The Language of Stories
How do we read stories? How do they engage our minds and create meaning? Are they a mental con...
153 downloads
1742 Views
1MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
The Language of Stories
How do we read stories? How do they engage our minds and create meaning? Are they a mental construct, a linguistic one, or a cultural one? What is the difference between real stories and fictional ones? This book addresses such questions by describing the conceptual and linguistic underpinnings of narrative interpretation. Barbara Dancygier discusses literary texts as linguistic artifacts, describing the processes which drive the emergence of literary meaning. If a text means something to someone, she argues, there have to be linguistic phenomena that make it possible. Drawing on blending theory and construction grammar, the book focuses its linguistic lens on the concepts of the narrator and the story, and defines narrative viewpoint in a new way. Textual use of pronouns and other referential expressions also comes under investigation, including their occurrence in constructions of speech and thought representation. The examples come from a wide spectrum of texts, primarily novels and drama, by authors such as William Shakespeare, Margaret Atwood, Philip Roth, Dave Eggers, Jan Potocki, and Mikhail Bulgakov. Barbara Dancygier is Professor in the Department of English at the University of British Columbia, Canada. Her publications include Conditionals and Prediction (Cambridge, 1998) and Mental Spaces in Grammar (with Eve Sweetser, Cambridge, 2005).
The Language of Stories A Cognitive Approach Barbara Dancygier
cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Tokyo, Mexico City Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9781107005822 © Barbara Dancygier 2012 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2012 Printed in the United Kingdom at the University Press, Cambridge A catalog record for this publication is available from the British Library Library of Congress Cataloging in Publication data Dancygier, Barbara. The language of stories : a cognitive approach / Barbara Dancygier. p. cm. Includes bibliographical references and index. ISBN 978-1-107-00582-2 1. Language and languages – Style – Psychological aspects. 2. Cognitive grammar. 3. Discourse analysis, Literary. I. Title. P301.5.P75D36 2011 8080 .036–dc22 2011015550 ISBN 978-1-107-00582-2 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.
for Jacek and Szymek
Contents
List of figures Acknowledgements
page ix x
Introduction
1
1
4
Language and literary narratives 1.1 1.2 1.3 1.4 1.5 1.6 1.7
2
Blending, narrative spaces, and the emergent story 2.1 2.2 2.3 2.4
3
4
Where does narrative meaning come from? Literary analysis and linguistic analysis Literature, language, and human nature Literary texts and communication Why is fiction special? Narrative and grounding Approaching narratives
Applying blending to fictional narratives Narrative spaces as mental spaces Narrative spaces – an example Emergent story
6 8 11 16 21 23 29
31 32 35 40 53
Stories and their tellers
58
3.1 3.2 3.3 3.4 3.5 3.6 3.7
Narrators, narrative spaces, and viewpoint Types of teller and epistemic viewpoint Second-person narratives The teller, the author, and the character Multiple tellers Narrative space embedding Narrative viewpoint and narrative spaces
60 64 76 79 80 84 86
Viewpoint: representation and compression
87
4.1 4.2 4.3 4.4 4.5 4.6 4.7
91 96 100 102 106 108 112
Viewpoint and representation Viewpoint compression Decompression for viewpoint Fictive vision, causation, and change The micro level, the macro level, and viewpoint compression Speech, thought, and multiple levels of representation Narrative thought and intersubjectivity
viii
5
Contents
Referential expressions and narrative spaces 5.1 5.2 5.3 5.4 5.5 5.6
6
Fictional minds and embodiment in drama and fiction 6.1 6.2 6.3 6.4 6.5
7
Deictic ground in literary discourse Mental spaces, physical spaces, and dramatic narratives Materiality of the stage and fictional minds From dramatic narratives to novelistic narratives Fictional minds, bodies, and brains
Speech and thought in the narrative 7.1 7.2 7.3 7.4
8
Compression, decompression, and cross-space mappings Proper names, frame metonymy, and the status of a character Role-value mappings as cross-space connectors Common nouns Personal pronouns, viewpoint, and the narrator Deictic I and the construal of subjectivity
Types of discourse spaces in the narrative Speaking for thinking Levels of embedding in thought representation Viewpoint compression and constructional compositionality
Stories in the mind 8.1 8.2 8.3 8.4
The linguistics of literature The storyworld reality Blending and narrative analysis A bridge to the truth
Notes to the text References Literary works cited Index
117 118 119 121 128 129 136
139 139 141 146 164 168
171 172 174 178 183
195 195 200 201 203
205 210 223 225
Figures
2.1 2.2 2.3 3.1 3.2 3.3 3.4 3.5 3.6 4.1 4.2 4.3 5.1 6.1 6.2 6.3 7.1 7.2 7.3 7.4
