. Most reliably identi®ed as such in¯uences thus far are cardiovascular disease, diabetes, neoplasms, and arthritis. All of these diseases are risk factors for the occurrence of early

318

Schaie

cognitive decline, as is a low level of overall health. Persons functioning at high cognitive levels are also more likely to seek earlier and more competent medical intervention in the disabling conditions of late life, and they are more likely to comply more effectively with preventive and ameliorative regimens that tend to stabilise their physiological infrastructure. They are also less likely to engage in high-risk lifestyles and to respond more readily to professional advice that maximises their chances for survival and reduction of morbidity (e.g., Bosworth & Schaie, 1999).

Environmental circumstances Candidates for circumstances that might account for individual differences in cognitive ageing, for example, have been all those aspects of the environment that are likely to enhance intellectual stimulation (cf. Schaie & O'Hanlon, 1990). There is considerable evidence to suggest that the onset of intellectual decline is postponed for individuals who live in favourable environmental circumstances, as would be the case for those persons characterised by a high socioeconomic status. These circumstances include above-average education, histories of occupational pursuits that involve high complexity and low routine, and the maintenance of intact families. Likewise, risk of cognitive decline is lower for persons with substantial involvement in activities typically available in complex and intellectually stimulating environments. Such activities include extensive reading, travel, attendance at cultural events, pursuit of continuing education activities, and participation in clubs and professional associations (Arbuckle, Gold, Andres, & Schwartzman, 1992; Gribbin, Schaie, & Parham, 1980). Intact families, our most important individual support system, also reduce risk of cognitive decline. In addition, it has been found that cognitive decline is less severe for those married to a spouse with high cognitive status, with the lower-functioning spouse at the beginning of a marriage tending to increase his or her level vis-aÁ-vis the higher-functioning spouse (Gruber-Baldini, Schaie, & Willis, 1995).

Psychological characteristics Cognitive styles Associated also with differential intellectual ageing have been individual differences in the cognitive style of rigidity±¯exibility. It can now be concluded that an individual's self-report of a ¯exible personality style at midlife, as well as ¯exible performance on objective measures of motor± cognitive perseveration tasks, is predictive of a reduction in the risk of cognitive decline. It seems that the availability of a more ¯exible response style is useful when it one must cope with the vicissitudes of advanced age.

16. Impact of longitudinal studies

319

Perceptual and response speed Ageing effects on many cognitive abilities tend to be confounded with the perceptual and response speed required to process the tasks used to measure these abilities. Thus, individuals who remain at high levels of perceptual speed are also at an advantage with respect to the maintenance of such other abilities. Life satisfaction Finally, those individuals who rate themselves as being satis®ed with their life's accomplishment in midlife or early old age seem to be at an advantage when assessed at a later age. Also, individuals who overestimate the rate of their cognitive decline might well be engaging in self-ful®lling prophecies if they reduce their active participation in life to compensate for decline that is perceived by the individual but has not actually occurred.

Developmental interventions Once we understand some of the factors that in¯uence individual differences in adult development, we are then challenged as scientists to attempt to obtain experimental control by means of targeted interventions. Whereas investigators in child development might be interested in accelerating normal development of children's competencies, adult developmentalists' primary goal is to postpone functional decline or compensate for cohort differences (see later). An essential feature of the design of interventions with the older adult is the need to determine whether the individual has declined from a previously attained higher level or whether he/she is simply functioning at a low level that represents that person's developmental asymptote. If longitudinal data are available, it becomes possible to determine not only whether an intervention has resulted in improved levels of function, but also to determine whether remediation or new learning has occurred. And again only longitudinal studies can inform us whether or not developmental interventions have had long-lasting effects (cf. Willis & Nesselroade, 1990; Willis & Schaie, 1994).

The impact of generational differences One of the major contributions made by developmentalists interested in adulthood has been the attention given to generational (or cohort) differences in psychological characteristics, including shifts in the rate of developmental change across successive cohorts. What is at issue here is that we cannot rest certain that once we have explicated developmental

320

Schaie

mechanism and life course trajectories that these will stay put. As Riley (Riley et al., 1999) vividly portrays, changing societies change the life course of individuals, who in turn during their life course modify the formative nature of society. The cohort issue has long been of central concern in sociology and demography. In developmental psychology the cohort was ®rst seen as a confound that created the unwelcome discrepancies between cross-sectional and longitudinal ®ndings. Hence, initial concerns were with attempts to control for what were perceived to be ``experimental artefacts''. Cohort variance in studies of infancy and childhood may indeed be no more than a minor disturbance, unlikely to overshadow or hide universal developmental laws. By contrast, cohort variance often assumes a substantively meaningful role in the study of adult development. Individual differences in adulthood, prior to advanced old age, are largely moderated by environmental context. It is therefore important to understand how successive cohorts differ from one another (Schaie, 1996a; Willis, 1989). Examples of major contexts that differ dramatically for successive generations can be found in the increased level of educational attainment, the adoption of healthy lifestyles with respect to exercise and diet, and major advances in healthcare that contribute to the extension of life and functionality. The increase in societal support during early old age has resulted in compensatory behaviours that optimise selective psychological functions (cf. Baltes & Carstensen, 1996). The increased functionality, however, also expands demands by society upon the individual's development. Societal implications of this interplay can be seen in the current discussions in the United States on the need to postpone to later ages the eligibility for social security payments. It is argued that individuals now live longer and remain functional to later ages than was the case in the past. The developmental psychologist, when asked to apply expertise to this issue, is faced with the question as to what assumptions must be made to support these contentions. I would argue that two assumptions are required. First, it is necessary to show that older individuals today function at higher levels than their previous age peers, and second, evidence is needed that the rate of declining competence has slowed for successive cohorts. Figure 16.2 shows some relevant data from the Seattle Longitudinal Study bearing on these questions. What should be noted in Figure 16.2 is the pattern of increased levels of performance on two important dimensions of cognitive competence (inductive reasoning and spatial ability) for successive cohorts over the age range from 60 to 74. This, of course, is the age span during which most individuals could typically be expected to retire, and the data support the assumption that performance over this age range has indeed risen. On the other hand, there is little support for the second assumption of changes in rates of ageing. The data suggest that the rate of change from age 60 to age 74 has not changed markedly over what represents an almost 30-year period of time.

16. Impact of longitudinal studies

321

Inductive reasoning

T-score means

55.0

50.0

45.0 Cohort 1896 Cohort 1903 Cohort 1910

Cohort 1917 Cohort 1924

40.0 60

67

74

Age Spatial orientation 55.0

T-score means

52.5

50.0

47.5

Cohort 1896

45.0

Cohort 1903

Cohort 1917

Cohort 1910

Cohort 1924

42.5 60

67 Age

74

Figure 16.2 Changes in level and rate of ageing across ®ve cohorts for the mental abilities of inductive reasoning and spatial orientation.

FUTURE IMPLICATIONS The study of adult psychological development is increasingly informed by relevant neighbouring disciplines that investigate the genetic basis, physiological infrastructure, and societal context of the developing individual. Hence, I would predict that the study of the life course of single

322

Schaie

psychological variables that was common in the ®rst two-thirds of the past century will be largely displaced by multivariate, multidisciplinary efforts. Indeed, most of the more recent longitudinal studies of adults already display these characteristics (e.g., Baltes & Mayer, 1999). Cross-sectional investigations will continue to have a role as exploratory pilot studies or as the ®rst stage of a prospective longitudinal study. Given the willingness of public agencies to invest in more comprehensive investigations we are likely to see more programmatic long-range investigation that may frequently include experimental paradigms and in particular interventions designed to modify rate of development. With our increasing sophistication in psychological measurement, we will take advantage of the work on structural invariance to develop better scales. The introduction of item response theory to educational measurement was instrumental in improving the precision of measurement scales and their applicability to populations at different stages of the range of human talent. But item response theory remains essentially a univariate procedure, and its proponents have not yet dealt with matters such as changes in item differences across age and time. Once we have used con®rmatory factor analyses to identify those robust marker variables that do well in measuring behaviour across the entire adult lifespan, we must then return to improving the measurement characteristics of those scales that we will continue to use. We are beginning to develop common archives that will make available to our colleagues large data sets from many different populations and covering a wide range of psychological attributes. This development calls even more urgently for the development of ``gold standards'' for a core set of measures. Such measures can then be used to link disparate data sets and provide the basis for substantively meaningful meta-analyses. I would also predict that the investigation of adult development will increasingly turn to the identi®cation of mechanisms and processes that underlie developmental interventions and that will be relevant to public policy questions. Many of the mechanisms that may lead from full functionality to increasing impairment are likely to become measurable ®rst in midlife (cf. Willis & Reid, 1999). Paradoxically, such investigations may rekindle interest in age-comparative and short-term longitudinal studies to be able to respond to societal problems of an immediate nature. However, age-comparative studies of this nature will not use the young/old comparison paradigms currently so common in the experimental literature. Rather, they must introduce designs that compare groups differing in age modestlyÐhence, making it possible to control or match for other individual difference variables. Hence, we should see more studies comparing different segments of midlife and of old age, rather than attempting to cover broad age slices. Recent theorising on losses and gains in advanced old age (cf. Baltes, 1997), suggests that we have been successful in identifying and introducing

16. Impact of longitudinal studies

323

behavioural and environmental compensations for the increasing physiological frailness all of us can expect as we age that seem to work well in early old age. We have been far less successful in dealing with the needs of the very old, where the compensatory methods that may work well at retirement no longer suf®ce. Some would argue that the only solution here is to ®nd ways to enhance the spiritual experience and psychological accommodation of the very old to the reality of end-of-life losses. Nevertheless, my inveterate optimism would argue that even here other possibilities for successful interventions may still lie ahead. Hence, studies of the very old, including investigation of behavioural and environmental prostheses, would seem to be part of the need for a strong applied psychology of adult development, one that will ®nd ways to enhance the quality of our existence towards the very end of human existence. REFERENCES Allison, P.D. (1984). Event history analysis: Regression for longitudinal event data. Beverly Hills, CA: Sage. Arbuckle, T.Y., Gold, D.P., Andres, D., & Schwartzman, A. (1992). The role of psychosocial context, age, and intelligence in memory performance of older men. Psychology and Aging, 7, 25±36. Baltes, P.B. (1987). Theoretical propositions of life-span developmental psychology: On the dynamics between growth and decline. Developmental Psychology, 23, 611±626. Baltes, P.B. (1997). On the incomplete architecture of human ontogenesis: Selection, optimization and compensation as foundations of developmental theory. American Psychologist, 52, 366±381. Baltes, M.M., & Carstensen, L.L. (1996). The process of successful ageing. Ageing and Society, 16, 397±422. Baltes, P.B., & Lindenberger, U. (1997). Emergence of a powerful connection between sensory and cognitive functions across the adult life span: A new window to the study of cognitive aging. Psychology and Aging, 12, 12±21. Baltes, P.B., & Mayer, K.U. (Eds.). (1999). The Berlin Aging Study: Aging from 70 to 100. Cambridge: Cambridge University Press. Bayley, N., & Oden, M. (1955). The maintenance of mental ability in gifted adults. Journal of Gerontology, 10, 91±107. Berg, S. (1996). Aging, behavior and terminal decline. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (4th ed., pp. 323±337). San Diego, CA: Academic Press. Binet, A., & Simon, T. (1905). MeÂthodes nouvelles pour le diagnostic du niveau intellectuel des anormaux. L'AnneÂe Psychologique, 11, 102±191. Birren, J.E. (Ed.). (1959). Handbook of aging and the individual. Chicago: University of Chicago Press. Birren, J.E. (1961). A brief history of the psychology of aging (Part II). Gerontologist, 1, 127±134. Birren, J.E., & Birren, B.A. (1990). The concepts, models and history of the

324

Schaie

psychology of aging. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (3rd ed., pp. 3±20). New York: Academic Press. Birren, J.E., & Schroots, J.J.F. (2001). On the origins of geropsychology. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (5th ed., pp. 3±28). San Diego, CA: Academic Press. Bosworth, H.B., & Schaie, K.W. (1997). The relationship of social environment, social networks, and health outcomes in the Seattle Longitudinal Study: Two analytic approaches. Journal of Gerontology: Psychological Sciences, 52B, P197± 205. Bosworth, H.B., & Schaie, K.W. (1999). Survival effects in cognitive function, cognitive style, and sociodemographic variables in the Seattle Longitudinal Study. Experimental Aging Research, 25, 121±139. Bosworth, H.B., Schaie, K.W., Willis, S.L., & Siegler, I.C. (1999). Age and distance to death in the Seattle Longitudinal Study. Research on Aging, 21, 723±738. Bryk, A.S., & Raudenbush, S.W. (1987). Application of hierarchical linear models to assessing change. Psychological Bulletin, 101, 147±158. Cattell, R.B., & Cattell, A.K.S. (1955). Factor rotation for proportional pro®les: Analytic solution and an example. British Journal of Statistical Psychology, 8, 83± 92. Cerella, J. (1990). Aging and information-processing rate. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (3rd ed., pp. 201±221). San Diego, CA: Academic Press. Collins, L.M. (1996). Measurement of change in research on aging: Old and new issues from an individual change perspective. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (4th ed., pp. 38±56). San Diego, CA: Academic Press. Cowdry, E.V. (1939). Problems of aging. Baltimore, MD: Williams & Wilkins. Cristofalo, V.J., Tresini, M., Francis, M.N.K., & Volker, C. (1999). Biological theories of senescence. In V.L. Bengtson & K.W. Schaie (Eds.), Handbook of theories of aging (pp. 98±112). New York: Springer. Cronbach, L.J., & Furby, L. (1970). How should we measure changeÐor should we? Psychological Bulletin, 74, 68±80. Eichorn, D.H., Clausen, J.A., Haan, N., Honzik, M.P., & Mussen, P.H. (1981). Present and past in middle life. New York: Academic Press. Finkel. D., Pedersen, N.L., McGue, M., & McClearn, G.E. (1995). Heritability of cognitive abilities in adulthood: Comparison of Minnesota and Swedish twin studies. Behavior Genetics, 25, 421±431. Gribbin, K., Schaie, K.W., & Parham, I.A. (1980). Complexity of life style and maintenance of intellectual abilities. Journal of Social Issues, 36, 47±61. Gruber-Baldini, A.L., Schaie, K.W., & Willis, S.L. (1995). Similarity in married couples: A longitudinal study of mental abilities and ¯exibility±rigidity. Journal of Personality and Social Psychology: Personality Processes and Individual Differences, 69, 191±203. Guttman, L. (1952). Multiple group methods for common-factor analysis: Their basis, computation and interpretation. Psychometrika, 17, 209±222. Hall, G.S. (1922). Senescence, the last half of life. New York: Appleton. Harris, C.W. (Ed.). (1963). Problems in measuring change. Madison, WI: University of Wisconsin Press.

16. Impact of longitudinal studies

325

Hultsch, D.F., Hertzog, C., Dixon, R.A., & Small, B.J. (1998). Memory changes in the aged. New York: Cambridge University Press. Jones, H.E. (1959). Intelligence and problem solving. In J.E. Birren (Ed.), Handbook of aging and the individual (pp. 700±738). Chicago: University of Chicago Press. Jones, H.E., & Conrad, H.S. (1933). The growth and decline of intelligence: A study of a homogenous group between the ages of ten and sixty. Genetic Psychology Monographs, 13, 223±298. JoÈreskog, K.G. (1971). Simultaneous factor analysis in several populations. Psychometrika, 36, 409±426. Lord, F.M. (1956). The measurement of growth. Educational and Psychological Measurement, 16, 421±437. Maitland, S.B., Intrieri, R.C., Schaie, K.W., & Willis, S.L. (2000). Gender differences in cognitive abilities: Invariance of covariance and latent mean structure. Aging, Neuropsychology and Cognition, 7, 32±53. Magnusson, D. (1998). The logic and implication of a person-oriented approach. In R.B. Cairns, L.R. Bergman, & J. Kagan (Eds.), Methods and models for studying the individual (pp. 33±64). Thousand Oaks, CA: Sage. Mason, K.G., Mason, W.H., Winsborough, H.H., & Poole, W.K. (1973). Some methodological problems in cohort analyses of archival data. American Sociological Review, 38, 242±258. Meredith, W. (1993). Measurement invariance, factor analysis and factorial invariance. Psychometrika, 58, 525±543. Miles W.R. (1931). Measures of certain human abilities through the life-span. Proceedings of the National Academy of Sciences, 17, 627±632. Nesselroade, J.R., Stigler, S.M., & Baltes, P.B. (1980). Regression towards the mean and the study of change. Psychological Bulletin, 88, 622±637. Nguyen, H. (2000). Family in¯uences and environmental similarities in adult parent± offspring and sibling dyads. Unpublished doctoral dissertation, Pennsylvania State University, PA. Owens, W.A., Jr. (1953). Age and mental abilities: A longitudinal study. Genetic Psychology Monographs, 48, 3±54. Owens, W.A., Jr. (1966). Age and mental abilities: A second adult follow-up. Journal of Educational Psychology, 57, 311±325. Palmore, E., Busse, E.W., Maddox, G.L., Nowlin, J.B., & Siegler, I.C. (1985). Normal aging III. Durham, NC: Duke University Press. Pedersen, N.L., & Reynolds, C.A. (1998). Stability and change in adult personality: Genetic and environmental components. European Journal of Personality, 12, 365±386. Plomin, R., & Daniels, D. (1987). Why are two children in the same family so different from each other? The Behavioral and Brain Sciences, 10, 1±16. Poon, L.W., Sweeney, A.L., Clayton, G.M., & Merriam, S.B. (1992). The Georgia Centenarian study. International Journal of Aging and Human Development, 34, 1±17. Rabbitt, P. (1993). Does it all go together when it goes? Quarterly Journal of Experimental Psychology: Human Experimental Psychology, 46A, 385±434. Reinert, G. (1970). Comparative factor analytic studies of intelligence through the human life-span. In L.R. Goulet & P.B. Baltes (Eds.), Life-span developmental psychology: Research and theory (pp. 468±485). New York: Academic Press. Riegel, K.F. (1977). History of psychological gerontology. In J.E. Birren & K.W.

326

Schaie

Schaie (Eds.), Handbook of the psychology of aging (pp. 70±102). New York: Van Nostrand Reinhold. Riley, M.W., Foner, A., & Riley, J.W., Jr. (1999). The aging and society paradigm. In V.L. Bengtson & K.W. Schaie (Eds.), Handbook of theories of aging (pp. 327± 343). New York: Springer. Rogosa, D., Brandt, D., & Zimowsky, M. (1982). A growth curve approach to the measurement of change. Psychological Bulletin, 92, 726±748. Rogosa, D., & Willett, J.B. (1985). Understanding correlates of change by modeling individual differences in growth. Psychometrika, 50, 203±228. Rott, C. (1993). Three components of intellectual development in old age: Results from the Bonn Longitudinal Study on Aging. Zeitschrift fuÈr Gerontologie, 26, 184±190. Rowe, J.W., & Kahn, R.L. (1987). Human aging: Usual and successful. Science, 237, 143±149. Rudinger, G., & Rietz, W. (2001). Structural equation modeling in longitudinal research on aging. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (5th ed.). San Diego, CA: Academic Press. Ryder, N.B. (1965). The cohort as a concept in the study of social change. American Sociological Review, 30, 843±861. Salthouse, T. (1999). Theories of cognition. In V.L. Bengtson & K.W. Schaie (Eds.), Handbook of theories of aging (pp. 196±208). New York: Springer. Schaie, K.W. (1958). Rigidity±¯exibility and intelligence: A cross-sectional study of the adult life-span from 20 to 70. Psychological Monographs, 72 (462, Whole No. 9). Schaie, K.W. (1965). A general model for the study of developmental problems. Psychological Bulletin, 64, 92±107. Schaie, K.W. (1977). Quasi-experimental designs in the psychology of aging. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (pp. 39±58). New York: Van Nostrand Reinhold. Schaie, K.W. (1983). The Seattle Longitudinal Study: A twenty-one year exploration of psychometric intelligence in adulthood. In K.W. Schaie (Ed.), Longitudinal studies of adult psychological development (pp. 64±135). New York: Guilford Press. Schaie, K.W. (1988). The impact of research methodology on theory-building in the developmental sciences. In J.E. Birren & V.L. Bengtson (Eds.), Emergent theories of aging: Psychological and social perspectives on time, self and society (pp. 41± 58). New York: Springer. Schaie, K.W. (1989). The hazards of cognitive aging. Gerontologist, 29, 484±493. Schaie, K.W. (1992). The impact of methodological changes in gerontology. International Journal of Aging and Human Development, 35, 19±29. Schaie, K.W. (1996a). Generational differences. In J.E. Birren (Ed.), Encyclopedia of gerontology (pp. 567±576). San Diego, CA: Academic Press. Schaie, K.W. (1996b). Intellectual development in adulthood: The Seattle Longitudinal Study. New York: Cambridge University Press. Schaie, K.W. (1996c). The natural history of a longitudinal study. In M.R. Merrens & G.G. Brannigan (Eds.), The developmental psychologists (pp. 232±249). New York: McGraw-Hill. Schaie, K.W., & Caskie, G.I.L. (2001). Methodological issues in aging research. In

16. Impact of longitudinal studies

327

D.M. Teti (Ed.), Handbook of research methods in developmental psychology. Oxford: Blackwell. Schaie, K.W., & Hertzog, C. (1986). Toward a comprehensive model of adult intellectual development: Contributions of the Seattle Longitudinal Study. In R.J. Sternberg (Ed.), Advances in human intelligence, Vol. 3 (pp. 79±118). Hillsdale, NJ: Lwarence Erlbaum Associates Inc. Schaie, K.W., & Hofer, S.M. (2001). Longitudinal studies in research on aging. In J.E. Birren & K.W. Schaie (Eds.), Handbook of the psychology of aging (5th ed., pp. 55±77). San Diego, CA: Academic Press. Schaie, K.W., & Labouvie-Vief, G. (1974). Generational versus ontogenetic components of change in adult cognitive behavior: A fourteen-year cross-sequential study. Developmental Psychology, 10, 305±320. Schaie, K.W., Maitland, S.B., Willis, S.L., & Intrieri, R.L. (1998). Longitudinal invariance of adult psychometric ability factor structures across seven years. Psychology and Aging, 13, 8±20. Schaie, K.W., & O'Hanlon, A.M. (1990). The in¯uence of social-environmental factors in the maintenance of adult intelligence. In R. Schmitz-Schertzer, A. Kruse, & E. Olbrich (Eds.), AgingÐA lifelong process of social interaction (pp. 55±66). Darmstadt, Germany: Steinkopf. Schaie, K.W., Plomin, R., Willis, S.L., Gruber-Baldini, A., & Dutta, R. (1992). Natural cohorts: Family similarity in adult cognition. In T. Sonderegger (Ed.), Psychology and aging: Nebraska Symposium on Motivation, 1991 (pp. 205±243). Lincoln, NE: University of Nebraska Press. Schaie, K.W., & Strother, C.R. (1968). A cross-sequential study of age changes in cognitive behavior. Psychological Bulletin, 70, 671±680. Schaie, K.W., & Willis, S.L. (2002). Adult development and aging (5th ed.). Upper Saddle River, NJ: Prentice Hall. Schaie, K.W., & Zuo, Y.L. (2001). Family environments and adult cognitive functioning. In R.L. Sternberg & E. Grigorenko (Eds.), Context of intellectual development (pp. 337±361). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Schmitz-Scherzer, R., & Thomae, H. (1983). Constancy and change of behavior in old age: Findings from the Bonn Longitudinal Study. In K.W. Schaie (Ed.), Longitudinal studies of adult psychological development (pp. 191±221). New York: Guilford Press. Shanan, J. (1993). The Jerusalem Longitudinal Study on mid-adulthood and aging. Zeitschrift fuÈr Gerontologie, 26, 251±255. Singer, J.D., & Willett, J.B. (1991). Modeling the days of our lives: Using survival analysis when designing and analyzing longitudinal studies of duration and the time of events. Psychological Bulletin, 110, 268±290. SoÈrbom, D., & JoÈreskog, K.G. (1978). Con®rmatory factor analysis with model modi®cation. Chicago: International Educational Services. Svensson, T., Dehlin, O., Hagberg, B., & Samuelsson, G. (1993). The Lund 80+ study: Some general ®ndings. In J.J.F. Schroots (Ed.), Aging, health and competence: The next generation of longitudinal research. Amsterdam: Elsevier. Terman, L.M. (1916). The measurement of intelligence. Boston: Houghton. Thorndike, E.L. (1924). The in¯uence of chance imperfections of measures upon the relation of initial scores to gains or loss. Journal of Experimental Psychology, 7, 225±232.

328

Schaie

Thurstone, L.L. (1947). Multiple factor analysis. Chicago: University of Chicago Press. Tucker, L.R. (1958). Determination of parameters of a functional relation by factor analysis. Psychometrika, 23, 19±23. Wechsler, D. (1939). The measurement of adult intelligence. Baltimore, MD: Williams & Wilkins. Welford, A.T. (1951). Skill and age. Oxford: Oxford University Press. Willett, J.B. (1989). Some results on reliability for the longitudinal measurement of change: Implications for the design of studies of individual growth. Educational and Psychological Measurement, 49, 587±601. Willett, J.B., & Sayer, A.G. (1994). Using covariance structure analysis to detect correlates and predictors of individual change over time. Psychological Bulletin, 116, 363±381. Willis, S.L. (1989) Cohort differences in cognitive aging: A sample case. In K.W. Schaie & C. Schooler (Eds.), Social structure and aging: Psychological processes (pp. 94±112). Hillsdale, NJ: Lawrence Erlbaum Associates Inc. Willis, S.L., & Nesselroade, C.S. (1990). Long-term effects of ¯uid ability training in old-old age. Developmental Psychology, 26, 905±910. Willis, S.L., & Reid, J. (Eds.). (1999). Life in the middle. San Diego, CA: Academic Press. Willis, S.L., & Schaie, K.W. (1994). Cognitive training in the normal elderly. In F. Forette, Y. Christen, & F. Boller (Eds.), Plasticite ceÂreÂbrale et stimulation cognitive [Cerebral plasticity and cognitive stimulation] (pp. 91±113). Paris: Fondation Nationale de GeÂrontologie. Wohlwill, J. (1973). The study of behavioral development. New York: Academic Press. Yerkes, R.M. (Ed.). (1921). Psychological examining in the Unites States Army. Memoirs of the National Academy of Sciences, 15, 1±890.

17 Growing points in developmental science: A summing up Willard W. Hartup

University of Minnesota, Minneapolis, USA

BEGINNINGS Developmental science began to achieve coherence during the last decade of the ninteenth century and the ®rst decade of the twentieth. The empirical foundations of this ®eld were still thin, and remained so for many years afterward. Nevertheless, developmental issues began to be written about more frequently by behavioural scientists, theoretical foundations formulated, journals and scienti®c societies begunÐall for the purpose of gaining a better understanding of developmental changes in human behaviour. Everybody knows the names of the scientists responsible for the outpouring of in¯uential works that appeared between 1890 and 1910: James Mark Baldwin (1895, 1897), Alfred Binet (1905), Charles Cooley (1902), Sigmund Freud (1900/1953, 1905/1953), John Dewey (1899), Emile Durkheim (1897/1951), Wilhem Wundt (1910±1920), G. Stanley Hall (1904); and a bit earlier, Francis Galton (1883), and Wilhelm Preyer (1882, 1888±1889), and a bit later, William Stern (1911), Arnold Gesell (1928, laboratory established 1911), and John B. Watson (1914). Others may belong on this list, and no claim should be made that everything important in developmental science was accomplished during these decades. Jean Piaget, to name one individual, did not begin publishing until the 1920s and no one has had more in¯uence on the ®eld during the last 100 years than he. Nevertheless, the work conducted at the turn of the century was propaedeutic; the theoretical contributions of this time were so sturdy that they served as major foundations for developmental thinking during most of the ensuing century. Revolutions in scienti®c thinking, however, did not produce modern developmental science all by themselves. By the turn of the last century, physicians, social workers, and teachers were all struggling to improve the lives of children and families but discovering that science had little to tell them about optimum growth in children and how to nurture it. The empirical science of the times was too thin to provide much assistance to practitioners or parents. By this time, however, advances in the physical sciences had brought about technological revolutions throughout the

330

Hartup

Western world and these were far greater than the small steps being taken in developmental science. Consequently, many individuals (including mothers) began to think that the application of scienti®c methods to human growth and development might provide information that would be useful in raising children, managing adolescents, and providing for older citizens. The record shows that, shortly after the turn of the century, it was women and the child guidance community (social workers, mental health workers, and teachers mostly) who successfully brought pressure to bear on provincial and state governments for funds to study normal human development. Mothers captured the interest of the great foundations in the United States (e.g., the Laura Spelman Rockefeller Memorial and the Payne Fund), inducing them to support scienti®c work in child development and family social science. Mothers even came up with the organisational models needed to support research on better child rearing and improved family life (e.g., centres of basic and applied research modelled on the agricultural experiment station). Social workers brought major pressure to bear in European countries. Beginning, then, in the 1920s in both North America and Europe, research in child development began to pick up speed. Lifespan development was emphasised to a greater extent in Europe (especially Germany) owing to the fact that scientists in the United States had been lobbied to give most of their attention to children. Scienti®c societies were formed, journals founded, and normative studies conducted in Budapest, Berlin, Moscow, Vienna, and Geneva; New Haven, Detroit, Iowa City, Berkeley, and Minneapolis. But then came the Great Depression and World War II. Both talent and funding immediately departed from developmental science and moved elsewhereÐto industry; to waging war; to survival. Some investigators managed to continue their work through this period but the enterprise was a lonely one (see Senn, 1975). The end of World War II brought unprecedented economic and educational expansion throughout the countries of Western Europe, North America, and Australia, as well as parts of Asia and South America. This time, the social sciences were ready to take advantage of the resources that economic expansion made available. For example, the Marshall Plan supported cooperative longitudinal studies of behavioural development in a number of European countries. Federal governmentsÐthrough academies of science, medical research organisations, and science foundationsÐbegan to provide funding for the social sciences as well as the biological and physical ones, earlier in some countries than in others. One must not underestimate the importance to developmental science of the rise of clinical psychology and the centrality of psychoanalytic theory to that discipline. During the post-war period, psychoanalytic theory had a good deal of popular and scienti®c credibility. Developmental theorems were part of it. Although psychoanalysis ultimately lost its hold over

17. A summing up

331

clinical psychology in most countries, professional psychology remained strong through the ensuing years and developmental psychology did, too. The economic and scienti®c Zeitgeist was thus optimal for a surge of activity within behavioural developmental science in the last half of the twentieth centuryÐand that surge has lasted until the present. This volume contains essays describing the state of the art within developmental science as the twentieth century turned into the twenty-®rst. These writings inform readers about questions and issues of current interest in behavioural development but also examine these issues in both retrospect and prospect. In some cases, the authors establish linkages between contemporary science and ideas that emerged during the ``golden age'' of theory-building that occurred between 1890 and 1910. In other cases, the authors connect contemporary thinking to even earlier times, for example, to the Classical era or the Enlightenment. And, in still other cases, the historical view in these essays goes back only a few decades but the author nevertheless extends the reader's attention beyond the last year or two. Considered together, the essays contained in this volume provide a contemporary view about certain intellectual dialectics that dominated the last century in thinking about developmental problems, that is, ideas that played off against one another and controversies that sometimes raged. In developmental science, as elsewhere, theoreticians need past explanations to throw new explanations up against. New ideas have little signi®cance without an understanding of old ones. We turn our attention now to a number of the dialectical tensions that surface in this book.

THE BIG QUESTIONS

What changes? Behaviour changes with age. But which changes are signi®cant? The child's use of language, for example, changes within a range extending from vocalisation to the understanding of grammar and narrative text. Early investigators were meticulous in charting age changes in vocabulary, sentence length, and grammar (e.g., McCarthy, 1931) but little interest was shown until recently in the deeper structural coherence of language and in language acquisition. Although these different levels of analysis each have theoretical and utilitarian signi®cance, it took many years to sort out the aspects of language that are most useful in understanding mental development, on the one hand, and social communication and understanding, on the other. Similarly, early investigators were interested in the young child's differentiation among social stimuli and the acquisition of a repertoire of adaptive social behaviours. Early studies focused on sociability, dominance, honesty, anger and aggression, and independence. These constructs did not

332

Hartup

come from developmental theory so much as from commonly held beliefs that these dispositions are important to the growth and development of competent individuals. As it turned out, the range of behaviours examined by investigators during the 1920s and 1930s was immense, and these choices had an enormous in¯uence on the directions that social development research took subsequently. Generally, however, the choice among social variables made between 1920 and 1940 was made on the basis of armchair reasoning rather than demonstrated theoretical signi®cance (Collins, 2002). Construct selection still dominates the ®eld and the search for new and better ones continues: The essays in this volume show that researchers are still trying to settle on the constructs that best describe memory, knowledge of the world, personality, and motor development as well as aggression and gender development (most especially gender). In many instances (e.g., temperament), construct selection is tentative and likely to remain so. In others (e.g., agency), it is theory-driven and traditions are relatively well established. Normative change or individual differences? The behavioural changes encompassed by developmental science include both normative changes and individual differences. Normative changes, in which emphasis is given to age-related changes in groups of individuals rather than variations within them, commanded much attention through the twentieth century, especially among cognitive researchers. Variations among individuals also command large amounts of attention, especially from investigators in social and personality development as well as psychopathology. Given that many aspects of psychopathology represent branchings from normal development rather than orthogonal adaptations, investigators have spent much time examining both genetic and environmental sources of variation across individuals. Behaviour genetics is one way in which developmental science has been concerned with individual differences; studies of early experience and social relationships are others. Both normative changes and individual differences are dealt with extensively in this book. From the specific to the general or from the general to the specific? Most models of developmental change include explanations of how these changes are instigated (e.g., equilibration, observational learning). Associationist theories, for example, have been utilised to account for the development of higher-order competencies in cognition and language, representing these competencies as build-ups of associations or habits. Developmental trajectories described with this model thus usually are progressions from

17. A summing up

333

simple stimulus±response sequences to more integrated and ``holistic'' structures. Beginning as early as a century ago, developmental change was described by other investigators in terms of differentiation. Most famously, Werner (1957) depicted developmental change in terms of the ``orthogenetic principle'': Namely, the notion that development moves from a state of relative globality and lack of differentiation to a state of differentiation through which the capacity for dealing with a vast array of adaptational challenges is achieved. This differentiation, including what is called ``hierarchical integration'' (a state in which mental and affective processes are organised as integrated ``systems''), makes possible higher levels of conceptualisation and thought as the individual grows older. Although few developmentalists draw their hypotheses directly from notions like the orthogenetic principle, much of their work is devoted to describing outlines or models of development resembling it. Continuity or discontinuity? Developmental views expressed earlier in the twentieth century strongly emphasised continuities believed to exist across time in behavioural adaptation. Early experience, both cognitive and social, was believed to establish substrates that constrain later intellectual and social functioning; manifestations of temperament in the early years were believed to forecast personality orientation and social relations in later life; intelligence was believed to be relatively stable across time; and so on. Serious questions, however, began to be raised some 40 years ago concerning the extent to which human development is actually marked by such continuity. And, indeed, hundreds of studies show extremely modest correlations between traits measured at one time and the same ones measured at others. Many developmentalists then began (circa 1970) to argue that the most important developmental continuities are not revealed in direct carry-over from time to time in behavioural traits or cognitive skills but, rather, in the organisation and integration of cognitive and social adaptations. According to these views, infant intelligence, as measured by the Bayley Infant Scales, should not be expected to correlate highly with Stanford-Binet IQ scores at age 5 because early experience and orientation are unlikely to have direct effects on later outcome. Similarly, social deprivation is now regarded as having few long-term effects that are not dependent in some way on the organisation of intervening life experience. Instead, current views stress that early social relationships are likely to establish ``patterns'' of adaptation that dispose the individual towards success or failure in a variety of other relationships and situations; these are not necessarily revealed by phenotypic continuities in any single aspect of functioning. Newer views also make room for the possibility that continuities in human development vary according to behavioural domain and contextual circumstance.