Narrative spaces (example 2) Narrative spaces in The Blind Assassin Referential connections across spaces in The Blind Assassin Intrusive narrator (example 2) Ego-viewpoints Present time and present tense (example 7) Ego-viewpoint in the present tense (example 8) Second-person narrative (examples 11 and 12) Second-person narrative (examples 13–15) Reported voices and viewpoint (example 1) Representation and compression 1 (example 11) Representation and compression 2 (example 11) Narrative spaces and roles in American Pastoral The mental space schema of a theatrical performance Acting/Speaking blend Feeling/Speaking blend Discourse blend (examples 4 and 5) The ‘crow’ blend (examples 9, 10, and 11) Third-person report (example 15) First-person report (example 18)
page 39 41 52 65 68 71 72 77 78 91 98 99 127 145 156 157 176 180 184 188
ix
Acknowledgements
A project like this one requires that the author finds friends in unexpected places, and I was lucky to meet many, and usually when I needed them most. I was writing this book feeling cautious about what various groups of people might say about it, and they invariably gave me their support and encouragement. I thus want to thank all of those who could have dampened my enthusiasm and didn’t, and who gave me good advice instead. The list is too long to quote here. I owe special thanks to colleagues who read major fragments of earlier drafts and helped with very specific responses: Uri Margolin, Eve Sweetser, Sean McAlister, Patsy Badir, Lieven Vandelanotte, Margery Fee, Suzanne Fagel, Maarten van Leeuwen, and my Cambridge University Press reviewers. Their careful consideration of my ideas and structural choices has led me through many hours of doubt. I owe much to the scholarship and inspiration of Mark Turner, Gilles Fauconnier, and Eve Sweetser. I also want to thank David Herman, who encouraged me and trusted in the final outcome before I had the courage to call this project a book, and Alan Palmer, who listened and shared thoughts when I needed it badly. I am grateful to Elena Semino and Paul Simpson for their sustained support of my various projects and their interest in my work. One of the chapters of this book could never have come into existence if not for the inspiration and friendship of Michael Booth and Amy Cook. I have also been lucky to know Lieven Vandelanotte, whose work, guidance, and support were with me all along. I have been given much support by the literature and linguistics colleagues in my own department – the Department of English at the University of British Columbia. This book was conceived against the background of their vast knowledge and wonderful open-mindedness, and I am very grateful to them all for their encouragement, friendship, and sense of humor. My graduate students have been my day-to-day audience, and have sat through many hours of workshops in which we discussed their ideas and mine. They have asked excellent questions, and have helped me focus when I was slipping. I can only mention some of them here: Jennifer Schnepf, Sean McAlister, Mike x
Acknowledgements
xi
Borkent, Slade Stolar, Tiffany Johnstone, Janet Lermitte, Robin Steen, Eve Preus, and Mark Deggan. Last, but not least, my family has helped me in more ways than I can count. My husband, Jacek, supported me with his impeccable literary taste and his boundless computer expertise – both were indispensable in the completion of this book. I also want to thank him for the motto to Chapter 8. My son, Szymek, infected me with his admiration for Melville and Poe, and also his passion for traditional Japanese theatre. Many ideas in this book owe much to his careful reading and willingness to share thoughts. This project has been supported financially with a grant from the Social Sciences and Humanities Research Council of Canada, number 410-20051644. I am grateful to SSHRC officers for making the funding available and providing prompt answers to all my questions. Last but not least, I want to thank the editorial staff at Cambridge University Press, and especially my editor, Andrew Winnard. His professional insight helped me see the big picture and guided me through various rapids. Finally, I owe much to the writers of all the brilliant stories I got a chance to read again and again. Thanks to them, writing this book was an exciting experience. It would not be fair to say that all these great novels were just excellent materials to work on, as on occasion I had to leave my analyst persona aside, and let the text speak to me. The stories just wouldn’t let me do anything but read. I am truly grateful for these moments. I have no doubt that the book has many faults, which my critics will be quick to point out. But I am sure I would have made many more mistakes if not for all the people who were willing to listen, read, and talk. I am deeply grateful to all of them.
Excerpts from American Pastoral by Philip Roth. Copyright © 1997 by Philip Roth. Used by permission of Houghton Mifflin Harcourt Publishing Company. All rights reserved. Excerpts from The Human Stain by Philip Roth. Copyright © 2000 by Philip Roth. Used by permission of Houghton Mifflin Harcourt Publishing Company. All rights reserved. Excerpts from The Blind Assassin by Margaret Atwood are reproduced with permission of Curtis Brown Group Ltd., London on behalf of Margaret Atwood. Copyright © Margaret Atwood 2000 Details of all the literary texts referred to are given in the ‘Literary works cited’ section of this book.