334

Hartup

Common pathways or many roads to Rome? When earlier developmental psychologists wanted to develop mini-theories of dependency, achievement, or aggression, relevant variables (e.g., negative reinforcement in parent±child interaction and exposure to socially unskilled antisocial peers in the case of aggression) were entered into predictive equations and developmental progression was modelled upon them (see Patterson, 1982). Model-building in development and psychopathology as well as cognitive and language development has been traditionally concerned more with modal pathways than individually speci®c ones, that is, with ``nomothetic'' images rather than ``idiographic'' ones (Allport, 1937). Speci®cally, the objective for the developmental scientist has been to account for more and more of the outcome variance within a single developmental model rather than to posit that development may follow different pathways in different individuals. Single-model efforts, however, have never accounted for much more than 40% of the variance in outcomes such as achievement or aggression; the unexplained variation remaining with most developmental models is considerable. Quite different notions about developmental pathways have taken hold in developmental science recently, notions borrowed from developmental biology. ``Equi®nality'', for example, means that: (a) individuals who have different early experiences may reach the same adaptational endpoint depending on intervening circumstances, and (b) individuals with the same initial status may reach identical endpoints by different routes or pathways. The argument is, indeed, that there are many roads to Rome. Investigators are thus increasingly challenged to look for multiple developmental pathways when pursuing the origins of delinquency, for example, school failure, or depression, as well as seeking to describe the antecedents of success in romantic relationships, for example, or outstanding achievement in science. Developmental modelling in almost all areas has become an exercise in seeking alternative trajectories (and the conditions shaping them) rather than modal ones. Strong hints of these ``systems'' views pervade the essays in this book and represent one direction in which theory-building in the developmental sciences is likely to move over the next several decades. Child development or lifespan development? Some tension has existed within developmental science concerning the ageperiods that deserve greatest attention. Owing mainly to the social forces that supported the beginning of developmental science in North America (which were child-oriented), primary attention was given in that region to normative and differential change occurring during childhood and adolescence. Some of this attention has ideological foundations, too: Many investigators believe strongly that development during the early years has pervasive and long-lasting consequences. But much of this attention was

17. A summing up

335

generated by activists who believed that social change depends on acquiring better information about children and families; these included educators and social workers, child welfare providers, and parents. Although the earliest American developmental scientists (e.g., G.S. Hall and J.M. Baldwin) were, for their own reasons, more interested in children than adults, the science devoted to child development in the United States would not have emerged as it did had mothers not teamed with foundation executives during the 1920s to start things moving in that direction. Social activism on behalf of child development research was more muted in Europe and not as in¯uential in shaping the course of developmental science as in North America. Rather, developmental psychology in Western Europe had antecedents in eighteenth- and nineteenth-century writing about lifespan developmental psychology (Baltes, Lindenberger, & Staudinger, 1998). And psychoanalysis, emerging in the late ninteenth century, was a lifespan theory, not a theory of child development. More recently, developmental psychology traces to William Stern's (1911) ``differential'' psychology. Since differential psychology was not age-speci®c in its tenets, the developmentalists within this sphere of in¯uence (e.g., Charlotte BuÈhler) tended to be lifespan in orientation rather than child-oriented. Much later, in the 1960s, a new generation of investigators working within the lifespan tradition began to raise questions about the wisdom of de®ning developmental science mainly in terms of childhood and adolescence. Lifespan theorists take two observations as their starting points: (a) human development does not cease at the end of the second decade; and (b) understanding human adaptation requires developmental perspectives that extend beyond childhood and adolescence. Attention certainly must be paid to early growth and development simply because it comes prior to development in other age periods. But, although speci®c age-periods are legitimate domains of investigation, it is argued, the adaptational coherence of human development can best be understood in lifespan terms that include attention to both growth and decline in behavioural function. Although there are exceptions (Baltes et al., 1998), most investigators still follow developmental agendas within age periods, e.g., infancy, childhood, adolescence, or old age. This parochialism may re¯ect tradition and inertia. It is also the case, however, that one needs a tremendous knowledge about the organism one wishes to study before being able to contribute substantially to new experimentation. Acquiring the necessary experience with that organism, the folk wisdom about it, the contents of the scienti®c literature, and the ``know-how'' for conducting work with babies, adolescents, newly married couples, and the elderly is a formidable task in each case. Age-graded developmental science is simply easier to do than lifespan research. That said, no intrinsic con¯icts exist between the aims of child development research and lifespan developmental science: both contribute to better understanding of developmental continuity and change; both deal with the

336

Hartup

mechanisms responsible for developmental change; both are concerned with normative change as well as individual differences; both deal with the relation between events occurring early and later in individual lives. Child development and lifespan developmental science have more in common than one might think.

What are the determinants of behaviour change? The ultimate goal for many developmental scientists has been to explain the conditions responsible for normative development. Although everyone recognises the necessity of conducting both descriptive and explanatory studies, current investigators, as well as their predecessors, can be divided into two groups: (a) those whose efforts are devoted to describing the behaviour that changes, and (b) those who devote themselves to explaining how and why these changes occur. As it turns out, explanation in behavioural science is extraordinarily dif®cult. Not only is most developmental change multidetermined, but causation is buried deeply within the genome and the mind. Moreover, the time separating ``cause'' from ``effect'' in human development is frequently decades, not moments, and change within one behavioural domain is entangled with changes in others. Becoming a developmental scientist, especially one who is interested in the determinants of behavioural development, thus requires brashness and intrepidity as well as whatever talent one can muster. Classifying the great explanatory principles that have emerged within developmental science is dif®cult. Although one frequently reads about superordinate categories such as ``organismic'' principles as opposed to ``mechanistic'' ones, boundaries are fuzzy. Such classi®cations are not attempted here. Rather, in these pages, several ``tensions'' or ``dialectics'' are discussed that were pervasive in theorising about developmental process during the twentieth century: (a) the continuing tension between nativist and empiricist developmental explanations; (b) the tension between general theories and those that are domain-speci®c; (c) how to account for socialisation; (d) the tension existing between views of human development as ``person-driven'' as opposed to ``persons in context''; and (e) the necessity for ``grand theories'' in formulating integrated explanations for developmental change as opposed to beliefs that ``mini-theories'' or ``theories of action'' will serve better. Nativism or empiricism? Understanding the nature and acquisition of knowledge has preoccupied philosophers and psychologists for centuries. Notions that human beings begin life with ``innate ideas'' and ``inborn receptivities'' appeared ®rst in philosophical and religious texts: Common beliefs in classical and medieval

17. A summing up

337

times were that nativistically established capacities were necessary to the formation of adaptations in earliest infancy, that some of this innate knowledge remains basic to adaptation throughout the lifespan, and that new knowledge is built on these innate structures. Extreme empiricist views, in contrast, are based on the assumption that all knowledge and many capacities (including perceptual and motor ones) are abstracted from the individual's experience in the world. The newborn mind is a tabula rasa, this argument goes, or a ``buzzing, booming confusion'' (James, 1912). Sometimes, the nativism±empiricism dialectic has been framed in terms of the ``heredity versus environment'' debate. Careful reading shows, however, that this dialogue is better described as occurring between individuals, on the one hand, who believe that intrinsic structures are necessary to the acquisition of knowledge in many domains and those, on the other hand, who believe that the main objective is to describe the world in which the individual lives and its bearing on these intrinsic structures. The debate is far more lively and complex than arguments concerning whether heredity or environment is the more important source of developmental differences among individuals. Even so, one frequently encounters views that the ``debate'' between nativism and empiricism in developmental science is either: (a) a pseudodebate or (b) a controversy that has been resolved in favour of empiricism because nativism is either ``incoherent, empty, unparsimonious, in¯exible, false, or socially and politically dangerous'' (Spelke & Newport, 1998). Every developmental change ever documented is the product of gene± environment interactions, so the argument goes. Current views, however, are more encompassing and varied than the heredity±environment question. Some concern the hypothesis that there is an innate substrate for knowledge acquisition (e.g., some innate capacity for learning language or for learning itself ) that is shared by all human beings but, at the same time, makes differences among them possible. Other notions concern the nature of the processes through which innate structures (whether aspects of perception or capacities for constructing interpersonal relationships) are deployed in behavioural development. According to these views, the task is to specify both nativist and empiricist contributions to developmental acquisitions as well as their interaction. What is socialisation? Usually described from an empiricist point of view, socialisation is widely regarded as the individual's internalisation of social constructs and representations of the world; his or her constructions of expectations, attitudes, and values about others and about social dilemmas; and the acquisition of communications skills and close relationships. Among psychologists, socialisation was viewed through most of the twentieth century as John Watson described itÐa process of ``social moulding''

338

Hartup

through which caretakers ``teach'' the naive young child the adaptational skills needed for meeting the challenges of living in the world. Children, in other words, were regarded as creatures who are socialised through interpersonal processes that are essentially one-way in terms of the direction of behavioural in¯uence: Parents teach and children learn, not the reverse. Although there were many indications in earlier writings that children are ``socialising creatures'' as well as ``socialised'' (Baldwin, 1895; Piaget, 1926), a consensus concerning a two-way or bilateral view of childhood socialisation was reached only recently (Bell, 1968). The notion that children are active in their own socialisation, and that adaptation is the outcome of interaction between personal characteristics (e.g., temperament and linguistic structures) and external in¯uences, then swept the ®eld. Socialisation was re-conceptualised as a dyadic process, the importance of peer relations in working out developmental adaptations was recognised in addition to the importance of parent±child relations, and an integration between cognitive and social themes in development began to occur. These convergences are the centre of attention in socialisation research today, and re¯ections about them occur frequently in this book.

Stages anyone? Another ``tension'' or dialectic running through the history of developmental thought involves the notion that behavioural development (whether in cognitive functioning or social/emotional skills) should be conceived as a progression of ``stages'', i.e., times in the individual's life when a particular mode of perceiving, thinking, or feeling dominates adaptation in many different contexts. Sometimes, stage theories account for developmental transformations in terms of endogenous forces, e.g., the development of the brain or the psychoneuroendocrinological system, and sometimes in terms of exogenous ones, for example, when con¯ict occurs between one's own views of the world and the views of others. Psychoanalytic writers thus sought to account for major transformations in the child's emotional and social life in terms of psychosexual development that was driven by biological mechanisms. In contrast, cognitive theories sought to account for major shifts in perception and reasoning in terms of socially induced cognitive con¯ict (equilibration). The evidence increasingly suggests that development seldom occurs broadly in stage-like progressions even though there may be some regularity from individual to individual in cognitive, emotional, and social changes. Variations across individuals in the order and timing of developmental change as well as variation within individuals related to task or situation are too great to support traditional stage notions. Although the essays in this volume do not evaluate ``stage models'' directly, the authors offer readers a chance to evaluate such developmental models.

17. A summing up

339

Domain generality or specificity? With the increasing recognition that the course of development varies across and within individuals, alternatives to stage models of developmental change were needed. Piaget and others struggled to modify their theories to recognise the existence of ``decalage''. Later, however, large numbers of investigators proposed that the notion of generalised developmental advance should be abandoned. And along with the abandonment of stage theories came increasing adherence to the notion that psychological processes are not organised to apply to knowledge acquisition universally, but within circumscribed ``domains'' of knowledge or expertise. As various critics have pointed out (cf. Fischer & Bidell, 1998), domain-speci®c theories of knowledge acquisition get around the dif®culties of stage theories in accounting for developmental variations simply by arguing that stages do not exist. As it turns out, however, domain-speci®c theories have dif®culties of their own. First, the problem of variations in developmental change is only super®cially solved by these notions since patterns of variation are not explained. Second, specifying domain boundaries is a major problem. How ``local'' must a domain be in order to reveal developmental process clearly? Must domains always be regarded as speci®c to content and cultural circumstance? Are any domains universal so as to substantiate what Wundt (1910) called ``folk psychology'' or what Heider (1958) called ``naive psychology''? Wrestling with these issues has proved daunting, and investigators still argue about the best ways to explain developmental variations, on the one hand, and developmental generalities, on the other. Once again, re¯ections on these issues can be found in several chapters in this book.

Does context make a difference? Common sense establishes that human beings and the world that surrounds them are separate entities. Everyone understands that events in the world instigate changes in mental and behavioural activity in individuals and, consequently, person and environment must be regarded as separate. Given this dualism, traditional cognitive and information-processing theorists attempted to understand qualities of the mind and their development as universals, that is, as common structures that characterise all human beings and that transcend differences in lifestyle and culture. These personoriented theories dominated research in cognitive development and language through the century, until an ecological revolution swept through research dealing with perceptual development that was instigated by Eleanor Gibson (1969) and James Gibson (1979). In social and personality development, environment was characterised as ``impinging'' or ``moulding'' the individual's development and most

340

Hartup

investigators concerned themselves with the mother's in¯uences over the child. Fathers and other family members were largely dismissed as insigni®cant contextual elements, and family structure was given scant attention except by certain sociologists. Cultural differences in child rearing drew the attention of some anthropologists, but cultural views of human development development were rareÐeven among students of social development. The fact that contexts themselves change over time received virtually no attention until recently (Elder, Modell, & Parke, 1993). Scientists who helped change this ``individualistic'' view of development toward a more ``contextual'' view include early writers like James Mark Baldwin, and major ®gures whose work dates from the 1930s including Kurt Lewin (Lewin, 1935), L. S. Vygotsky (1939), and Roger Barker (Barker, 1968). In these works, the ®rst argument is that as much attention must be paid to parsing the environment as to the manner in which the mind is structured. This argument has led to extended efforts to map the environment in which psychological growth occurs, although agreement as to how to map and assess it has still not been achieved. Complex models of the ``individual-in-environment'' have been proposed that include the DNA as well as social relationships, groups, and culture (see, for example, Bronfenbrenner & Crouter, 1983; Hinde, 1997). Initially, environmental analysis was undertaken because the environment was thought to support or shape development, including both mental and social development. Although not a particularly revolutionary notion as related to social development, these ideas were more controversial in relation to cognitive development: Is it really the case, for example, that mental structures originate as ``gestures'' in social interaction (Mead, 1934) or that cognitive function appears ®rst between the child and other persons before appearing psychologically in the child (Vygotsky, 1939)? Most cognitive developmentalists seemed to be concerned primarily with mapping out the fundamentals of mind, memory, and language, believing that context can be added later (including both setting or situation and the historical context). As the contents of this book make plain, though, that assumption has turned out to be naive. Currently, the most advanced ideas being pursued concerning contextualism and development involve discarding the notion that person and environment should be conceived dualistically. Gone are notions that persons cause environmental change and vice versa. Instead, the modern view treats the person and the environment as integrated or fused; neither is an entity that can be described sensibly without reference to the other, and development is the story of change over time in this integrated system. One must continue to examine development as it involves structures of varying levels of complexity including psychobiological processes, behaviour, dyadic relationships, groups, and cultures. One must consider systems in development, however, rather than development as a property of either individuals or contexts.

17. A summing up

341

Grand theories or mini-theories? Through three-quarters of the twentieth century, a series of comprehensive theories dominated thinking about behavioural development. These socalled ``grand theories'' were encompassing, directed at relevant domains ranging from socio-emotional development to cognitive development, based on interesting assumptions, and sometimes (not always) produced hypotheses that were empirically testable. Cognitive and maturational theories developed by James Mark Baldwin and Arnold Gesell held sway early in the century; psychoanalytic theory, as developed by Sigmund Freud and various associates, grew steadily in popularity until after World War II. Cognitive-developmental theory, as constructed by Jean Piaget, was introduced in the mid-1920s, attracted a number of early adherents, and then exploded in full force during the 1960s. The so-called learning theories were applied to human development, beginning with the work of John Watson in the 1920s, and were modi®ed later by John Dollard, Neal Miller, and their associates in the immediate pre- and post-World War II years (Dollard, Miller, Doob, Mowrer, & Sears, 1939), and after that by Albert Bandura and others (Bandura & Walters, 1963). Ecological and ®eld theories, formulated by Lewin (1935), commanded considerable attention ranging from the 1930s to the present. The record thus shows that more than one theory has usually been taken seriously at any one time during the century; theoretic hegemony did not exist. Some writers contend that behaviourism was the dominant developmental theory between 1940 and 1965, even though other theories were also vying for attention during those years. Grand theories now in¯uence developmental science to a much lesser degree than in earlier decades. As they grew to prominence, most of these theories were found to have limitations. Cognitive-developmental theory depended too rigidly on its ``stages'', and failed convincingly to explain why stage-to-stage change occurs. Learning theory was regarded by many critics as extremely limited in accounting for language acquisition and other aspects of cognitive development. That sexual vicissitudes undergird socioemotional developmentÐone of the best-known ideas in psychoanalytic theoryÐnow seems greatly overstated. Although revisionists worked long and hard to modify the more questionable aspects of these theories and to extend them to a broader range of phenomena (see, for example, Kohlberg's 1969 extension of cognitive-developmental theory into the range of moral and social thinking), these efforts have not been entirely successful. Developmental science has gradually moved into an era in which theory-construction is directed at more circumscribed issues. Nevertheless, vestiges of the grand theories remain. One cannot thoroughly appreciate today's interest in children's ``theories of mind'' without understanding egocentric thinking and its modi®cations as enunciated by Jean Piaget. Similarly, one cannot understand contemporary theories of attachment formation without understanding what Sigmund

342

Hartup

Freud wrote about object relations and other aspects of relationships and emotion regulation. One does not know whether anyone will succeed in constructing another grand theory, especially soon. One suspects we will remain involved with an eclectic series of mini-theories for the foreseeable future. Nevertheless, the continuing viability of certain core ideas from earlier theories must be acknowledged. To dismiss the contributions of psychoanalytic theory, maturational theory, or learning theory as contributions to contemporary developmental science is naive.

CONCLUSION The essays included in this volume describe the ideas of past scientists in order to provide current investigators with a sense of identity and links to both past and future. No one embarks on a scienti®c career or sustains one for any length of time by working on issues and ideas de novo. One must know the ``roads taken and the roads not taken'' by our predecessors (Cahan, 1997). In this book, we sought to illuminate current work by showing its antecedents as well as similarities between past and present. The authors review some of the intellectual debates and tensions that have occurred over the last century in thinking about developmental issues. The ideas that have worked and not worked, and ideas that have been played off against one another (many still regarded in opposition), receive attention. These dialectics must be appreciated by contemporary scientists who wish to tackle the big, explanatory questions in developmental science. Seldom can one give science a nudge without a good working knowledge of the tensions that surround such growing points.

REFERENCES Allport, G.W. (1937). Personality: A psychological interpretation. New York: Holt. Baldwin, J.M. (1895). Mental development in the child and the race: Methods and processes. New York: Macmillan. Baldwin, J.M. (1897). Social and ethical interpretations in mental development: A study in social psychology. New York: Macmillan. Baltes, P.B., Lindenberger, U., & Staudinger, U.M. (1998). Life-span theory in developmental psychology. In W. Damon (Series Ed.), R.M. Lerner (Vol. Ed.), Handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed., pp. 1029±1144). New York: John Wiley. Bandura, A., & Walters, R.H. (1963). Social learning and personality personality development. New York: Holt, Rinehart & Winston. Barker, R.G. (1968). Ecological psychology: Concept and methods for studying the environment of human behavior. Stanford, CA: Stanford University Press.

17. A summing up

343

Bell, R.Q. (1968). A reinterpretation of the direction of effects in studies of socialization. Psychological Review, 75, 81±95. Binet, A., & Simon, T. (1905). Methods nouvelles pour le diagnostic du niveau Intellectuel des anormaux. L'Anne Psychologique, 11, 191±244. Bronfenbrenner, U., & Crouter, A. (1983). The evolution of environmental models in developmental research. In P.H. Mussen (Series Ed.), W. Kessen (Vol. Ed.), Handbook of child psychology: Vol. 1. History, theory and methods (4th ed., pp. 357±414). New York: John Wiley. Cahan, E.D. (1997). On the uses of history for developmental psychologists or on the social necessity of history. Newsletter of the Society for Research in Child Development, 40, 2±8. Collins, W.A. (2002). Historical perspectives on contemporary research in social development. In P.K. Smith & C.H. Hart (Eds.), Blackwell 's handbook of social development (pp. 3±23). Oxford: Blackwell. Cooley, C.H. (1902). Human nature and the social order. New York: Free Press. Dewey, J. (1899). The school and society. Chicago: University of Chicago Press. Dollard, J., Miller, N.E., Doob, L.W., Mowrer, O.H., & Sears, R.R. (1939). Frustration and aggression. New Haven, CT: Yale University Press. Durkheim, E. (1897/1951). Suicide. New York: The Free Press. (Original work published 1897.) Elder, G.H., Modell, J., & Parke, R.D. (Eds.). (1993). Children in time and place. Cambridge: Cambridge University Press. Fischer, K.W., & Bidell, T.R. (1998). Dynamic development of psychological structures in action and thought. In W. Damon (Series Ed.), R.M. Lerner (Vol. Ed.), The handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed., pp. 467±562). New York: John Wiley. Freud, S. (1901/1953). The interpretation of dreams. London: Hogarth. (Original work published 1900.) Freud, S. (1905/1953). Three essays on the theory of sexuality. London: Hogarth. (Original work published 1905.) Galton, F. (1883). Inquiries into human faculty and its development. London: Macmillan. Gesell, A. (1928). Infancy and human growth. New York: Macmillan. Gibson, E.J. (1969). Principles of perceptual learning and development. Englewood Cliffs, NJ: Prentice Hall. Gibson, J.J. (1979). The ecological approach to visual perception. Boston: Houghton Mif¯in. Hall, G.S. (1904). Adolescence: Its psychology and its relations to physiology, anthropology, sociology, sex, crime, religion, and education (2 vols.). New York: Appleton. Heider, F. (1958). The psychology of interpersonal relations. New York: John Wiley. Hinde, R.A. (1997). Relationships. Hove, UK: Psychology Press. James, W. (1912). Essays in radical empiricism. New York: Longmans, Green. Kohlberg, L. (1969). Stage and sequence: The cognitive-developmental approach to socialization. In D.A. Goslin (Ed.), Handbook of socialization theory and research (pp. 347±480). Chicago: Rand McNally. Lewin, K. (1935). A dynamic theory of personality. New York: McGraw-Hill. McCarthy, D. (1931). Language development. In C. Murchison (Ed.), A handbook of child psychology (pp. 278±315). Worcester, MA: Clark University Press.

344

Hartup

Mead, G.H. (1934). Mind, self and society. Chicago: University of Chicago Press. Patterson, G.R. (1982). Coercive family process. Eugene, OR: Castalia Press. Piaget, J. (1926). The language and thought of the child. New York: Harcourt Brace. (Original work published 1923.) Preyer, W. (1882/1888±1889). The mind of the child (2 vols.). New York: Appleton. Senn, M.J.E. (1975). Insights on the child development movement in the United States. Monographs of the Society for Research in Child Development, 40 (Serial No. 161). Spelke, E.S., & Newport, E.L. (1998). Nativism, empiricism, and the development of knowledge. In W. Damon (Series Ed.), R.M. Lerner (Vol. Ed.), The handbook of child psychology: Vol. 1. Theoretical models of human development (5th ed., pp. 275±340). New York: John Wiley. Stern, W. (1911). Die differentielle Psychologie in ihren methodeischen Grundlagen. Leipzig, Germany: Barth. Vygotsky, L.S. (1939). Thought and speech. Psychiatry, 2, 29±54. Watson, J.B. (1914). Behavior: An introduction to comparative psychology. New York: Henry Holt. Werner, H. (1957). The concept of development from a comparative and organismic point of view. In D.B. Harris (Ed.), The concept of development: An issue in the study of human behavior (pp. 125±148). Minneapolis, MN: University of Minnesota Press. Wundt, W. (1910±1920). Volkerpsychologie (10 vols.). Leipzig, Germany: Engelmann.

Author index

Abbot-Shim, M. 92 Abu-Lughod, L. 280 Achenbach, T.M. 242 Adler, B. 65 Adolph, K.E. 2, 4, 13, 14, 15, 18 Agnetta, B. 115 Aguirre, D.M. 69 Ahadi, S.A. 179, 184 Ahern, F. 49 Ainsworth, M. 30 Akkari, A. 279 Aksan, N. 188 The Alan Guttmacher Institute 294 Alcock, N. 68 Alessandri, S.M. 244 Aligne, A. 67 Allard, T.T. 11 Allen, A. 178 Allen, J.P. 294 Allison, P.D. 310 Allport, G.W. 178, 334 Alpert, J. 65 Alpert, R. 203 Alsaker, F.D. 295 American Psychiatric Association 245±246, 246 Ames, E.W. 32 Anderson, D.I. 16 Andersson, B. 86, 88 Andreasen, N.C. 59 Andres, D. 318 Andrews, B. 25 Angleitner, A. 48, 181, 190 Angulo-Kinzler, R.M. 14 Anisman, H. 216

Anooshian, L.J. 157 Anthony, J.C. 298 Apostolaris, N.H. 213 Applegate, B. 246 Arbuckle, T.Y. 318 Archer, J. 245 Ardila, A. 277 Arnett, J. 92 Arnett, J.J. 291, 292 Aronson, E. 15 Arsenio, W.F. 173, 289 Asakawa, K. 290 Asbury, K. 52 Asendorpf, J.B. 187, 190, 195 Ashmead, D.H. 15 Ashworth, A. 73 Assuncao, A. 73 Astington, J.W. 105, 106, 167, 168, 171 Atran, S. 130, 134 Auinger, P. 67 Ausley, J.A. 157 Austin-La France, R. 74 Avia, M.D. 181 Avis, J. 115, 272 Avolio, A.M. 14 Ayers, T. 294 Bachman, J.G. 291, 292, 297 Baddeley, A. 152 Bahrick, L.E. 150 Bai, D.L. 15 Baillargeon, R. 125, 127, 131, 136 Bakan, D. 189 Baksh, M. 75 Baldwin, D.A. 111

Baldwin, J.M. 329, 335, 338, 340, 341 Bales, R.F. 213 Baltes, M.M. 320 Baltes, P.B. 192, 226, 226±227, 227, 232, 235, 309, 310, 315, 322, 335 Bandura, A. 206, 207, 208, 223, 243, 244, 247, 253, 341 Banerjee, M. 112 Banich, M.T. 289 Bank, L. 253, 254 Barbu-Roth, M.A. 16 Barker, D. 66 Barker, R.G. 340 Barkow, J.H. 139 Barnes, H. 86, 87, 255 Barnes, J. 91 Barnes, R. 73 Barnett, W.S. 86, 87 Baron, R.M. 236 Baron-Cohen, S. 106, 109, 115, 117 Barraclough, B.M. 35 Barrett, B.J. 87 Barrett, D. 73 Barrett, K.C. 16 Barry, H. 266 Bartsch, K. 106, 109, 112, 114, 169 Bassily, N. 71 Bates, E.A. 18, 129, 130, 139 Bates, J.E. 27, 179, 183, 184, 191, 193, 196, 244, 252 Batth, R. 289

346

Author index

Bauer, P.J. 151 Baumert, J. 224, 232±233, 233 Bayley, N. 1, 309, 314 Beam, M. 291 Becker, B. 294 Bedi, K. 70 Beeghly-Smith, M. 105 Bekman, S. 86 Bell, J. 57 Bell, R.Q. 26, 338 Beller, E. 86 Belsky, J. 89 Bem, D.J. 178 Bem, S.L. 207 Ben-Artzi, E. 292 Benenson, J.F. 213 Bennett, J. 105 Bennett, R.T. 113 Benson, J.B. 117 Berg, S. 49, 316 Bergenn, V.W. 3 Bergman, L.R. 251 Berkowitz, L. 241 Bernstein, N. 6 Berrueta-Clement, J.R. 87 Berry, J.W. 266, 267, 268, 269, 271, 275, 276, 277, 278, 280 Bertenthal, B.I. 13, 15, 16, 17 Beyth-Marom, R. 289 Bibby, R.W. 300 Bidell, T.R. 339 Bifulco, A. 37 Binet, A. 4, 308, 329 Binet, H. 146, 256 Birch, D. 67 Birch, E. 67 Bird, H.R. 246 Birren, B.A. 307 Birren, J.E. 307, 308 Bishop, D.V.M. 51 Bishry, Z. 71, 73, 75, 76 Bjorklund, D.F. 104, 149, 153, 154, 156, 157, 160, 161, 215 Bjorkqvist, K. 244, 245 Black, J. 74 Black, M. 67, 68, 73 Block, J. 178, 179, 180, 181, 182, 184, 187, 251, 296

Block, J.H. 178, 180, 181, 182, 184, 187 Bloom, F. 74, 289 Blumstein, A. 246 Blurton-Jones, N. 214 Blythe, D.A. 191 Boesch, C. 216 Boesch, E.E. 267 Boesch, H. 216 Boivin, M. 229, 249, 253 Boker, S.M. 17 Bolger, K.E. 40 Borgatta, E.F. 213 Borkenau, P. 48 Borker, R.A. 218 Borutta, M. 13 Boston, M. 30 Bosworth, H.B. 310, 316, 318 Bouchard, T.J., Jr. 49 Boudreau, J.P. 15 Boulerice, B. 245, 249, 250 Bourdon, B. 145 Bous®eld, W.A. 148 Bowlby, J. 30, 40, 214 Bradley, R.H. 54, 92 Brain, P. 244 Brainerd, C.J. 158, 159 Brame, B. 252 BrandstaÈdter, J. 225 Brandt, D. 309 Brandt, J. 67 Bransford, J.D. 104 Braungart, J.M. 54 Bream, L.A. 244 Bredekamp, S. 92 Breen, G. 56 Brendgen, M. 255 Breniere, Y. 13 Bretherton, I. 105 Brewin, C.R. 25 Bridges, K.M.B. 248, 249 Briggs, E.C. 291 Bril, B. 13, 274 Brislin, R.W. 278 Broberg, A. 88 Broidy, L.M. 252 Bronfenbrenner, U. 224, 225, 230, 277, 289, 340 Bronzino, J. 74 Brooks-Gunn, J. 87, 294 Brown, A.L. 104, 124 Brown, B.B. 219, 291

Brown, C.H. 299 Brown, G. 37 Brown, G.W. 37 Brown, J. 136 Brownlee, E.M. 89 Bruck, M. 146, 159, 160 Brumann, C. 280 Bruner, A. 67 Bruner, J. 86, 267 Bruner, J.S. 15 Brunswik, E. 146, 149 Bryant, A.L. 292 Bryant, D.M. 87, 91 Bryk, A.S. 93, 231, 310 Bucholz, K.K. 242 BuÈhler, K. 144, 145, 147 Bukowski, W. 219 Bukowski, W.M. 247 Burchinal, M.R. 86, 91 Burgard, P. 270 Burnside, L.H. 3 Burt, C.L. 241 Burton, R.V. 26 Bushnell, E.W. 15 Buss, A.H. 241 Buss, K.A. 172, 185 Busse, E.W. 315 Bussey, K. 207, 208 Butterworth, G. 15 Bwibo, N. 75, 76 Byler, P. 86, 91 Bynner, J. 291, 295 Cadoret, R.J. 54, 61 Cahan, E.D. 342 Cairns, B.D. 244, 251, 256 Cairns, R.B. 216, 224, 225, 226, 241, 244, 251, 255±256, 256, 297 Caldwell, B.M. 25, 54, 92 Calkins, S.D. 185±186 Call, J. 115 Calzi, F. 70 Camilleri, C. 267, 275 Campbell, F.A. 87 Campbell, J. 88 Campbell, J.D. 26 Campione, J.C. 104 Campos, J.J. 16, 172, 185 Canino, G. 246 Capage, L. 115 Capaldi, D. 254 Cardon, L.R. 49, 56, 57

Author index Carey, S. 124, 126, 129, 131, 134 Carlo, G. 195 Carlson, E.A. 33 Carlson, S.M. 109 Carlson, V. 131 Carmichael, L. 1, 241 Carpendale, J.I. 113, 117 Carpenter, M. 106, 111 Carraher, D. 275 Carroll, J.B. 50 Carrothers, P. 167 Carson, C. 172 Carstensen, L.L. 320 Carter, M.C. 32 Casagrande, C. 249 Casas, J.F. 245 Casasso, R. 70 Case, R. 152 Casella, J. 67 Caskie, G.I.L. 309 Caspi, A. 38, 39, 180, 186, 187, 188, 189, 190, 191, 246 Cassel, J. 76 CastanÄeda-English, P. 291 Castle, J. 251 Castle, P. 15 Catalano, R.P. 255 Cattell, A.K.S. 310 Cattell, R.B. 251, 310 Caul®eld, L. 67, 69 Ceci, S. 275 Ceci, S.J. 146, 156, 159, 160, 289 Cerella, J. 316 Chacon, M. 70, 75, 76 Chamberlain, A.H. 145 Chandler, M.J. 113, 117 Chang, S. 66, 67, 69, 71, 74±75, 75 Chapman, D. 14 Chapman, M. 223, 225, 226, 227 Charman, A. 171 Chase-Lansdale, P.L. 89 Chavez, A. 71, 75 Chen, C. 291 Chen, X. 68 Cherny, S.S. 49 Chi, M.T.H. 124, 135, 139, 156 Chipuer, H.M. 52

Chisholm, K. 32 Chiu, C.Y. 115 Chodorow, N. 203 Chomsky, N. 124 Choquet, M. 249 Chorney, K. 56 Chorney, M.J. 56 Christensen, L. 65, 66, 70 Christiansen, M. 71 Cicchetti, D. 234, 294 Cillessen, A.H.N. 193 Cintra, L. 74 Clark, J.E. 13 Clark, L.A. 182 Clark, M.M. 86 Clark, R. 188 Clarke, A.D.B. 28, 40 Clarke, A.M. 28, 40 Clarke-Carter, D. 169, 171 Clarke-Stewart, K.A. 85 Clarkson, M.G. 15 Clausen, J.A. 309, 314 Claussen, A. 65 Clayton, G.M. 315 Clement, J. 71 Clements, W.A. 114, 117 Clifford, R. 91 Clifford, R.M. 86, 91 Clifton, R.K. 15 Cloninger, C.R. 179, 182 Coatsworth, J.D. 294 Coghill, G.E. 2 Cohen, A.H. 17 Cohen, D.J. 106, 115, 117 Cohen, J. 246 Coie, J.D. 229, 241, 243, 244, 248, 253, 254, 266 Coldham, C. 76 Cole, M. 125, 267 Cole, R. 255 Coleman, P. 207 Collier, J. 216 Collins, L.M. Collins, W.A. 332 Colombo, M. 72 Colvin, M. 67, 69 Condelli, L. 87 Conger, R.D. 54, 61 Connolly, K. 72 Conrad, H.S. 308, 309, 314 Cook, J. 65 Cooley, C.H. 329

347

Coontz, P.D. 247 Corbetta, D. 11, 13, 14 Corbin, S.B. 39 Corder, E.H. 56, 58 Cosmides, L. 139 Costa, F.M. 291, 294 Costa, P.T., Jr. 179, 181, 190 Costigan, C.L. 290, 291 Cowan, N. 153 Cowdry, E.V. 308 Crabbe, J.C. 59 Craig, W. 255 Craig, W.M. 290 Crick, N.R. 245 Cristofalo, V.J. 307 Cronbach, L.J. 309 Cross, D. 102, 167 Crouter, A. 340 Crowell, J. 33 Cryer, D. 91 Csikszentmihaalyi, M. 290 Cueto, S. 69 Culkin, M.L. 86, 91 Cumberland, A. 195 Cummings, E.M. 251 Curtiss, S. 32 Cutting, A. 169, 171 Cutting, A.L. 168, 172 Cynader, M. 250 Dale, P.S. 51 Dalton, T.C 3 Damon, W. 266 D'Andrade, G. 280 Daniels, D. 311 Daniels, J. 56 Darlington, R. 87, 88 Darwin, C. 1, 2 Dasen, P.R. 266, 268, 269, 271, 272, 274, 275, 277, 278, 279, 280 Davidson, G.R. 274, 275 Davies, L. 251 Dawe, H.C. 248, 249 Day, D.M. 244 Dayal, H. 68 de Andraca, I. 65 de la Parra, A. 72 De Leeuw, J. 231 de Raad, B. 179, 183 de Ribaupierre, A. 272 de Salis, W. 26

348

Author index

de Villiers, J.G. 171 de Villiers, P.A. 171 Dean, R. 67 Deary, I.J. 56 Deater-Deckard, K. 51, 54, 204 Deaux, K. 208 DeBaryshe, B.D. 252 Deci, E.L. 224, 225, 226 DeFries, J.C. 51, 56 Dehlin, O. 315 DelVecchio, W.F. 186 Demetriou, H. 169 Dempster, F.N. 152 Dennett, D.C. 105 Dennis, M.J. 117 Dennis, W. 28, 39 Denny, M.A. 13 Desai, S. 89 Desmet, H. 271 Detterman, D.K. 50, 56 Devich-Navarro, M. 297 DeVries, M. 76 Dewey, J. 329 deWit, J. 241, 243, 244, 248 Diamond, A. 243 Diaz-Cintra, S. 74 Dichgans, J. 13 Dickson, N. 246 Digman, J.M. 179 DiModugno, M. 67, 69 Dinwiddie, S.H. 242 DiPietro, J. 67, 69 Dishion, T.J. 247, 253, 255 Dixon, R.A. 315 Doak, C. 65 Dobkin, P.L. 246 Dodge, K.A. 27, 173, 229, 241, 243, 244, 248, 252, 253, 254, 266 Dollard, J. 341 Donelan-McCall, N. 169 Dong, Q. 272, 291 Donovan, J.E. 291 Doob, L.W. 341 Dorea, J. 72 Dorenberger, C.H. 294 Dornbusch, S.M. 291 Douglas, R.N. 154 Duggan, A. 67 Dulcan, M.K. 246 Dumas, J.E. 253