Introduction
The relationship between language and literature is a contentious issue. On the one hand, it may simply be described as a relationship between raw material and a finished product – language provides the basis on which creative and unique works of literature emerge. On the other hand, once we look at meaning, the dividing lines begin to fade – it is difficult to define a sharp boundary separating the meaning of literary works and the meaning of other texts. One way of downplaying the obvious links is to claim that fiction engages knowledge in a much broader and much more culturally specific way than everyday use of language does. But that would be an exaggeration. One could not follow an ordinary discussion of, say, climate change if one did not have any prior knowledge of the issue. Terms such as global warming or carbon neutral mean something only in the context of the debate and the science underlying it. It is also not true that literature is different simply because it is not intended to describe facts which have actually occurred. Linguistic constructions such as conditionals do that too – they set up imaginary scenarios and explore their consequences. The dividing line is either too fine to see, or shifts too much, depending on the assumptions. In this book I attempt not to debate those differences, but instead to talk about literary texts as linguistic artifacts, focusing on the description of language processes which participate in the emergence of literary meaning. My point of departure is the belief that if a text means something to someone, there have to be linguistic phenomena that make that possible. I found much inspiration and help in two cognitive theories of language: construction grammar, with its interest in the interaction between lexis and grammar and its trust in the meaningfulness of syntactic choices, and conceptual integration, with its focus on the mechanism of the emergence of meaning, regardless of the means that lead to the initiation of the process. Both theories have been applied to various works of fiction. However, the goal of this book is not to discuss specific interpretations of specific texts in terms of blending and construction grammar, but rather to explain how such interpretations might be arrived at. In other words, this book explores the processes of meaning construction which yield (all the varied) interpretations 1
2
Introduction
of literary texts. Crucially, I attempt to start with sentence-level phenomena and study the ways in which they contribute to text-wide meanings, relying on blending through all the levels. This project is thus an attempt to propose a blending interpretation of fictional storytelling, along with the linguistic strategies it involves. In order to talk about linguistics and literature, I had to accept the necessity of zooming in and out all the time. Some of the fragments in this book are closer to linguistics, especially when the analysis focuses on specific sentence-level choices, while other parts discuss the structure of entire novels, with limited attention to the sentence-level language they rely on. Such shifts of focus may seem somewhat distracting, but I also hope they will keep various kinds of readers engaged. This book is addressed to cognitive linguists, stylisticians, narratologists and literary critics alike, though these groups will surely find different parts of it relevant to a different degree. It also attempts to facilitate a conversation across these divisions. The structure of the book also reflects its multiple foci. Chapter 1 discusses recent work which sheds some light on the role of narratives in our understanding of language, literature, and culture; it contextualizes the ensuing discussion against a broader range of research. Chapter 2 briefly introduces blending theory and constructs some of the main tools of my methodology. Among others, the chapter discusses primary mechanisms of story construction and explains my understanding of the story as a conceptual construct. In Chapter 3, I propose the analysis of the concept of a ‘narrator’ in terms of epistemic viewpoint. Going against the tradition which assumes that every story needs a teller, I discuss ‘narratorship’ as a cluster of epistemic and linguistic choices which organize the text-wide viewpoint structure of the story. Chapter 4 investigates viewpoint configurations at lower levels of narrative structure, and argues for an understanding of narrative discourse as directed primarily at prompting conceptualizations of various kinds. In this chapter I also introduce the primary mechanism of story emergence – viewpoint compression. Then, Chapter 5 discusses the story-constructing role of referential expressions, especially pronouns and role descriptors. Chapter 6 briefly introduces narrative underpinnings of the discourse of theatre, arguing for tight conceptual connections between the text of the play, the space of the theatre, and the bodies of actors. I also show how the cognitive specificity of the theatre set-up affects linguistic choices, especially the emergence of specialized constructions. In the remainder of the chapter I attempt to trace the route which leads from on-stage narratives, where all the modalities contribute to the emergence of the story, to novelistic, or on-page, narratives, relying exclusively on linguistic forms. In this context, I also discuss the primary importance of the forms through which fictional minds are represented in narratives. In Chapter 7, I continue the discussion of the ways in which
Introduction
3
representing characters’ thoughts and emotions (not just words) lies at the core of narrative discourse. I give a brief overview of the constructions used to represent conceptualizations held by various story participants, to conclude that such textual choices are subordinate to the overall mechanisms of viewpoint compression. Finally, Chapter 8 provides some closure to the main themes of the book. While there may be only limited explicit attention given to the issue, the ensuing text also argues for a mode of analysis which treats the goals of literary, linguistic, and cognitive studies as possibly overlapping or even shared. There was much uncharted territory and a vast volume of research to contend with, but some of the central questions have emerged clearly. How does the mind negotiate the divide between everyday discourse and fiction? How does our creativity work in these two contexts? What assumptions about the nature of mind and language yield results potentially relevant to the humanities? What exactly can literary work contribute to the discussion? There are clearly more such questions than any single book can raise. This book attempts to identify and discuss the few that seemed important.
1
Language and literary narratives
There is nothing so strange it cannot be true, and no story so unlikely it cannot be told. No story is a lie, for a tale is a bridge that leads to the truth. (The Arabian Nights, retold by Neil Philip)
This book is about how the centuries-old thoughts quoted above are indeed true. Why are stories not lies, even though they don’t tell the truth? How do they help us to learn from our experience and the experience of others? And how does language support the meaning of stories? The structure of that “bridge … to the truth” is what I will try to understand. The human ability, or even desire, to tell, understand, watch, and create stories has engaged a number of disciplines, each of which poses a different set of questions about the core of the phenomenon. Why do we enjoy stories? What’s in it for us as a species? Could our culture exist without stories? Are they a mental construct, a linguistic construct, or a cultural construct? Is there a difference between real stories and fictional stories? These are just some of the questions of interest to anthropologists, psychologists, narratologists, philosophers, linguists, and literary scholars. The answers have been, by necessity, partial, and directed at the interests of the disciplines they emerge from, but it is becoming increasingly clear to all concerned that some cross-disciplinary dialogue is necessary. This book is an attempt to bring together at least some of the questions out there while focusing on one central aspect of storytelling: how do stories construct meaning? The answer to a broad question like this requires making some assumptions about the general mechanisms of meaning construction and about the correlations between formal choices and the meanings prompted. This approach would naturally define the project as a basically linguistic or stylistic one, if not for the fact that while meaning construction is a process prompted by forms (linguistic, visual, artistic, and otherwise), it is in fact a cognitive process, involving various faculties of the mind. Elucidating the emergence of meaning in various storytelling contexts thus requires adopting a methodology which is sensitive enough to fine-grained linguistic details, while opening itself up to questions about cultural context, our cognitive capacities, and 4
Language and literary narratives
5
the simple fact of there being various ways to interpret any storytelling artifact. The very term ‘meaning construction’ requires a brief explanation, so as to be clearly distinguished from literary and anthropological work reliant on ‘social constructivism.’ Throughout this book I will be using ‘meaning construction’ to refer to the processes which yield meanings of language expressions based on the frames evoked, constructional meaning patterns, and, among others, blending. Meaning construction in this sense relies on the specific word and grammar choices. Even in the simplest case such as a choice of I will do it, which commits the speaker, versus I would do it, which does not commit her at all, the construal of the former expression as a promise and of the latter as advice relies on meaning-construction processes crucially dependent on the local language choices. The construction depends on all the linguistic forms used, from the choice of the first person I (rather than third person he, which would not commit the speaker in any way), and of the modal will (rather than might, which would not involve a commitment), to the choice of tense. The present tense form will prompts a more committed meaning than the past form would. Also, while a social constructivist reading of a text is a matter of a critic uncovering frames possibly (but not necessarily) existing in the writer’s mind, the framing of past tense is a different phenomenon. Past tense forms are commonly used to mark epistemic or emotional distance, and thus are often used for politeness (so that Could you call me tomorrow? is more polite than Can you call me tomorrow?). The framing of the past tense such that it marks the speaker as less committed is a matter of standard language use, independent of any particular language user.1 The methodological framework to be used in this book relies primarily on two theories – mental spaces theory and its extension known as blending, or conceptual integration.2 Both theories assume that meaning is not a set of discrete conceptual packets neatly correlated with linguistic forms, especially lexical forms, but that in each case it emerges as a result of one’s use of formal signals (verbal or not). It also relies on processes which economically use the available complexity of possible aspects of meaning to yield an interpretation which optimally fits the needs of ongoing discourse. In actual discourse, though, interpretations are often negotiated rather than simply communicated, and thus any account of meaning needs also to consider signals built into the form being communicated which address the argumentative and inferential aspects of meaning construction. The fact that such mechanisms take the form of specialized grammatical constructions (cf. Verhagen 2005, 2008; Dancygier 2008c) should further convince us of the centrality of meaning negotiation to any meaning-emergence processes. As this book argues, understanding narratives relies on very similar processes: emergence, construction, and negotiation of meaning through specific language choices. The choices may be found at the lowest level of linguistic structure (such as using the first or the third person in
6
Language and literary narratives
the representation of narrative voice) or they can pertain to text-wide decisions on narrative structure (whether the story is presented chronologically or in a fragmented manner), but they are language choices readers rely on in the construction of narrative meaning. 1.1
Where does narrative meaning come from?
Meaning is often talked about in terms which seek to divorce it from the situation in which it emerges, and to propose formal constraints on what counts as a meaningful utterance. In the context of stories, this approach would not yield any understanding of their meaning, simply because stories do not open themselves to this kind of analysis. The same can be said about attempts, especially in structuralist narratology, to arrive at clear taxonomies and terminological distinctions. These taxonomies range over a very broad spectrum of phenomena, starting from representations of time flow to a typology of ‘narrators.’ While this kind of work has alerted linguists and stylisticians to phenomena which can now conveniently be talked about as ‘flashbacks’ or ‘intrusive narrators,’ it has not explained adequately why these forms would be used and how they are understood. Trivializing things a little, one could ask why flashbacks are so common in novels or movies (and, yes, also in natural conversations and oral stories), if most narratologists claim at the same time that the story is (roughly) a temporal sequence of linked events? And is a narratorial voice which (who?) addresses the reader directly just a frill, a naïve stage in the development of narrative form, or a charming attempt to make written fiction feel like a conversation? Admittedly, recent work in cognitive narratology (cf. Herman 2002, 2003a), Fludernik’s natural narratology (1997, 2003) and stylistics3 marks a move away from taxonomies and in the direction of cognitive explanations, but there is still much to say about the connection between the lowest-level language choices and the narrative, and, primarily, about how this connection is mediated by cognitive mechanisms involved in the emergence of meaning. It seems to be an accepted view in many disciplines (including some linguistic and cognitive theories) that language is a separate module of the mind and thus cannot be considered in the context of art, literature, or culture. The fallacy is compounded by a common belief that linguistic analyses are too formal and too focused on representing structural patterns to be useful in an enquiry into meaning in any context broader than a sentence. While this is true in many cases, it is not always true. The linguistic models to be relied on in this book, within the broadly understood discipline of cognitive linguistics, assume that meaning construction is a process which affects numerous levels, and that the overall meaning does not arise as a result of simple addition or compounding of the lower-level constructs. On the contrary, meaning-construction processes