Dunn, J. 52, 136, 167, 168, 169, 170, 171, 172, 173 Dunn, J.F. 53 Dunne, M.P. 242 Durkheim, E. 329 Dutta, R. 307, 317 Dweck, C.S. 115 Eaves, L.J. 51 Ebbinghaus, H. 143, 145 Eccles, J.S. 293, 297 Eckenrode, J. 255 Eckensberger, L.H. 270 Eckert, P. 218 Edelbrock, C. 242 Edelman, G.M. 11, 12 Edelstein, W. 187, 195 Egeland, B. 33 Ehrenberg, M.F. 294 Eichorn, D.H. 309, 314 Eisenberg, N. 195, 289, 290 El-Najar, S. 92 Elder, G. 72 Elder, G.H. 294, 340 Eley, T.C. 51, 242 Elliott, D.S. 246 Elliott, E. 153 Ellis, L. 247 Elman, J.L. 18, 129, 130, 139 Emde, R.N. 172 Endler, N.S. 178 Engle, P. 68, 72, 73, 76 English and Romanian Adoptees (ERA) Study Team 29, 32 Ensminger, M.E. 293 Epstein, A.S. 87 Erickson, M.F. 299 Eron, L.D. 242, 247, 253 Esbensen, B.M. 113 Escobar, C.M. 86 Espinosa, M. 67, 69 Essex, M.J. 188 Esterson, J. 148 Estes, D. 114, 169 EURONET 235, 295 Evans, A.C. 289 Evans, D.E. 179 Evans, G. 72 Eysenck, H.J. 179, 247

Fabes, R.A. 289, 290 Fabricius, W.V. 112, 117 Farran, D.C. 89 Farrington, D.P. 246, 247, 255 Fava, M. 65 Fein, G.G. 85 Fein®eld, K.A. 103 Feingold, A. 190 Felt, B. 69 Ferguson, L.L. 251 Fergusson, D. 252 Fergusson, D.M. 34, 38, 293 Fernald, L. 65, 66, 67, 68 Fernandez, J.W. 280 Fernandez, T. 69 Fernyhough, C. 169, 171 Ferrera, R.A. 104 Feshbach, S. 244 Fetterman, D.M. 298 Field, T. 88 Figueroa, A. 67, 69 Finkel, D. 317 Fischer, M.M.J. 280 Fischer, K.W. 339 Fischhoff, B. 289 Fisher, C.B. 294, 299 Fisher, D.M. 7, 8, 14 Fisher, P.J. 56 Fitzgerald, J.A. 92 Fitzpatrick, P. 17 Flammer, A. 295 Flavell, E.R. 103, 105, 112, 114 Flavell, J.H. 102, 103, 104, 105, 106, 110, 111, 112, 113, 114, 116, 117, 153, 155, 172, 243, 272 Floyd, F.J. 291 Fodor, J.A. 51 Fonagy, P. 171 Foner, A. 307, 320 Fontaine, R. 13 Forman, D.R. 193, 195 Forsyth, J. 67, 69 Fox, N. 85 Fox, N.A. 185±186 Foy, H. 88 Francis, M.N.K. 307 Freebody, P. 274 Freedland, R.L. 13 Freedman-Doan, C. 293, 297

Author index Freeman, H. 89 Freud, S. 329, 341 Frick, P.J. 246 Frost, B. 250 Fuerst, D. 87 Fuerst, J.S. 87 Fulker, D.W. 49, 51, 54, 204 Furby, L. 309 Furstenberg, F.F., Jr. 294 Galal, O. 71, 73, 75 Galambos, N.L. 287, 290, 292, 293, 294, 297 Galler, J. 74 Gallistel, C.R. 124 Galsworthy, M.J. 51 Galton, F. 329 Gandour, M.J. 26 Garber, H.L. 87 Garcia, J. 124, 132 Garcia, L. 76 Gardiner, H. 278 Gardner, J. 74±75 GarieÂpy, J.-L. 223, 251 Gaskell, P.C. 56, 58 Gathercole, S.E. 152, 162 Gauvin, M. 275 Gay, J. 125 Ge, X. 54, 61 Geary, D.C. 215 Geber, M. 67 Gelman, R. 124, 125, 127, 131, 134 Gelman, S.A. 106, 109, 123, 124, 127, 129, 207 Gendreau, P.L. 244 Georgieff, M. 74 Gesell, A. 1, 2, 4, 329, 341 Giambrone, S. 115 Gibson, E.J. 8, 9, 15, 339 Gibson, J.J. 8, 339 Giedd, J.N. 289 Gillis, J.J. 51 Girardot, J.J. 249 Glaser, R. 124 Glick, J. 125 Gluckman, P. 66 Goddall, J. 216 Godfrey, K. 66 Goelman, H. 88, 90 Gold, D.P. 318 Goldberg, L.R. 179, 183 Golden, M. 89

Gold®eld, E.C. 17 Goldscheider, L. 146, 149 Goldsmith, H.H. 48, 172, 185, 188, 196 Goldstein, H. 93 Golinkoff, R.M. 131 Gonzales, N. 290 Goodenough, F.L. 248, 249 Goossens, C. 92 Gopnik, A. 103, 105, 109, 111 Gordeeva, T.O. 227, 228 Gorman, K. 67, 68, 72, 73 Gotlieb, I.H. 25 Gottlieb, G. 216, 224, 225, 226 Goyer, R. 71 Graham, J.W. 233 Graham, S. 226 Grantham-McGregor, S. 64, 65, 66, 67, 68, 69, 71, 72, 73, 74±75, 75 Grasshof, M. 227, 228 Graumann, C.F. 177 Gray, J.A. 183, 243 Graziano, W.G. 195 Green, F.L. 103, 105, 112, 114 Green, S.M. 246 Greenberger, E. 291 Green®eld, P.M. 269, 271 Greenough, W. 74 Gribbin, K. 318 Grob, A. 235 Grossi, M.T. 89 Gruber-Baldini, A.L. 307, 317, 318 Guba, E.G. 267, 268 Gudjonsson, G.H. 247 Guilarte, T. 68, 70 Guilford, J.P. 178, 179, 180 Gurtner, J.-L. 279 GuÈss, D. 275 Guthrie, I.K. 195, 290 Guttentag, R.E. 157 Guttman, L. 310 Haan, N. 309, 314 Hadeed, J.A. 86 Hagberg, B. 315 Haight, W.L. 291 Haines, J.L. 56, 58

349

Hair, E.C. 195 Haith, M.M. 117 Hala, S. 106 Halfon, O. 92 Hall, G.S. 287, 308, 329, 335 Hall, N.W. 299 Halterman, J. 67 Halverson, C.F. 207 Halverson, C.F., Jr. 179, 184, 196 Halverson, H.M. 13 Hamilton, M.A. 296 Hamilton, S.F. 296 Hannerz, U. 280 Harden, P.W. 245, 249, 250 Harding, C.G. 131 Harding, J. 66 Hare, B. 115 Harkness, S. 274, 278 Harlow, H.F. 214 Harman, G. 105 Harms, T. 91 Harnish, J.D. 244 Harnishfeger, K.K. 153, 159 Harre, R. 270 Harris, C.W. 309 Harris, J.R. 53, 54 Harris, P.L. 105, 106, 109, 112, 115, 272, 273 Harris, R. 72 Harris, T. 37 Harris, T.O. 37 Harrison, G. 71, 73, 75 Hart, D. 187, 195 Hart, E.L. 246 Harter, K.S.M. 291 Hartup, W.W. 183, 212, 241, 243, 244, 248, 249, 329 Haselager, G.J.T. 187, 188, 190, 193 Haselager, W.S.G. 273, 280 Hasselhorn, M. 154 Hatano, G. 123, 126, 129, 133, 134, 135, 137, 138 Hathaway, S.R. 179 Havill, V.L. 179 Hawkins, J.D. 246, 255

350

Author index

Hawley, P.H. 223, 224, 225, 226, 229, 231, 232, 234 Hawrylewicz, E. 70 Hay, D.F. 246, 251 Heath, A.C. 51, 242 Heelas, P. 273 Heider, F. 339 Held, R. 15 Helwig, C.C. 243 Henderson, C.R. 255 Henderson, H.A. 185±186 Henderson, V.K. 88 Henri 144 Henrich, C.C. 224, 225, 226, 229 Henrichon, A.J. 113 Heron, A. 272, 274 Herre, K. 294 Herrera, C. 168 Herrera, M. 71 Hertenstein, M.J. 16 Hertzog, C. 315 Hetherington, E.M. 35, 39, 51, 53, 54, 172 Hetherington, M.E. 242 Heymans, P.G. 192 Hicks, L. 15 Himes, J. 71, 74±75, 75 Hinde, R.A. 30, 168, 214, 340 Hirsch, N. 168, 172 Hirschfeld, L.A. 124, 127 Hix, H.R. 109 Hobbes, T. 249 Hodges, J. 31 Hofer, S.M. 233, 316 Hoffman, D. 67 Hofmann, V. 187, 195 Hofstee, W.K. 179, 183 Holmes, C.J. 289 Holt, K.G. 9 Hong, Y.Y. 115 Honzik, M.P. 309, 314 Hood, K.E. 216 Horn, J.M. 52 Horwood, J. 252 Horwood, L.J. 34, 293 Howe, M.L. 159 Howes, C. 86, 88, 90, 91, 92 Hrdy, S.B. 244 Hrebickova, M. 181

Huang, J. 75 Hubbard, E.M. 16 Hubbard, J.J. 294 Huberman, A.M. 271 Huesmann, L.R. 242, 247, 253, 254 Hughes, C. 109, 116, 168, 170, 172, 173 Hultsch, D.F. 315 Hurtado, E. 65 Husaini, M. 75 Huston, A.C. 204 Hutchinson, J.E. 135, 139 Hutt, C. 88 Hutt, S.J. 88 Hwang, C. 88 Hwang, C.P. 88 Hyde, J.S. 188 Iacono, W.G. 49, 242 Ialongo, N. 255, 299 Iannotti, R.J. 251 Imbens-Bailey, A.L. 117 Inagaki, K. 123, 126, 129, 133, 134, 135, 137, 138 Intrieri, R.C. 310, 316 Intrieri, R.L. 310, 316 Irvine, S.H. 276 Jacklin, C.N. 204, 210, 212, 244 Jacobs, E. 88 Jacoby, E. 69 Jacques, T.Y. 27 Jahari, A. 75 Jahoda, G. 266, 269, 278 James, W. 337 Japel, C. 249 Jenkins, W.M. 11 Jensen, A.R. 50 Jensen, J.L. 13, 14 Jensen, P.S. 246 Jensen-Campbell, L.A. 195 Jernigan, T.L. 289 Jerome, N. 71, 73, 75, 76 Jessor, R. 291, 293, 294 Joffe, A. 67 Johansson, B. 49 Johansson, I. 86 John, O.P. 187, 188 Johnson, C.N. 105 Johnson, M. 161

Johnson, M.H. 18, 129, 130, 139 Johnson, M.J. 129 Johnston, L.D. 291, 292, 297 Johri, R. 272 Joireman, J. 182 Jones, H.E. 307, 308, 309, 314 Jones, K. 92 Jones, S. 290 Jones, S.M. 226, 229 Jordan, E. 92 JoÈreskog, K.G. 310 Jost, J.T. 106 Jowett, S. 87 Juvonen, J. 226 Kaczorowski, J. 67 Kafterian, S.J. 298 Kagan, J. 40, 243 Kagan, S.L. 86, 91 KagitcËibasi, C. 86, 268 Kahn, R.L. 307 Kail, R. 289 Kail, R.V. 149, 153, 154 Kamm, K. 11, 13, 14 Kaplan, B. 64 Karmiloff-Smith, A. 18, 123, 129, 130, 132, 139 Karp, R. 65 Karraker, K.H. 205 Kasarda, S. 56 Kasser, T. 224, 225, 226 Katz, H. 148 Katz, L.G. 85 Kaukiainen, A. 245 Kay, B.A. 17 Keating, D.P. 288 Keenan, K. 249, 251 Keil, F.C. 124, 127, 128, 129, 133, 138 Kellam, S.G. 255, 298, 299 Keller, H. 277 Keller, M. 187, 195 Kelso, J.A.S. 9, 10, 14 Kemper, T. 70, 74 Kendler, K.S. 51 Kenny, D.A. 232, 236 Kerr, M. 247 Kessler, R.C. 51 Kett, J.F. 287 Ketterlinus, R.D. 293

Author index Keusch, G. 71 Kinderman, T.A. 219 King, A.P. 216 Kirksey, A. 71, 73, 75 Kitzman, H. 255 Klein, M.H. 188 Klein, N. 70, 75, 76 Klein, R. 73 Kleinman, J. 66 Klinnert, M.D. 185, 196 Knowlton, B. 158 Kochanska, G. 27, 193, 195 Koelling, R.A. 124, 132 Koening, A.L. 27 Koffka, K. 144 Kohlberg, L. 207, 341 Kohnstamm, G.A. 179, 184, 196 Konczak, J. 13 Kontar, F. 249 Koppenaal, R.J. 148 Korman 149 Kosmitzki, C. 278 Kosslyn, S. 48, 51 Kotovsky, L. 131, 136 Kracke, B. 295 Kraft, M. 182 Krasnor, L.R. 254 Kreft, I. 231 Kreppner, K. 232 Krewer, B. 269, 274 Kritchevsky, S. 92 Krucher, R. 92 Kruger, A.C. 137 Kruglanski, A.W. 106 Kruglyak, L. 57 Krull, A. 148 Ku, H.-C. 245 Kugler, P.N. 9, 10 Kuhlman, D.M. 182 Kuhlmann, F. 155 Kuhn, D. 104, 106, 161 Kuperminc, G. 294 Kupersmidt, J.B. 40 Kurtz, L. 255 Laboratory of Comparative Human Cognition (LCHC) 127 Labouvie-Vief, G. 315 LaFrance, M. 208 Lagattuta, K.H. 172 Lagerspetz, K.M. 244

Lahey, B.B. 246 Laible, D.J. 171 Laird, R. 252 Lalonde, C. 243 Lamb, M.E. 88, 92, 293 Land, K.C. 246, 252 Lau, A. 243 Laurent, D. 249 Lawrence, J. 276 Lay, K.-L. 33 Lazar, I. 87, 88 Leadbeater, B.J.R. 287, 290, 291, 292, 294, 298, 300 Leaper, C. 212, 213 Leaper, S. 56 Ledoux, S. 249 Lee, D.N. 15 Lee, P.P. 103 Lee, V.E. 87 Leffert, N. 293 Lefkowitz, M.M. 242 LeMarquand, D. 255 Lemerise, E.A. 173, 289 Lemery, K.S. 172, 185, 196 Lemmon, H. 56 Lengua, L.J. 290 Lerner, R.M. 224, 230, 293, 294, 299 Leslie, A.M. 109, 113 Lester, B. 75 Levin, H. 25, 203 Levitsky, D. 65, 70, 73, 74 Lewin, K. 340, 341 Lewis, J.M. 39 Lewis, M. 244 Lewis, M.D. 223 Li, J. 68 Lichenstein, P. 242 Lickliter, R. 216 Lillard, A.S. 114, 115, 168, 272, 273 Lin, N.T. 114 Lindenberger, U. 227, 310, 335 Ling, X. 299 Lipsitt, L.P. 3 Lira, P. 73 Little, T.D. 223, 224, 225, 226, 226±227, 227, 228, 229, 230, 231, 232, 232±233, 233, 234, 235

351

Litvinovic, G. 268 Lloyd, B. 211 Lloyd, E. 88, 89, 90, 92 Lochman, J.E. 244, 255 Locke, J. 249 Lockman, J.J. 17 Loeber, R. 243, 246, 247, 249, 252, 255 Loehlin, J.C. 51, 52, 178, 185 Lopez, D.F. 226, 228 Lopez, I. 72 Lord, F.M. 309 Lorenz, K. 30, 247 Losoya, S. 290 Lozoff, B. 67, 68, 69, 70, 73, 74, 75, 76 Luciana, M. 18 Luckey, D. 255 Luscher, K. 72 Luthar, S.S. 294 Lutz, C. 168 Lykken, D.T. 49 Lynam, D. 191, 252 Lynch, S. 71 Lynskey, M.T. 34, 38, 293 Lytton, H. 204 Maccoby, E.E. 25, 26, 202, 203, 204, 207, 210, 212, 213, 215, 216, 244 MacDonald, D. 289 MacMillan, H.L. 291 Madden, P.A.F. 242 Maddox, G.L. 315 Maggs, J.L. 297 Magnusson, D. 178, 180, 191, 224, 226, 251, 296, 297, 316 Mahady Wilton, M.M. 290 Maitland, S.B. 310, 316 Malewska-Peyre, H. 267, 275 Maltz, D.N. 218 Manke, B. 172 MaÈnnikoÈ, K. 191 Manuel, M. 70, 75, 76 Marchant, L.F. 216 Marcoen, A. 33 Marcoulides, G.A. 224, 233 Markman, E.M. 111

352

Author index

Marolla, F. 67 Marsland, K. 224, 225, 226 Martin, C.L. 206, 207, 210, 212 Martin, K. 65 Martin, N.G. 242 Martin, R.P. 179, 184, 196 Martin, S. 88, 89, 90, 92 Martinez, C. 71, 75 Martorell, R. 68, 72, 73 Mason, K.G. 309 Mason, W.H. 309 Massey, C.M. 131 Masten, A.S. 34, 294 Matama, H. 216 Mathews, R. 65 Maton, K. 300 Matthews, C.G. 277 Mayer, K.U. 235, 315, 322 Mayer, L.S. 255 Mayer-Kress, G. 17 Mbsse, L.C. 255 McArdle, J.J. 233 McBurnett, K. 246 McCabe, G. 71, 73, 75, 76 McCarthy, D. 331 McCartney, K. 40, 90, 92 McClearn, G.E. 49, 56, 317 McClelland, J.L. 129 McClintock, J. 291 McClish, D. 70, 75, 76 McConkey, C. 87 McCord, J. 247, 255 McCrae, R.R. 179, 181, 190 McCullough, A. 71 McDonald, M. 75, 76 McDuff, P. 245, 249, 250 McFarlane, A.C. 35 McGraw, M.B. 1, 2, 3, 5 McGrew, W.C. 216 McGue, M. 49, 242, 317 McGuf®n, P. 56 McGuire, J. 92 McGuire, S. 54, 172, 242 McKey, H.R. 87 McKinley, J.C. 179 McLin, D. 17 McLoyd, V. 291, 294 McNew, S. 105