1.1 Where does narrative meaning come from?
7
invariably involve selection and narrowing on the one hand, and emergence of new configurations on the other. There is a significant range of linguistic work using mental spaces and blending to account for meanings of nominal compounds and sentential or phrasal-level constructions,4 and such analyses typically involve uncovering the composite constructs at the lower levels. Research has also been done on the emergence of meaningful blends in discourse and conversation (cf. papers in Oakley and Hougaard 2008; Pascual 2002, 2006a, 2006b, 2008). What the analysis proposed in this book aims at is describing the consequences of lower-level linguistic choices at the higher levels of narrative discourse. While there are some limits to treating the whole text as a linguistic artifact, the focus on processes rather than forms opens longer and more complex stretches of discourse to linguistic analysis. Thus the intended contribution of the analyses proposed in this book is to offer a view of meaningconstruction processes in multidimensional artifacts such as novels or plays. The processes of selection, narrowing, and emergence of new meanings are complex and depend on more than the understanding of specific language items. As Deacon argues (2006), blending is an interpretive process which relies on our generic symbol-processing and referential capacities. That is, the ability to use linguistic symbols in new combinations, to juxtapose them or compose them into new configurations opens our capacity for coining new meanings to nearly limitless possibilities. One might read this to mean that the potential for our linguistic abilities to be used in the construction of new meanings is not inherently limited and relies on the fact that manipulation of symbolic concepts is the daily bread of our cognition. Indeed, even a very simple expression can illustrate that potential. For example, the expression emotional roller-coaster cannot be explained without an analysis of the meaning of the component concepts, especially roller-coaster. It is a fully conventionalized expression which refers to a machine found in amusement parks which takes people on a seemingly dangerous ride involving abrupt ups and downs, lack of control of one’s movements, as well as high speed. While any such experience in real life would be unwelcome, it is treated as ‘fun’ in the context of an amusement park. The addition of the adjective emotional changes the meaning significantly. There is no longer any physical object involved, there is nothing suggesting entertainment, or even an experiencing person’s conscious decision to be engaged in the activity. An emotional roller-coaster evokes negative (rather than amused) emotional response parallel to what roller-coasters do (jerking people up and down at a breakneck speed); it applies to a person’s psychological state, unpleasantly and abruptly altering between elation and depression. The composition process of this blend creates a concept not available in the component expressions, and, even more important, evokes an emotional response opposite to the original. This is also one of the points Deacon makes when he stresses the fact that emergent reference may involve conflicting emotions.
8
Language and literary narratives
The brief discussion above does not do justice to the complexity of the blend described, even though it remains at a relatively low level of complexity, which is the lexicon. And yet, the analyses throughout this book will attempt to look at narrative structure in its entirety through the lens of such low-level choices. While longer texts will be the target, sentence-level phenomena such as pronominal choice, the use of tense, the impact of referential expressions, or representation of discourse will be considered as the lowest level of meaning construction. In general, the assumption of this book (and other cognitive linguistic texts) is that grammar is as meaningful as lexis, and thus grammatical choices impact the meaning of texts. That is, a text’s reliance on interior monologue or on direct discourse is a stylistic choice which, via the interpretation of the specific constructions, impacts the way the narrative is understood. This analysis will also go a step further in arguing that a use of a specific form, associated with a meaningful construction, may ‘infect’ other aspects of the expression and carry constructional meaning while only one bit of the construction is present. The linguistic process whereby chunks of constructions spread constructional meaning has been described as the mechanism of constructional compositionality (Dancygier and Sweetser 2005; Dancygier 2005b). Examples are common. The simplest instance I can recall is an advertisement for Schweppes tonic in the London underground, many years ago. It consisted primarily of the letters Sch . . ., running on the wall along the tracks. The combination of letters, unusual in English, brought up the brand name, and its flagship drink. It seems a long way from this to a novel, but I will argue that strategic (rather than random) linguistic choices, such as the choice of firstperson or third-person narration, do propagate through narrative structure to orchestrate narrative viewpoint throughout the levels. In other words, anaphoric and cataphoric use of pronouns is a concept which explains a host of referential meanings affecting sentences, but in a longer text these concepts do not suffice (cf. Emmott 1997). Similarly, a consistent choice of past tense as the narrative form affects not only the meaning of every sentence used as ‘narration,’ but also the viewpoint construction in any sentence elsewhere in the narrative which could be described as representing speech or thought. For these aspects of meaning to be accounted for, we need to assume that grammatical choices can single-handedly construct (not just participate in) meaning configurations that they are ordinarily a part of. 1.2
Literary analysis and linguistic analysis
To be able to make its point, this book will engage with a number of storytelling artifacts. I will be relying on material mainly from novels, but also from travel narratives and plays; there will also be some short examples from poetry. Texts will be chosen to represent particular phenomena, though some questions
1.2 Literary analysis and linguistic analysis
9
regarding the historical emergence of various narrative strategies will also be addressed. Crucially, none of the literary periods will be specifically foregrounded. This evokes the question often raised in the discussions of cognitively motivated attempts to talk about literary texts. As Weber 2004 (among others) points out, cognitive methodology often downplays or disregards the historical and cultural embedding of the texts, or differences among readers. If it is indeed the case, then does it mean that such analyses are lacking something essential? Certainly not, and there are many reasons for this. First, it does not seem to be the case that everything that can be said about a historically situated text is period specific. If anything at all is indeed wrong, it is the suggestion that only period studies can reveal frames and cognitive patterns. What cognitive methodology intends to capture is patterns of thought, and not exclusively patterns of Victorian thought or modernist thought. Still, this approach does not preclude cognitively informative work looking at a specific period. In fact, quite a few Renaissance scholars (Hart and McConachie 2006; Cook 2007, 2010; Moschovakis 2006) are doing cognitively informed work themselves, and the conversation between the generic and the specific is most fruitful for both sides. Second, it is not true that ‘cognitive’ equals ‘culture neutral.’ Some of the most thorough cognitive analyses of literary texts (cf. Sweetser 2004; Turner 2004; Canning 2008) are also deeply immersed in the beliefs of their respective periods. Still, there are probably many cognitive analyses where writers ignore literary/cultural/historical expertise, either because it is not relevant to their topic or because they have different goals, and it would seem unreasonable to assume that all cognitive work on literary texts has to be supported with indepth considerations of their historical context. Also, there is the question of bridging the gap across different ways in which language is talked about in linguistics and in literary studies. Current approaches to language in the work of philosophers such as Derrida or Lacan are not easy to reconcile with routine linguistic analysis. However, cognitive approaches to meaning are less formal than most and also more open to questions of cultural or emotional framing, so it seems possible that the linguistic and literary scholars might find shared interests here. Linguistic work on literature is often seen as restricting the freedom to offer a new interpretation (cf. Jackson 2005). Yet, while meaning is not entirely indeterminate, it is also not determinate. It is a perfectly natural reaction on the part of those engaged with various sources of interpretation to feel that someone arguing for a single interpretation is missing the point. And it is true that many stylistic and cognitive analyses can be easily read to claim that what they discuss is the ‘prescribed’ interpretation (even though I am not aware of research where such claims are made explicitly). The assumption of this book, for comparison, is that while it is impossible to predict all reader responses and
10
Language and literary narratives
possible interpretations of a text, the focus on cognitively based processes of interpretation should in principle open the floor to this discussion. Indeed, what is it that counts as an interpretation? Are there perhaps interpretations out there which depart excessively from optimal avenues of text processing? Certainly. And, also, are there perhaps language-based interpretations out there which are too restrictive? Of course, there are many. The point in these cases seems to be that the interpretive tools employed are either not specific enough or too rigid, and, I argue, these limitations are the result of broader assumptions about . . . yes, about language. Whether one believes that linguistic conventions are irrelevant to meaning or that they determine meaning entirely via grammar and discrete lexical packets, the distortion is almost equally dangerous. The model of language which at least attempts to avoid these kinds of traps is a model which assumes that all meaning is constructed, regardless of the level we are looking at. The semantics which would be satisfied with the characterization of the word such as a rock as [−animate] [+concrete] [+count] will not be able to open itself to the contexts where a rock is viewed as a geological specimen, material suitable for sculpture, material suitable for construction, a weapon, a tool, or a symbol of stability and certainty, or to the possible contextual ambiguities. By the same token, if a linguistic model assumes that the present tense is simply an aspect of the morphology of English and that it is different from the imperative form, it will not even address the possible interpretation of a common usage where these forms are difficult to distinguish, as in an airport BMW ad featuring an attractive picture of a car and the line Miss your flight; still make the meeting. The slogan is obviously not issuing an imperative, and it is not describing a situation which can be construed as the use of ‘present tense.’ What is required is an understanding that a similar use of a present form appears in non-predictive clauses of predictive constructions (If you miss the flight, you can still make the meeting), and that the verb form is to be understood as neither future nor present nor imperative in meaning, because the specific constructional meaning evoked is ‘non-predicted assumption to be used in further reasoning.’ Dancygier and Sweetser (2005) propose a broader discussion of how a verb form may take its meaning from a construction it participates in and how the use of the form may prompt the meaning of a whole construction. The point I am trying to reiterate here is that grammatical and lexical choices may carry more meaning than some grammars allow. On the other hand, from the language perspective, an interpretation may appear unnecessarily open-ended. Results vary, but there are clear cases where interpreting passes over central aspects of narrative meaning because of overreliance on cultural concepts. In Margaret Atwood’s novel The Blind Assassin the actual plot, or ‘what really happened,’ is not narrated directly, but requires construction by the reader. The crucial question of the plot is whether a man accused of arson actually did commit it. He is wanted for the crime, and the
1.3 Literature, language, and human nature
11