Meacham, J.A. 148 Mead, G.H. 340 Meany, M.J. 216 Medrano, Y. 76 Meeks-Gardner, J. 71, 75 Meijer, A. 65 Meins, E. 169, 171 Meisels, S.J. 86 Melhuish, E.C. 85, 86, 88, 89, 90, 91, 92, 93, 94, 95 Meltzoff, A.N. 109, 111, 151 Mena, P. 67 Merali, Z. 216 Meredith, W. 316 Merialdi, M. 67, 69 Merikangas, K.R. 57 Merisca, R. 299 Merline, A.C. 292 Merriam, S.B. 315 Mervielde, I. 179 Merzenich, M.M. 11 Metcalfe, J. 106 Meumann, E. 145 Meyer, K. 72 Michael, R.T. 89 Miles, M. 271 Miles, W.R. 308 Miller, J.G. 271 Miller, L.S. 247, 253, 255 Miller, N.E. 341 Miller, P.H. 103, 104, 106, 110, 112, 113, 114, 116, 117, 153, 154, 243 Miller, S.A. 104, 153 Mills, P. 89 Mills, W. 256 Mischel, W. 178, 206, 230 Mishra, K. 276 Mishra, R.C. 266, 272, 276, 278, 279, 280 Mitchell, C.M. 290 Mitchell, P. 106 Mitchell, R. 106 Modell, J. 340 Moen, P. 72 Mof®tt, T.E. 187, 191, 246, 252 Molenaar, P.C.M. 11, 17 Montagner, H. 249 Monteior, A. 72 Montplaisir, J. 249

Mooney, A. 85, 88, 89, 90, 92 Moore, C. 106, 169 Moore, J.D. 280, 281 Mora, J. 71 Morgan, A.G. 92 Morgan, R. 9 Morgan, S.P. 294 Morgane, P. 70, 74 Morison, S.J. 32 Morris, P. 255 Morris, S. 73 Moses, L.J. 109, 111 Moshman, D. 104 Moshmann, D. 274 Moss, P. 85 Mostofsky, D. 70, 74 Moussa, W. 75, 76 Mowrer, O.H. 341 Mrazek, D.A. 185, 196 Muir, D. 15 Munakata, Y. 129 Munton, A.G. 85 Murphy, B.C. 195, 290 Murphy, G. 178 Murphy, L.B. 249 Murray, H. 87, 88 Murray, K. 27 Musen, G. 158 Mussa-Ivaldi, F.A. 18 Mussen, P.H. 309, 314 Mutter, J. 278 My Lien, N. 72 Myrtle, B. 3 Nagel, K. 106, 111 Nagin, D.S. 246, 252 Nagy, L. 145 Naito, M. 114 Neale, M.C. 51, 233 Neckerman, H.J. 251 Needham, A. 131, 136 Neiderhiser, J.M. 52, 53, 54, 61 Nelson, C. 18, 74, 289 Nelson, C.A. 162 Nelson, E. 70, 75, 76 Nelson, K. 151 Nelson, L. 66 Nelson, T.O. 106 Nesselroade, C.S. 319 Nesselroade, J.R. 309 Netschajeff, A. 147 Neubauer, A.C. 48

Author index Neumann, C. 75, 76 Newell, A. 124 Newell, K.M. 11, 17 Newport, E.L. 132, 337 Newton, P. 168, 170 Nguyen, H. 313 NICHD Early Child Care Research Network 89, 90, 92 Nichida, T. 216 Nicholson, C. 213 Nigg, J.T. 48 Nitz, K.A. 293 Nixon, C.L. 115 Noack, P. 295 Northwood, A. 294 Norton, J. 66 Nowlin, J.B. 315 Nsamenang, B. 279 Nugent, L. 153 Nurmi, J.E. 191 O'Connell, J.C. 89 O'Connor, T. 168, 172 O'Connor, T.G. 29, 48, 51, 54, 204, 242 Odbert, H.S. 178 Oden, M. 309, 314 O'Donnell, A.K. 114 È sterman, K. 245 O Oettingen, G. 226±227, 227, 228 Offner, M. 144, 145, 147 Ogay, T. 275 O'Hanlon, A.M. 318 Olds, D. 255 Olencik, M. 88 Oligny, P. 244 Oliver, B. 51 Olson, D. 105, 106 Olson, D.R. 167 Olson, R.K. 56 Olthof, T. 192 Olweus, D. 244, 250±251, 255 O'Malley, P.M. 291, 292, 297 O'Nell, T.D. 290 Orlansky, H. 25 Ostendorf, F. 181, 190 Otero, G.A. 69 Owen, M.J. 56 Owens, J. 66 Owens, W.A., Jr. 309, 314

Pagani, L. 255 Pagani-Kurtz, L. 247 Paine, P. 72 Pal, A. 244 Palmore, E. 315 Palti, H. 65 Pandey, J. 267 Panksepp, J. 212 Parent, J. 291 Parham, I.A. 318 Parisi, D. 18, 129, 130, 139 Parke, R.D. 241, 243, 340 Parker, J.G. 219 Parkin, A.J. 157 Parkin, L. 114 Parks, Y. 69 Parnass, J. 213 Pasquali, J. 72 Pasupahti, M. 232 Patterson, C.J. 40 Patterson, G.R. 247, 252, 253, 254, 334 Paya-Cano, J.L. 51 Peake, P.K. 178 PeÃcheux, M. 17 Pedersen, N.L. 49, 317 Pedroso de Lima, M. 190 Peirano, P. 67 Peisner-Feinberg, E.S. 86, 91 Peltonen, T. 244 Pence, A.R. 88, 90 Penland, J. 68 Pennington, B.F. 56 Pepler, D.J. 290 PeÂrez, S. 279 Periano, P. 65 Pericak Vance, M.A. 56, 58 Perner, J. 105, 109, 114, 117 Perry, D.G. 207 PeÂrusse, D. 245, 249, 250 Petersen, A.C. 293 Peterson, C. 92 Peterson, C.C. 115, 136 Peterson, R. 92 Petrill, S.A. 48, 49, 50, 56 Petrogiannis, K. 86 Pettit, G.S. 27, 244, 252 Pettitt, L. 255 Pfeiffer, S. 66 Philliber, S. 294

353

Phillips, D. 90 Phillips, S.J. 13 Phinney, J.S. 297 Piaget, J. 5, 10±11, 15, 16, 103, 127, 138, 171, 329, 338, 339, 341 Pickens, J.N. 150 Pierrehumbert, B. 92 Pihl, R.O. 245, 246, 249, 250, 255 Pike, A. 54, 172 Pilek, E. 146, 149 Pillow, B.H. 112, 113 Pine, D. 289 Pinker, S. 51 PitkaÈnen, L. 241, 244, 245 Plantz, M. 87 Plomin, R. 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 58, 59, 172, 178, 204, 242, 257, 307, 311, 317 Plunkett, K. 18, 129, 130, 139 Policare, H.J. 89 Pollitt, E. 65, 66, 67, 68, 69, 72, 73, 75 Ponomarev, I. 153 Poole, W.K. 309 Poon, L.W. 315 Poortinga, Y.H. 266, 267, 268, 271, 275, 277, 278, 280 Popkin, B. 65 Porcayo, R. 69 Posada, G. 33 Posterski, D.C. 300 Poulin, F. 229, 255 Poulin, R. 290 Poulin-Dubois, D. 110 Pourtois, J.-P. 271 Povinelli, D.J. 115 Powell, C. 66, 67, 68, 72, 74±75 Powers, E. 255 Powers, J. 255 Powlishta, K.K. 210 Prabucki, K. 75 Premack, D. 104, 105 Prescott, E. 92 Pressley, M. 147, 149, 152, 154, 155 Preyer, W. 1, 7, 144, 161, 329

354

Author index

Price, J.M. 229 Price, T.S. 51 Pritchard, M. 117 Pugh, G. 86 Pulkkinen, L. 191, 244, 245, 251 Purcell, S. 51 Qi, S. 272 Quay, H.C. 246 Quetelet, A. 246 Quintana, S.M. 291 Quinton, D. 36, 37 Rabbitt, P. 315 Rabinowitz, S. 70, 74 Radke-Yarrow, M. 73 Ramey, C.T. 87, 89 Ramsey, E. 252 Ramstein, T. 92 Rao, R. 74 Rasmussen, K. 70 Ratner, H.H. 137 Rau, L. 203 Raudenbush, S.W. 93, 231, 310 Reaux, J.E. 115 Rebok, G.W. 255 Redfern, S. 171 Rees, E. 124 Reid, J. 313, 322 Reifel, S. 86 Reinert, G. 310 Reiser, M. 290 Reiss, D. 51, 53, 54, 172, 242 Rende, R. 178 Repacholi, B.M. 111 Resnick, L.B. 126 Resnick, O. 70 Restoin, A. 249 Reyna, V.F. 158, 159 Reynolds, C.A. 317 Rhodes, J.E. 291 Ricciuti, H. 72 Ricco, G. 15 Richards, D.D 135 Richman, N. 92 Richmond, J.B. 293 Ridderinkhof, K.R. 161 Ridge, B. 27 Ridley-Johnson, R. 14 Riegel, K.F. 307 Riemann, R. 48

Rietz, W. 310 Riggs, I.U. 106 Riksen-Walraven, J.M. 187, 188, 190 Riley, J.W., Jr. 307, 320 Riley, M.W. 307, 320 Risaldo, S. 216 Risch, N. 57 Rivera, J. 68, 72, 73 Roberts, B.W. 186 Robertson, R. 172 Robertson, S.S. 17 Robin, A.F. 135, 139 Robins, R.W. 187, 188 Robinson, J. 66 Robinson, J.L. 172 Rochat, P. 9 Rock, S.L. 92 Rodriguez, D. 249 Roeser, R.W. 293, 297 Rogoff, B. 137, 139, 269 Rogosa, D. 309, 310 Rogosch, F.A. 234 Romney, D.M. 204 Rose, J.L. 17 Rosenberg, D. 15 Rosenbluth, D. 30 Rosenbluth, L. 89 Rosenthal, M.K. 86 Roses, A.D. 56, 58 Rosicky, J.G. 111 Ross, M. 292 Roth, J.A. 246 Rothbart, M.K. 26, 179, 183, 184, 191, 193, 196 Rott, C. 315 Rovine, M.J. 52 Rowe, D.C. 54 Rowe, J.W. 307 Rowland, L. 85 Roy, R. 213 Royce, J. 87, 88 Rubin, K.H. 185±186, 219, 254 Ruble, D.N. 206, 207, 210, 212 Rudinger, G. 310 Ruffman, T. 114 Russell, J.T. 169, 171 Rustici, J. 86, 91 Rutter, M. 29, 32, 34, 36, 37, 38, 48, 51, 54, 58, 59, 87, 204, 257

Ryan, R.M. 224, 225, 226 Ryder, N.B. 309 Sabatier, C. 274 Saco-Pollitt, C. 75 Saenger, G. 67 Sahlins, M. 280 Salmivalli, C. 297 Salmona, M. 70 Salthouse, T. 313 Salthouse, T.A. 153 Sameroff, A. 72 Sammons, P. 91, 93, 94, 95 Samuelsson, G. 315 Sanchez, G. 76 Sandiford, P. 76 Sandler, I. 294 Sandler, I.N. 290 Sandstead, H. 68, 71 Sanson, A. 26 Sapru, S. 275 Sartorius, N. 278 Sasse, D.K. 288 Saudino, K.J. 172 Saults, J. 153 Saunders, A.M. 56, 58 Savelsbergh, G.J.P. 14 Savin-Williams, R.C. 291 Saxton, M.E. 131 Sayer, A.G. 311 Saylor, C.F. 35 Scarr, S. 27, 32, 40, 49, 52 Schaffer, H.R. 24, 26, 34, 35, 36, 39 Schaie, K.W. 307, 308, 309, 310, 311, 313, 314, 315, 316, 317, 318, 319, 320 Scheier, C. 11, 17 Schellenbach, C. 300 Scheper-Hughes, N. 76 Schlegel, A. 266 Schlieker, E. 88 Schliemann, A. 275 Schmechel, D.E. 56, 58 Schmidt, L.A. 185±186 Schmidt, R.C. 17 Schmitz-Scherzer, R. 308, 315 Schmuckler, M.A. 15 Schnabel, K.U. 224, 232±233, 233 Schneider, K. 11, 13, 14

Author index Schneider, W. 104, 143, 147, 149, 152, 154, 155, 156, 157, 161 Schnur, E. 87 Schoelmerich, A. 277 Scholte, R.H.J. 187, 188, 190 SchoÈner, G. 11, 17 Schore, A. 74 Schraw, G. 274 Schroots, J.J.F. 307 Schulenberg, J. 291, 292, 297 Schultz, A. 294 Schumacker, R.E. 224, 233 Schwanen¯ugel, P.J. 112 Schwartzman, A. 318 Schweinhart, L.J. 86, 87, 88, 255 Scott, K. 65 Sears, R.R. 25, 203, 341 Secord, P.F. 270 Segall, M.H. 266, 268, 269, 271, 275, 277, 278, 280 Seier, W.L. 154 Senn, M.J.E. 330 Serpell, R. 278 Sethuraman, K. 65, 66, 67, 68 Shadmehr, R. 18 Shaffer, D. 246 Shaheen, F. 73 Shanahan, M. 224 Shanahan, M.J. 291, 292 Shanan, J. 315 Shantz, C.U. 103 Sharp, D.W. 125 Shatz, M. 105 Shaw, D.S. 251 Shef®eld Morris, A. 292 Sheldon, A. 213 Sheldon, K.M. 224, 225, 226 Shepard, S.A. 195, 290 Shepherd, D.M. 35 Shinamura, A.P. 106 Shiner, R.L. 179, 196 Shirley, M.M. 1, 3, 14, 16 Shock, N. 308 Shoda, Y. 178, 230 Shonkoff, J.P. 86, 299 Shott, S. 66

Shulman, S. 292 Shweder, R.A. 267, 269 Sibley, A. 92 Sidora, K. 255 Siegal, M. 115, 136 Siegler, I.C. 310, 315, 316 Siegler, R.S. 127, 129, 135, 152 Sigman, M. 67, 69, 75, 76 Silber, S. 105 Silbereisen, R.K. 295 Silberg, J. 48, 51 Silva, P.A. 187, 190, 191 Silva, P.S. 246 Silver, A. 65 Silverman, A.J. 276 Simeon, D. 64, 68 Simmons, R.G. 191 Simoes, A. 190 Simon, H.A. 124 Simon, T. 308, 329 Simonoff, E. 48, 51 Simpson, A. 213 Sinclair, A. 112 Singer, J.D. 310 Sinha, D. 276, 278, 279 Siraj-Blatchford, I. 91, 93, 94, 95 Skala, K. 14 Skeels, H. 28, 39 Skinner, E.A. 225, 226, 227 Skinner, M.L. 247 Sklar, R. 65 Skuder, P. 56 Skuse, D.H. 28, 32, 33 Slaby, R.G. 241, 243 Slaughter, V. 117 Slomkowski, C. 136 Slutske, W.S. 242 Small, B.J. 315 Small, G.W. 56, 58 Small, S.A. 299 Smider, N.A. 188 Smirnov, A.A. 148, 149 Smith, C. 211 Smith, D.L. 56 Smith, J.S. 232 Smith, L.B. 10, 11, 17 Smith, P.K. 167, 244 Smith, R.S. 34 Smith, S.D. 56 Snipper, A. 87, 88 Sobhy, A. 75

355

Sodian, B. 154 Solarz, A. 300 SoÈrbom, D. 310 Sowell, E.R. 289 Sparling, J.J. 87 Spearman, C. 50 Spelke, E.S. 129, 132, 134, 337 Spencer, J.P. 11, 13, 14 Spinath, F.M. 48 Spitz, R.A. 28, 34 Spurr, G. 65 Squire, L. 158 Sroufe, L.A. 33 St Clair, D. 56 Stanford, C.B. 216 Stanley-Hagan, M. 35, 39 Stanton, W. 246 Starr, J.M. 56 Statham, D.J. 242 Stattin, H. 191, 297 Staudinger, U.M. 335 Stavy, R. 135, 136 Stein, Z. 67 Steinberg, L. 291, 292 Stern, C. 146 Stern, D.N. 211 Stern, M. 205 Stern, W. 70, 144, 146, 177, 179, 180, 196, 329, 335 Sternberg, K. 88 Stetsenko, A. 226±227, 227, 228 Stevenson, J. 38, 51, 242 Stewart, M. 68 Stewart, M.A. 54, 61 Stewart, P. 92 Stice, E. 290 Stigler, S.M. 309 Stoffregen, T.A. 15 Stolzfus, R. 64 Stoolmiller, M. 247 Stouthamer-Loeber, M. 187, 243, 246, 247, 249 Strayer, F. 214 Strayer, F.F. 214, 249 Strayer, J. 249 Strelau, J. 48 Streri, A. 17 Strittmatter, W.J. 56, 58 Strodbeck 269 Strohschneider, S. 275 Strother, C.R. 314

356

Author index

Strupp, B. 65, 70, 73, 74 Stunkard, A.J. 65±66 Styfco, S. 86 Subotnik, R. 192 Sullivan, M.A. 269 Sullivan, M.W. 244 Sully, J. 249 Sultana, R. 279 Sunar, D. 86 Suomi, S.J. 216 Super, C. 71, 278 Super, C.M. 274, 278 Susser, M. 67 Suter, J. 294 Sutherland, D.H. 13 Sveistrup, H. 14 Svensson, T. 315 Swallow, J.A. 276 Sweeney, A.L. 315 Swei®, E. 73 Swettenham, J. 115, 117 Sylva, K. 86, 87, 91, 92, 93, 94, 95 Szilagyi, P. 67 Taber, S. 291 Tacconi, M. 70 Tager-Flusberg, H. 106, 115, 117 Taggart, B. 91, 93, 94, 95 Talpain, B. 249 Tanner, J.M. 29 Tanon, F. 275 Taormina, J. 15 Taylor, J. 242 Taylor, M. 106, 113 Tellegen, A. 49, 179, 182 Terman, L.M. 314 Terry, R. 253 Tesla, C. 136 Teta, P. 182 Theall, L.A. 115 Thelen, E. 1, 2, 5, 7, 8, 10, 11, 13, 14, 16, 17 Thomae, H. 178, 181, 308, 315 Thompson, H. 2, 4 Thompson, L.A. 50, 56 Thompson, P.M. 289 Thompson, R.A. 171 Thorndike, E.L. 309 Thorne, B. 218 Thurstone, L.L. 310 Tidball, G. 111