consequences of the fire ripple through the lives of all the characters, but no answer is provided. It is in fact available in the text, but only if the reader assumes that all its narrative strands, even if apparently incongruous, are present for a reason and that the form of the novel can be interpreted in a way which reveals the whole plot. However, the indirectness of the narrative actually distracted quite a few critics (Wilson 2006; Perrakis 2006; Roy 2006) into thinking that the surface incongruence is in fact yet another instance of the ‘instability’ of meaning, and into focusing on one of the themes common in contemporary criticism – violated women’s bodies. The unfortunate consequence is that they failed to consider crucial aspects of the plot and essentially misrepresented the plight of the main character. In cases like these, the problem is not that the ‘interpretation’ simply lacks depth and convincing power, but that the interpretive avenues chosen impoverished the result, while an analysis using the complexity of the structure to reveal the complexity of meaning might have worked better. Some critical work might be read to suggest that writing challenging literature intends to background the very idea of a coherent interpretation. From the linguistic perspective, such a suggestion would be difficult to accept. Within the framework I’m assuming, there is no reason to believe even the most challenging literary texts to be beyond our interpretive strategies by their very nature, although in fact readers do give up on texts which do not engage them enough. As long as a reader is prepared to do what she does in colloquial contexts, that is, read out of a text but also read into it, the most difficult narratives can be read and coherently interpreted. With this assumption, fragmented and ostensibly incoherent stories are indeed ultimate exercises in language use, because they stretch the cognitive abilities of ‘making sense’ to their limits. It has been my experience, as well as that of many of my students, that some practical expertise in cognitive analysis opened higher levels of our abilities as readers, by making explicit the strategies which remain implicit in less challenging instances of reading. 1.3
Literature, language, and human nature
While the nature of literary and linguistic meaning is still a hotly disputed subject, more recent arguments revolve around a much broader set of concepts involving human nature, evolution, the origin of culture, et cetera. There is a growing interest in various areas of the humanities in the study of the individual mind, independent thought, genetic endowment, cognitive processes, and the bodily basis of cognition, but we are far away from an agreement on how these questions are to be incorporated into the existing body of scholarship. One proposed approach relies heavily on the findings in evolutionary psychology and related disciplines. For example, Steven Pinker’s book The Blank Slate
12
Language and literary narratives
(2002) is a passionate defense of what we used to call ‘human nature,’ our shared as well as individual cognitive and evolutionary heritage which is not subject to cultural or social molding. While the evidence he presents to talk about, for example, education is clear and convincing, the argument becomes less clear when it comes to the final discussion of the arts. Pinker criticizes the current state of the arts rather harshly, and further points out that esoteric concepts such as ‘artistic value’ have direct correlates in ‘human nature.’ To conclude his argument, he discusses some literary examples. Unfortunately, it is not likely that the way Pinker chose will convince anyone seriously interested in literature or language. Several anecdotal examples not supported with any specific argument cannot count as evidence for the above rather strong claim which the reader is expected to accept. Intuitively, his claim that literature reflects human nature is equally unsubstantiated as insisting that it only reflects human culture, and the earlier extensive discussion of the tensions between nature and nurture do not suffice in supporting this very specific claim. An off-hand statement that reading initiates a cascade of neuronal and cognitive events does not explain how this is different from the avalanche of such events prompted by an appetizing smell or a beautiful view, or how humans have come to enjoy interacting with fictional artifacts of various kinds and keep on perfecting the means of creating them. While I agree with the general gist of Pinker’s story, it is somewhat too cursory to be satisfactory. I appreciate, though, that the arts were not the main focus of his project. However, another claim implied in much of Pinker’s work is convincing – the humanities should return to an attempt to understand humans as they are. In view of serious advances in the sciences the goal seems to be somewhat more realistic, and there is more and more work pointing in that direction. For one, Slingerland’s argument (2006) for what he calls ‘vertical integration’ (of arts and sciences) presents much of the already available evidence for the viability of such endeavors. Also, the growing interest in cognition within the study of music, visual art, performance, film, et cetera suggests that there is an awareness of the need for the humanities to at least interact with the explanations science offers and to prompt new questions in science.5 There is, however, the tricky issue of how exactly science fits in the humanities and what humanities need to do to expand their current understanding of what culture contributes to literary reading. In an issue of Criticism, for example, Evans (2007) makes specific suggestions for a more refined view of culture, such that would reflect the complexity of our times and fit better into the interests of criticism (and not cultural anthropology). In the same issue, van Oort (2007) discusses the potential for interaction between humanities and hard sciences and argues that a close methodological fit, such as the one proposed by Slingerland (2006), is not desirable, because humanities need to focus on the specifically human ability to use symbols in communication and culture, even
1.3 Literature, language, and human nature
13