Tilton-Weaver, L.C. 292, 293, 297 Titzer, R. 17 Tizard, B. 31 Tjebkes, T.L. 193, 195 Toga, A.W. 289 Toller, S. 66 Tomasello, M. 106, 111, 115, 137, 272 Tonkiss, J. 74 Tooby, J. 139 Topka, H. 13 ToÈrestad, B. 180, 296 Tremblay, R.E. 241, 244, 245, 246, 247, 249, 250, 252, 255 Tresini, M. 307 Triana, N. 75 Tripathi, R.C. 278 Troughton, E. 54, 61 Trudel, M. 249 Tucker, L.R. 310 Turbin, M.S. 294 Turic, D. 56 Turvey, M.T. 9, 10 Twohey, J. 294 Tyler, S. 88 Uauy, R. 65, 67 Ullmann, V. 249 Ulrich, B. 10, 11, 13, 14 Underwood, B.J. 146 UNICEF 64 United Nations Subcommittee on Nutrition 64 Valenzuela, M. 70 Valsiner, J. 224, 268, 276 van Aken, M.A.G. 187, 188, 190, 195 van der Kamp, J. 14 Van der Molen, M.W. 161 Van Kammen, W.B. 246 van Lieshout, C.F.M. 177, 183, 187, 188, 190, 192, 193 Van Meter, P. 154 Vandell, D.L. 86, 88, 188 Vandergeest, K.A. 27 Varma, S. 67, 69 Vereijken, B. 13 Verschueren, K. 33

Vignetti, S. 56 Vinden, P.G. 272, 273 Visher, C.A. 246 Vitaro, F. 244, 246, 247, 252, 253, 255 Vitunski, E.T. 292 Volker, C. 307 von Eye, A. 295 von Hofsten, C. 13, 15 Vuori, L. 71 Vygotsky, L.S. 340 Wachs, T. 26 Wachs, T.D. 64, 67, 68, 69, 70, 71, 72, 73, 75, 76 Wadsworth, K.N. 291, 297 Wadsworth, S.J. 50, 51 Wahi, S. 272 Wahler, R.G. 253 Wahlsten, D. 216 Wainright, P.E. 69, 70 Waire, A. 213 Wakschlag, L.S. 249 Walder, L.O. 242 Waldman, I. 246 Walka, H. 75 Walker, S. 66, 67, 69, 72 Wallen, K. 217 Wallerstein, J.S. 39 Wallis, J. 216 Walsh, C.A. 291 Walters, R.H. 243, 341 Wandersman, A. 298 Wanner, B. 225, 226, 228, 235 Warren, W.H. 17 Wasik, B.H. 87 Wasserman, G.A. 211, 255 Wassmann, J. 272 Waters, E. 33 Waters, H.S. 154 Watkins, W. 67, 69 Watson, A.C. 115 Watson, D. 182 Watson, J.B. 329, 337, 341 Watson, J.K. 102, 167 Wauben, P.M. 69, 70 Wax, N. 135, 136 Way, N. 290, 291, 292, 294, 298

Author index Wearing, A.J. 235 Webster-Stratton, C. 255 Wechsler, D. 308 Weikart, D.P. 87, 255 Weikhart, D. 86, 87, 255 Weinberg, R.A. 52, 294, 299 Weinert, F.E. 161 Weisz, J.R. 226 Welford, A.T. 308 Wellman, H.M. 102, 103, 105, 106, 109, 112, 115, 123, 129, 155, 167, 169, 172 Werker, J.F. 243 Werner, E.E. 34 Werner, H. 144, 147, 148, 333 West, M.J. 216 West, S.G. 290 Whaley, S. 67, 69 Whalley, L.J. 56 Whang, S.-M. 291 Wharton, B. 69 White, A. 173 White, B. 15 White, D.R. 88 White, G.M. 168 Whitebook, M. 90 Whiting, J.W.M. 269, 278 Whitmarsh, G.A. 148 WHO (World Health Organization) 92 Widaman, K.F. 232

Wilcox, S.A. 105 Willatts, P. 67, 69 Willcutt, E. 56 Willerman, L. 52 Willett, J.B. 309, 310, 311 Williams, G.C. 244 Willis, S.L. 307, 310, 313, 316, 317, 318, 319, 320, 322 Wilpers, S. 195 Wilson, A. 115 Wilson, K.S. 88 Wimmer, H. 105 Windelband, W. 177 Winick, M. 72 Winick, N. 72 Winsborough, H.H. 309 Wiser, M. 126 Witelson, S.F. 276 Witherington, D. 16 Witmer, H. 255 Wohlwill, J. 310 Wolchik, S.A. 290 Wolf, A. 215 Wolfe, B. 86 Wolkind, S. 26 Woodhead, M. 85 Woodruff, G. 104, 105 Woods, R.P. 289 Woodward, A.L. 111 Woollacott, M. 14 Woolley, J.D. 103, 114 Wright, M.O'D. 294 Wright, V. 253

357

Wundt, W. 143, 329, 339 Yang, J. 68 Yarrow, M.R. 26 Yaschine, J. 71 Yates, W. 54, 61 Yazejian, N. 86, 91 Ybarra, V.C. 291 Ye, R.M. 184 Yehuda, S. 70, 74 Yendovitskaya, T.V. 149 Yerkes, R.M. 308, 314 Youngblade, L. 136 Yunis, F. 73, 75 Zaharia, M.D. 216 Zahn-Waxler, C. 172, 251 Zajonc, R.B. 212 Zavaleta, N. 67, 69 Zeitlin, M. 76 Zelazo, J. 86, 91 Zelazo, P.D. 167, 243 Zernicke, R. 11, 13, 14 Zhang, X.D. 272 Zhao, F. 68 Zigler, E. 86, 294 Zigler, E.F. 299 Zimowsky, M. 309 Zinchenko, P.I. 148 Zoccolillo, M. 245, 249, 250 Zou, H. 272 Zuckerman, M. 182 Zuo, Y.L. 311

Subject index

Abnormality, 51, 55 Abstract principles, 130, 132, 134 Abused children, 38, 40 Achieving overcontrollers, 188, 189 Action-control beliefs, 226±228 Action research, 299 Action sequence, 227 Actions, goal-directed, 226±227 Activation, 192 Activities, 318 Activity level, 35, 185 Adjustment, 190 Adolescence, 287±306 adaptation, 289±290 aggression, 246±247, 253±254 binge drinking, 297 cognition, 288 contexts of development, 292±296 co-occurrence of problems, 292±293 cultural differences, 290±291 delinquency, 34 disabled, 291 emotion regulation, 289±290 ethnic differences, 290±291 foster care, 291 gender differences, 290±291 historical and social contexts, 294±296 homosexuality, 291 as a life stage, 287 methodology, 296±298 mortality, 287 neurological changes, 289 positive psychosocial outcomes, 293±294 programme and policy-relevant research, 298±300

relationships, 172±173 resilience, 294 school-to-work transition, 296 time use, cross-national comparisons, 296 transition to adulthood, 291±292 Adult development, 307±328 advanced old age, 322±323 cognition, 311±313 cognitive ageing, 316, 317±321 history, 308 interventions, 319 methodology, 308±311 structural invariance of constructs, 316±317 theoretical framework, 311±313 Adult Involvement Scale (AIS), 92 Adulthood implicit theories, 292 transition to, 291±292 Affect positive/negative, 184 regulation, 193 Affectionless character, 30 Af®liativeness, 184 Affordances, 8 Age-crime curve, 246 Age-vulnerability relationship, 39 Agency, 189, 223±240 context, 224, 225, 230±232 modelling, 232±233 organismic models, 223, 224±230, 236±237 Agent-centred approaches, 180, 181±186, 196 Agentic action, 227 Agentic resilients, 188, 189 Agents, 180

Subject index Aggressive behaviour, 241±265 adolescence, 246±247, 253±254 de®nition, 241±245 developmental courses, 253±255, 256 emotion regulation and understanding, 173 environmental factors, 247, 257 family interactions, 253±254 genetic factors, 257 interventions, 254±255 means±end analysis, 229 onset, 245±249 personality, 195 rating scales, 242 seriousness, 249±250 sex differences, 244±245 social-information-processing models, 254 social learning theory, 247±248, 253 socially positive 251, 257 stability, 250±253 victims, 243 Agreeableness, 179, 182, 184, 193, 195 Alzheimer's dementia, 56, 58, 313 Anaemia, see Iron de®ciency anaemia Anger, 184, 185 control, 255 proneness, 183 Animals, theory of mind, 115 Anthropology, 267, 269, 280 Antisocial behaviour, 241±243 intervention, 255 Antisocial undercontrollers, 188, 189±190 Anxiety, depression and, 51 Apolipoprotein E, 56, 58, 313 Appearance±reality distinction, 105, 108, 113 Applied developmental psychology, 278±279 Applied theory of mind, 107 Apprehension span, 153 Arousability, 183 Articulatory loop, 152±153 Assessment Pro®le for Early Childhood Programs, 92 Associationism, 332 Associative clustering, 148 Attachment, 30±33 day care, 89 mindreading skills, 169±170, 171

359

Attention joint attention, 137 malnourishment, 75 span, 184 theory of mind, 111, 112 Autism, 103, 115 Autobiographical memory, 151 Autosuggestion, 146 Bayley Scales of Infant Development, 4 Behaviour change, 2, 331±336 determinants, 336±342 Behaviour execution, 192±193 Behaviour monitoring, 207 Behaviour regulation, emotion-related, 289±290 Behavioural Activation System (BAS), 183 Behavioural genetics, 47±63 de®nition, 47 developmental, 49 environmental, 52±55 molecular, 55±59 multivariate, 49±51, 54 quantitative, 48, 55 Behavioural genomics, 59 Behavioural Inhibition System (BIS), 183 Belief-desire psychology, 109 Beliefs, 102, 112±113, 116±117 action-control, 226±228 parental belief systems, 273±274 see also False beliefs Berkeley Growth and Guidance studies, 314 Biases, 129, 130 Big-®ve personality factors, 179, 182±183, 184, 186±187, 193, 194, 195 Binge drinking, 297 Biological beings, 180 Biology behaviour and, 51 gender development, 208±209 motor development, 2±3 nutrition and behavioural development, 70, 72 taxonomy, 134±136 Biomechanics, 14 Body memory, 17 Boot-strapping, 130, 131 Brain adolescence, 289 sex differences, 209

360

Subject index

Breastfeeding, 65 Brinley plots, 316 Bullies, 173, 244, 290 Calendars, 266±267 California Q-set, 179 Candidate genes, 57 Capacity beliefs, 227 Catch-up phenomenon, 28±29, 33 Catching, 13 Causality, 267 Causality beliefs, 227 Central nervous system, dietary in¯uences, 70, 74 environment links, 74 plasticity, 11±12, 18 Centre effects, 94 Change measurement, 309, see also Behaviour change Child abuse, 38, 40 Child care, see Pre-school experience Child Care Facility Schedule, 92 Child-rearing, 25±27 sex-typing, 203, 204 Chronic disease, cognitive decline, 317±318 Chronological age, 310 Circumplex models, 179 Clay moulding model, 24, 25 ``Clinker method'', 315 Cliques, 219 Cluster analysis adolescence, 297 personality types, 187 Clustering, 148 Cognition adolescence, 288±289 cognitions, 192 constrained, 137 cross-cultural differences, 125 deprivation, 28 environmental factors, 311, 313, 318 general cognitive ability (g), 49, 50±51, 52±53, 56 genetic factors, 48, 50, 313, 317 ``higher'', 17 motor skill development, 16, 17 Cognitive ageing, 316, 317±321 Cognitive competence, 195 Cognitive development adults, theoretical frameworks, 311±313 aggression regulation, 255 innate constraints, 139

Cognitive-developmental theory, 341 Cognitive primitives, 153 Cognitive science, 271, 273 Cognitive style, cognitive ageing, 318 Cognitive theories, 341 gender development, 206±208, 209±210 Cognitive triage, 158 Cohort variance, 319±320 Collaborative discourse, 213 Collective capacity beliefs, 227 Collective ef®cacy, 227 Colour clustering, 148 Committed compliance, 193 Common pathways, 334 Communal resilients, 188, 189 Communion, 189 Comparative cognitive psychology, 269 Complex traits, 56, 57 Compliance, committed/situational, 193 Compliant agents, 110 Computer-assisted movement analysis, 12 Computer metaphor, 150 Conceptual change, 126 Conceptual clustering, 148 Conceptual development, 123±142 domain-speci®c innate constraints, 128±133 history, 123±125 sociocultural constraints, 128, 133±138 Conduct disorder, early and late onset, 246 Con®gural frequency analysis, 188, 297 Con®rmatory factor analysis, 310, 316 Connectionism, 132 Conscience, 27 Conscientiousness, 179, 182, 184, 193, 195 Consciousness, 114 Consortium of Longitudinal Studies, 87 Constraints, 125, 127±128 Constructed knowledge, 126 Constructivism, 267 Constructs, 331±332 heterotypic versus homotypic, 234 structural invariance across age, 316±317

Subject index Context, 339±340 adolescent development, 292±296 agency, 224, 225, 230±232 nutrition and behaviour, 72±73 Continuity, 38±40, 49, 181, 185±186, 333 Control, 12±16 Control expectancy, 227 Controversies, 280±281 Conversational experiences, mindreading skills, 170±171 ``Conveyer belt'' phenomenon, 39 Coordination, 12±16 Coping behaviours, 228±229 Cortisol assays, 51 Cost-effectiveness, pre-school programmes, 86±87 Crawling, 13, 14, 16 Creative talent, 192 Criminology, 246 Critical period, 30 hypothesis, 30±33 Cross-cultural differences adolescence, 290±291 biological taxonomy, 135±136 cognition, 125 mindreading, 168 motor development, 18 theory of mind, 115, 272±273 Cross-cultural psychology, 266±286 applications, 278±279 ``emic'' approach, 269 folk psychology, 273±274, 279 goals, 269 lifespan developmental psychology, 276 metacognition, 274±275 methodology, 268, 270±271 neo-Piagetian theories, 272±273 neuropsychology, 276±277 paradigm shifts, 267±270 theoretical frameworks, 277±278 Cross-disciplinary integration, 235 Cross-sectional studies adult development, 309, 313±314, 322 memory, 154±155, 161 Cross-sex avoidance, 211±212 Crowds, 219 Cultural change, accommodation, 278 Cultural psychology, 267, 269, 270, 273, 281

361

Culture, 271±272 conceptualisation, 280 de®nition, 267 ``writing against'', 280 Day care, 86, 88±89 Day Care Environmental Inventory (DCEI), 92 Day Care Quality Assessment Inventory (DQAI), 92 Decalage, 339 Deceptive containers, 272±273 Decision-making, adolescence, 289 Declarative memory, 150, 152 Declarative metacognitive knowledge, 155 Deductive reasoning, 135 Deep-structure goals, 234 Deferred (delayed) imitation, 150±151 Delinquency, 34 Dementia, 56, 58, 313 Denver Developmental Screening test, 4 Depression anxiety and, 51 maternal, 26±27 predisposition, 37 Deprivation, 27±29 Desire-belief psychology, 109 Desire psychology, 109 Desires, 112 Developed countries, nutritional de®ciencies, 65 Development, 2 Developmental cognitive neuroscience, 18 Developmental domains, 191±192 Developmental genetics, 49 Developmental niche, 278 Developmental norms, 4±5 Developmental pathways, 36±38, 252, 334 Developmental psychology, 335 Developmental science, history, 329±331 Developmental trajectories, 332±333, 334 adults, 316 physical aggression, 252±253 Differential effects hypothesis, 33 Differential psychology, 335 Differentiation, 333 Direct perception, 8 Disability, 291

362

Subject index

Disadvantaged families adolescent delinquency, 34 pre-school experience, 86, 87, 88 Discipline techniques, 27 Discontinuity, 333 Discourse, 170±171, 213 Disease, cognitive decline, 317±318 Disequilibration, 11 Disinhibition, 182 Dispositions, 180 Disruptive behaviour, intervention, 255 Dissipative structures, 9 Divorce, 34±35, 39 DNA, 55±56, 58 samples, 58 Domain generality, 124, 138, 339 Domain knowledge, memory, 152, 156±157 Domain-speci®c constraints, 123, 127±128, 138, 139 Domain-speci®c innate constraints, 128±133 Domain-speci®c learning, 124, 130±131 Domain speci®city, 124±125, 339 Domains, 123, 127, 339 Dopamine neurotransmitters, 70 Duke Longitudinal Study, 315 Dyads agentic self, 231 play activity, 212±213 Dynamic systems, 9±11 Dysfunctional families, 34 Early Childhood Classroom Observation Scale (ECCOS), 92 Early Childhood Environment Rating Scale (ECERS/ECERS-R/ECERSE)), 91 Early experience, 24±46 Eco-cultural framework, 277, 278 Ecological models, 277±278, 341 Ecological psychology, 8±9 Effective Pre-school Provision in Northern Ireland (EPPNI), 93±94 Effective Provision of Pre-school Education (EPPE), 93±94 Ego-brittleness, 182, 184 Egocentrism, 103 Ego-control, 181±182, 184, 186 Ego-resilience, 181±182, 186 Elderly advanced old age, 322±323 cognitive decline, 316, 317±321 development studies, 315±316

Embodied cognition, 17 Emergentists, 129 Emerging adulthood, 291 ``Emic'' approach, 269 Emotional stability, 179, 182, 184, 193, 195 Emotions regulation, 173, 289±290 theory of mind, 112 understanding, 168±173 Empathy, 172 Empiricism, 336±337 Encoding strategies, 153, 155 Environmental continuity, 39±40 Environmental factors, 339±340 adult cognition, 311, 313, 318 aggression development, 247, 257 CNS development, 74 see also Genetic-environment interaction Environmental genetics, 52±55 Environmental involvement and nutrition, 73, 74±75 Episodic memory, 150, 151 ``Equi®nality'', 334 Equipotentiality, 234 Ethics, 58±59 Ethnic differences, adolescence, 290±291 Ethnocentrism, 266 Ethnopsychologies, 273 Ethnotheories, 274, 278 Ethological theories, gender development, 214±216 European Network Project (EURONET), 295±296 Event-related potentials, 289 Event-time analysis, 310 Evolution, 130 gender development, 214, 215 Executive functioning, theory of mind, 109±110, 116 Experience plasticity and, 12, 18 theory of mind development, 109 Experimental studies aggression intervention, 254±255 gender labelling, 204±205 memory development, 145±146 motor development, 14 Expert-novice paradigm, 156 Expert opinion, 93 Explicit memory, 150, 151±152 Exploitive efforts, 138, 139 Exploration, 9, 73, 75

Subject index Extraversion, 179, 182, 183, 185, 193, 195 Eyewitness memory, 146, 159±160 Facial anger, 244 Factor analysis, 310, 316 False beliefs, 102, 105, 108, 113, 117, 171 non-Western cultures, 273 Families aggression development, 253±254 cognitive ageing, 318 dysfunctional and disadvantaged, 34, 86, 87, 88 stress response, 35±36 Family Day Care Rating Scales (FDCRS), 91 Fantasy play, 212 Fatty acids, 65, 67, 69, 70 Fear, 183, 185 Feeling-of-knowing, 156 Field theories, 341 Fight±¯ight system, 243±244 Five-factor (big-®ve) model, 179, 182±183, 184, 186±187, 193, 194, 195 Flexibility, 318 Folk psychology, 102, 136, 339 cross-cultural studies, 273±274, 279 Food intake variability, 76 Forgetting, 145±146, 159 Foster care, 291 Fragmented pictures, 157 Frame problem, 273 Free recall, 144, 148 triage model, 158 French-speaking countries, migrant studies, 275 Friendships, 171, 213 Functional genomics, 59 Functional isolation, 73±76 Functional localisation, 18 Functional magnetic resonance imaging, 289 Fuzzy-trace theory, 158

g, 49, 50±51, 52±53, 56 Gait, 4, 10, 13 Gender constancy, 207 Gender development, 202±222 biological factors, 208±209 cognitive theories, 206±208, 209±210 ethological theories, 214±216

363

group behaviour, 210±214 psychobiological theories, 216±217 socialisation theories, 202±206, 208, 209±210 Gender differences action-control beliefs, 228 adolescence, 290±291 personality, 190±191 play behaviour, 212±214 see also Sex differences Gender identity, 207, 210 Gender labelling, 204±205 Gender schema theories, 207 Gender segregation, 212±213 Gender stereotypes, 206±207, 208, 210 Generational differences, 319±320 Genetic correlation, 50 Genetic±environment interaction, 49, 52±55 aggression development, 257 gender development, 216 Genetic factors adult development, 313, 317 cognition, 48, 50, 313, 317 individual differences, 50±51, 172, 317 see also Behavioural genetics Genetic risk prediction, 58 Genie, 32 Genome scans, 57 Genomics, 59 Genotyping, costs, 58 German reuni®cation, 295 Gibsonian psychology, 8 Gist, 158 Goal-directed actions, 226±227 Goal orientation, 193 ``Gold standards'', 322 ``Good enough'' parent, 27 Grand theories, 341±342 Group norms, 219 Group processes, 219 Group success/viability, 215±216 Groups, gender development, 210±214 Growth, dynamic principles, 2±3 Growth curve modelling, 195, 310±311 Guided participation, 269 Head Start, 87 Heritability throughout lifespan, 49 ``Hierarchical integration'', 333

364

Subject index

Hierarchical linear modelling, 93, 231, 297 Hierarchical models, 93 High/Scope Perry Pre-school Project, 87 ``Higher'' cognition, 17 Holistic approach, 180 Home Observation for Measurement of the Environment (HOME), 54 Homosexuality, 291 Hormones, 217 Human Genome Project, 56, 57 Humour, 172 Hunter-gatherers, 276 Hyperactivity, 173 Identity strategies, 275 Idiographic methods, 177, 334 Imagery, 144 Imagination, 114 Imitation, 137 gender development, 205±206 Immune system, nutritional links, 71 Implicit memory, 150, 157±158 Implicit theory, 112, 115, 292 Imprinting, 30 Impulse monitoring, 182 Impulsive undercontrollers, 188, 189±190 Incest, 215 Indigenous language, 271 Indigenous psychologies, 279 Individual differences, 332 adult development, 316, 317±321 genetics, 50±51, 172, 317 mindreading, 167±168, 169, 172 motor development, 14, 18 theory of mind development, 114±115 Individual-in-environment models, 340 Induction issues, 129 Infant/Toddler Environmental Rating Scale (ITERS), 91 Infantile amnesia, 144±145, 151 Information processing adolescence, 288±289 memory, 149, 152, 153, 161 quantitative genetics, 48 Inhibition, 195 Innate constraints, 125, 128±133, 139 Innate knowledge, 139 Inoculation theory, 33 Integrative efforts, 138

Intelligence peak, 314 Intent to harm, 243±244 Intentional beings, 180 Intentionality, 110±111, 267, 268 Interactions gender differences, 213 malnourishment, 75 sociocultural constraints, 136±138 Interculturation, 275 Interdisciplinary cooperation, 235 Interference, 145±146, 148 Internal representations, 40 Interpersonal domain, 192 Interventions aggression development, 254±255 cognitive ageing, 319 genetic disorders, 58 Intrauterine growth retardation, 66, 67 Introversion, 183 Inverted Q-factor analysis, 187±188 Involuntary memory, 148 Iodine de®ciency, 67, 68 Iodine supplementation, 68 Iron absorption, 71 Iron de®ciency, mild, 69 Iron de®ciency anaemia, 64, 65, 67, 68, 70, 73, 75±76 Iron supplementation, 68 Irreversibility, 27±29, 31 Irritability, 184 Isolation, functional, 73±76 Item response theory, 322 Jimmy and Johnny, 3 Joint attention, 137 Kauai, 34 Kicking, 7 Knowledge, 13 acquisition, 126±128 innate, 139 prior knowledge, 133, 139, 148, 156 theory of mind, 113 Language acquisition device, 132 Language development, 32, 331 mindreading development, 171 quality of day care, 88, 90 Late-onset Alzheimer's dementia, 56, 58 Lead absorption, 71 Learnability, 129

Subject index Learning constrained, 137 domain-speci®c, 124, 130±131 mechanical, 147, 148 plasticity and, 18 verbal, 148 Learning theory, 341 personality, 178 Leg movements, spontaneous, 7, 13 Levels-of-learning, 159 Lexicon, 178±179 Life satisfaction, 319 Lifespan development, 334±336 adult development, 315 cross-cultural studies, 276 Linkage approaches, 56±57 Localisation of function, 18 Logical memory, 147, 148 Longitudinal studies adult development, 307, 310, 313±316, 322 aggression, 251, 253±254 child care, 88, 93±94, 95 early experiences, 36±38 genetic analysis, 49, 52±53 memory, 154±155, 161 mindreading, 169, 173 motor development, 12±13 parental behaviour and emotion regulation, 290 personality development, 178, 193, 195 transition to adulthood, 291±292 Long-term memory, 144, 145±146, 150±151, 159 Long-term store, 150 Maladjustment, 34 Malnutrition, 64, 66, 67, 68, 75 Maternal characteristics and child care association, 89±90 Maternal depression, 26±27 Maturational theories, 2±3, 341 information processing speed, 153 Means±end analysis, 124, 228±230 Measurement of change, 309 Mechanical learning, 147, 148 Mechanistic models, 223, 270, 336 Mediation de®ciency, 154 Mediators, 236 Memories/memorising distinction, 161 Memory decay, 153 deceptions, 146

365

movement and, 17 multistore models, 150 span, 145, 152±153 strategies, 152, 153±155 strength, 147 trace, 144, 158 Memory development, 143±166 basic capacities, 152±153 domain knowledge, 152, 156±157 general course, 146±147 history, 143±149 memory strategies, 152, 153±155 metacognitive knowledge, 152, 155±156 nonstrategic factors, 157 social context, 151 Mental models, 40 Mental obedience, 146 Mental representations, 112±113 Mental world, 102±122 history, 103±107 theories, 107±115 Metacognition cross-cultural studies, 274±275 de®nition, 104 schooling and 274±275 theories of mind and, 106±107 Metacognitive knowledge, 152, 155±156 Metamemory, 104, 155±156 Metarepresentational, 106 Methodology adolescence research, 296±298 adult development, 308±311 cross-cultural studies, 268, 270±271 motor development studies, 3±4 Micronutrient de®cits, 68, 69 Migrant studies, 275 Mind-mindedness, 171 Mindreading, 167±176 cultural variation, 168 emotion and, 168±170 individual differences, 167±168, 169, 172 language and, 171 Mineral de®ciencies, 67, 68, 69, 71 Minnesota Multiphasic Personality Inventory, 179 Model-building, 334 Model ®tting, 233 Modelling agentic self in context, 232±233 motor development, 17 see also Social learning theory

366

Subject index

Modularity, 169 conceptual development, 124 theory of mind, 109 Molecular genetics, 55±59 Monitoring, 156, 207 Moral agency, 193 Mothers characteristics and child care association, 89±90 depression, 26±27 historical role in developmental research, 330 Motor development, 1±23 biology, 2±3 cognition and, 16, 17 coordination and control, 12±16 cultural differences, 18 developmental norms, 4±5 dynamic systems, 9±11 ecological psychology, 8±9 experimental studies, 14 individual differences, 14, 18 methodology, 3±4 modelling, 17 neural bases, 18 plasticity, 11±12, 18 technological advances, 12 Movements analysis, 12 patterns, 10 recording, 3±4 as synergies, 6, 9±10 Moving room paradigm, 15 Multilevel modelling, 93, 94, 310±311 Multimodal perception, 9, 17 Multiple developmental pathways, 334 Multistore memory models, 150 Multivariate genetics, 49±51, 54 Mx program, 233 Myelination, 70 Naive biology, 129, 131, 132, 134±136 Naive physics, 129, 133, 136 Naive psychology, 129, 132±133, 136, 339 Naive theory of mind, 102 Nativism, 129, 336±337 Natural disasters, 35±36 Natural selection, 214 Naturalistic methods, 2, 271 Neonates as a lump of clay, 24, 25 memory, 150 stepping re¯ex, 7±8, 14

NEO-PI, 179 Neo-Piagetian theories, 272±273 Neural maturation, 2±3 Neural networks, 11 Neuroimaging adolescence, 289 genetics, 48, 51 Neurological changes, adolescence, 289 Neuronal group selection, 12 Neuropsychology, cross-cultural studies, 276±277 Neuroscience approach to memory, 149, 162 NICHD, 89±91, 92 Nomothetic methods, 177, 334 Nonshared environment, 52±53 Nonshared Environment and Adolescent Development (NEAD), 53 Normality, 51 Normative change, 332 Novelty seeking, 182 Nutrient-nutrient interactions, 71 Nutritional de®ciencies, 64±84 Nutritional supplementation, 68, 72 Obesity, 65±66 Objectivity, 271 Observation du Lieu de Vie de l'Enfant (OLIVE), 92 Observation of Speci®c Children and Caregivers (ORCE), 92 Observational learning, gender development, 206 Omnipotentiality, 234 On-line, 10 Openness, 179, 182, 184, 185, 193, 195 Optimisation model, 158 Oregon Social Learning Center, 253 Organisational strategies, 153±154 Organismic models, 223, 224±230, 236±237, 336 Organismic speci®city, 26 ``Orthogenetic principle'', 333 Other-sex avoidance, 211±212 Outcome variability, 34±36 Overcontrollers, 182, 184, 186, 187, 188 subtypes, 188, 189 Paradigm shifts, 267±270, 308±311 Parent±child relations memory development, 151 nutritional links, 70±71, 75±76

Subject index Parental practices, 25±27 aggression development, 254 emotion regulation, 290 sex-typing, 203, 204 Parenting skills aggression regulation, 255 childhood experience link, 37 Parents belief systems, 273±274 ethno-theories, 274 ``good enough'', 27 loss of, 37 suicide, 35 Past, understanding, 171±172 Patterns adaptation, 333 dynamic systems, 10 Peer experiences, 53 Perception and action, 8±9, 15±16 multimodal, 9, 17 Perceptual speed, cognitive ageing, 319 Perceptuomotor skills, 32, 33 Performance, 192 Person-centred approaches, 339 adolescence, 296±297 aggressive behaviour, 251 personality, 180, 196 Person±situation interactions, 178 Personality agents, 180 assessment, 179 big-®ve factors, 179, 182±183, 184, 186±187, 193, 194, 195 consistency, 178, 181 continuity, 181 gender differences, 190±191 problem solving, 181±183, 192±196 stability, 186 subtypes, 188±190 traits, 178±179, 180 types, 178, 180, 186±190 Personality development, 177±201 agent-centred approach, 181±186 developmental outcome, 191±196 history, 177±181 types, 186±190 Perspective taking, 103±104, 108 Phenylketonuria, 58 Physical aggression, 249±250 developmental trajectories, 252±253 regulation, 255, 257 stability, 251

367

Planning skills, cross-cultural studies, 275 Plasticity, 11±12, 18 Play gender differences, 212±214 nutrition links, 75 pretend play, 113±114 Policy development, 298±300 Positivism, 267, 268 Postnatal malnutrition, 66, 67, 68 Post-positivism, 268 Postural control, 14, 15 Preference-for-novelty, 150 Preferences, 129, 130 Prehension, 15 Prematurity, 18 Prenatal malnutrition, 66, 67 Pre-school experience, 85±101 care-orientated, 86, 88±89 cost-effectiveness, 86±87 developmental effects, 86±91 educationally-oriented, 86±88 quality of care, 85, 86, 88±89, 90±91 quality of care measures, 91±95 quantity of care, 89, 90, 91 Pretence, 113±114 Prior knowledge, 133, 139, 148, 156 Proactive interference, 146 Problem solving domain-speci®city, 124±125 personality and, 181±183, 192±196 temperament, 183±184 Procedural memory, 150 Procedural metacognitive knowledge, 155 Process accounts, 11 Process characteristics, 91 Production de®ciency, 154 Programme development, 298±300 Proprioception, 15 Prosocial aggression, 251, 257 Prosocial behaviour, 193, 195 Prospective memory, adolescence, 289 Protein±calorie±macronutrient de®cits, 69 Protein-energy malnutrition, 66 Prototypes, 134 Psychoanalytic theory, 25, 330±331, 335, 341 Psychobiological theories, gender development, 216±217 Psycho-cultural models, 278

368

Subject index

Psychodynamic theory, 203 Psychological factors, cognitive ageing, 318±319 Psychological test performance, 276 Psychopathology comorbidity, 51 genetics, 48, 51 Psychosexual development, 203 Psychosocial mechanisms, nutritional de®cits, 70±71, 72±73 Psychosocial stimulation, 72 Q-factor analysis, 187±188 Qualitative research adolescence, 297±298 childcare, 93±94 cross-cultural studies, 271 Quality of care, 85, 86, 88±89, 90±91 measures, 91±95 Quality of Day Care Environment Scale (QDCE), 92 Quantitative approach, 270 Quantitative genetics, 48 , 55 Quantitative trait locus (QTL), 55, 56±57, 59 Random coef®cient models, 231 Reaching, 13, 14, 15 Reactivity, 183±184, 185 Reading disability, 56 ``Real world'' studies, 271 Reciprocal effects, direct versus indirect, 236 Recognition, 144, 150 Rehearsal rate, 153 strategies, 153±154 Relationships emotion and social understanding, 170±173 gender development, 211±214 personality factors, 195 Research mothers' role in promoting, 330 policy/programme relevance, 298±300 regional disparities, 279 teams and networks, 235 Resilients, 182, 186±187, 188 adolescence, 294 subtypes, 188, 189 Resource control theory, 234 Response speed, cognitive ageing, 319

Restructuring, 126 Retrieval strategies, 153, 155 Reversibility, 28±29, 32 Right frame of mind, 280 Rigidity, 318 Robotics, 17 Romanian orphans, 29, 32, 33 Safari-style cross-cultural research, 271 Salamander, 2±3 Scaffolding, 269 School achievement tests, 50 Schooling and metacognition, 274±275 Scienti®c method, 329±330 Scripts, 151, 206±207 Seattle Longitudinal Study, 314±315 Selective imitation, gender development, 206 Self, 192, 224, 225 Self-concept, 40 Self-con®dence, 172 Self-esteem, 172, 195 Self-evaluative feedback, 225 Self-monitoring, 156, 207 Self-organising systems, 9 Self-regulation, 183±184, 185, 224 Semantic memory, 150 Sensation seeking, 182 Sensitive period, 30 Sensorimotor intelligence, 32 Separation, 31, 36, 39 Sex differences aggression, 244±245 brain structure and function, 209 see also Gender differences Sex-typing, 202 child-rearing practices, 203, 204 Short-term memory, 145 Simulation theory, 109 Situated cognition, 139 Situational compliance, 193 Skeletal core principles, 134 Small for gestational age, 66 Social activism, 335 Social adjustment, 88 Social behaviour, 331±332 Social beings, 180 Social context, memory development, 151 Social cues, 254 Social development, 88

Subject index Social experiences, theory of mind development, 114±115 Social forces, 334±335 Social-information-processing models, 254 Social learning theory aggression development, 247±248, 253 gender development, 205±206 personality development, 178 Social moulding, 337±338 Social referencing, 111 Social skills, 115 Social understanding, 168±173 Socialisation, 24, 267, 337±338 gender development, 202±206, 208, 209±210 Sociocultural constraints, 128, 133±138 Soothability, 184, 185 Spatial ability, 276 Stability, 185±186, 250±253, see also Emotional stability Stage theories, 338 Statistical modelling, 224 Stepping re¯ex, 7±8, 14 Stimulus bound recognition memory, 144 Stimulus±response principles, 202 Strategic memory, 152, 153±155 Strategy beliefs, 227 Stress resistance, 34±36 Structural characteristics, 91 Structural equation modelling, 224, 233, 236, 310 Structural models, 179±180 Substance use, 291±292 Suggestibility, 146, 160 Suicide, parental, 35 Surface-structure goals, 234 Surprising objects, 272±273 Survival analysis, 310 Synergies, 6, 9±10 Systems approach to agency, 225±226 Target Child Observation, 92 Task persistence, 184 Teasing, 226 Temperament CNS structures, 70 emotion regulation, 290 genetic factors, 172 parental practices, 26±27

369

personality, 185±186, 196 problem solving, 183±184 stability, 185±186 stress resistance, 34±36 vitamin B intake, 69 ``Temperamental'' types, 188 Testosterone, 246 Text recall, 149 Theoretical frameworks adult development, 311±313 cross-cultural psychology, 277±278 Theory change, 126 Theory of mind, 104±109 cross-cultural studies, 115, 272±273 development, 110±115, 102 naive, 102 theories, 109±110 Theory theory, 109 Thinking, 114 Time adolescent use, 296 monitoring in adolescence, 289 understanding of, 146 Timing, 38±40 Trait theories, 178±179, 180 Traumatic events, 35±36 Turning points, 38 Typologies, 231±232 Undercontrollers, 182, 184, 186, 187, 188 subtypes, 188, 189±190 Undernutrition, 64, 67, 69, 70±71, 75 Unexpected transfer method, 105 Universality, 272 Utilisation de®ciencies, 154 Variable approach, aggressive behaviour, 251 Verbal aggression, 249 Verbal learning, 148 Verbal memory, 144, 161 Victimisation, 226 Video-assisted movement analysis, 12 Vision and postural control, 15 Visual ¯ow, 15 Visual memory, 144 Visual perception, 111±112 Vitamin de®ciencies, 67, 68, 69 A, 64 B, 69 B6, 68, 71 B12, 65

370

Subject index

Voluntary memory, 148±149 Vulnerability±resilience continuum, 34 Vulnerable overcontrollers, 188, 189

Walking, 3±4, 7, 13, 14 Working memory, 152 Zinc de®ciency, 68, 69, 73