at the cost of downplaying more traditional critical interests. While this is an interesting argument, I am not yet sure that the nature of symbolic representation is a rich enough base to enable the ‘after-theory’ humanities to flourish. It is also interesting that van Oort expects the humanities to include the discussion of language (I couldn’t agree more), but, surprisingly, puts cognitive linguistics among hard sciences, and thus outside the humanities. Further discussion of how cognitive linguistics might be useful to the humanities will be taken up throughout this book, but van Oort’s suggested dividing lines will probably surprise researchers on both sides. The point where I agree with van Oort is when he suggests that the humanities need their own methods, whether science-based or not. However, the call for the integration of disciplines is sometimes read as an encouragement to make humanities look like sciences. One such effort is represented by a group of literary critics now known as ‘Literary Darwinists.’6 Their claim is that literature is an evolutionary phenomenon and should be studied as such. Their main argument is that literature is a direct reflection of the human evolutionary set-up, so that literary texts reflect concepts such as mate selection, dominance, and other concerns of species survival and reproductive psychology. I am reading their work as a response to Pinker’s call for putting human nature first, but the specific choice made by this group is tantamount to substituting ‘mother nature’ with ‘mother evolution.’ As such, it is a misrepresentation of literary tradition and its cultural and humanistic role, and another bold attempt to find one key concept to explain a complex issue. The swing of the pendulum from culture to evolution does not even ask how culture is linked to evolution and glides gracefully over everything in between. Also, while the literary Darwinists focus on the few relevant aspects of human nature, and do not include historically determined cultural concepts in their discussion, they are also avoiding any recognition of what schools of psychology other than evolutionary psychology have to say. This also includes theories already used by other literary scholars, such as the theory of mind, cognitive psychology, or the psychology of emotions (Gerrig 1993; Zunshine 2006; Hogan 2003). The conclusion is that this specific attempt at marrying science and the humanities is not a particularly promising avenue. My own understanding of the problem is that building such bridges requires more than one step. One cannot go from evolution into literature directly, because too much is missing: human cognition, including the cognitive roots of culture; the cultural and cognitive role of art; and the role of language in representing both. The gap is not easily filled, but some attempts have been made. Throughout this book, I will rely on some of the concepts formulated by Donald (1991, 2001, 2006) and Deacon (2006). They both refer primarily to art, but many of their claims are naturally applicable to literature (some are not, but they do not undermine the use of those which do apply).
14
Language and literary narratives
Donald’s approach assumes that there is a dual explanation to the emergence of art. One is mimesis, as the primary and evolutionarily oldest mode of representation, while the other is self-reflection. He argues that art, which relies on both, is an activity intended to influence other minds, and it functions in the context of distributed cognition. It thus has a dimension that is both individual and social, as mental representations are also necessarily shared and requiring a mode of transmission (such as the early mimetic forms involving the body or purposefully created representation, and, in the later stages, a specialized form of communication, that is, language). The self-reflective nature of art applies not only to the individual, but also to the society as a whole, and, in later stages, to human nature. Donald also stresses that art does not emerge solely from an individual mind, but that the source of creativity is in fact in the distributed network of cultural memory and its technological development. Thus a cave drawing and a cathedral are both rooted in the cultural system the artists participate in, and even though they rely on currently available technology, they serve the reflective needs of the society as well as creating specific cognitive outcomes – monitoring collective memory, giving salience to cultural values, reinforcing the social network. Narratives can naturally be viewed similarly (cf. Herman 2006). The self-reflective nature of art is further discussed by Deacon (2006), who sees the role of art in terms of the way it affects individuals emotionally. Interestingly, Deacon proposes a category of ‘emergent emotional states’ which rely specifically on the availability of more than one cognitive agent or representation. Candidates for such emotions are irony, nostalgia, righteous indignation, or humor, because they imply an interaction of various cognitive constructs. The source of nostalgia, for example, is being able to access two mental representations at the same time – the one reflecting the current state, and the other rooted in the emotionally loaded memory. To put it in cognitive linguistic terms, the tension between different construals or mental spaces creates an emotional/cognitive reaction which is different from basic emotions (see Damasio 1999, 2003) such as fear. Deacon’s emergent emotions can only appear in the network of distributed cognition, so that both the foregrounded construal and its competitor are available (as when I feel indignant but the cultural norm approves). The concept also seems to tap into Damasio’s theory (1999) postulating the emergence of human consciousness out of just such complex emotions. While Donald’s and Deacon’s work gives us a top-down view of how the cognitive architecture of our species affects human production and reception of art (including literature), there are more issues to be considered in the bottom-up view. Donald makes it clear that art works in a two-way interaction, where the artistic object offers its structure and framing and the viewer/reader adds her or his own, and Deacon also stresses the symbolic roots of art, where the
1.3 Literature, language, and human nature
15
representation emerges in the recipient’s mind. There is, however, a whole set of questions pertaining to the cognitive work done in the close interaction with narratives. For example, the work in cognitive psychology reported in Gerrig and Egidi (2003) focuses on the postulated automatic and necessary types of strategies of narrative comprehension, emerging out of experimental work. There seem to be two major possible strategies for reading narratives: either text processing is primarily memory-based, or it relies on a search for coherence which prompts the construction of causative chains. While testing memory mechanisms may seem a safer and more minimalist route, researchers working closer to actual literary narratives (Emmott 1997) are inclined to propose reading strategies which are constructivist and coherence-based. Evidence from the studies of the therapeutic role of narratives (Klein 2003) also supports this approach, as subjects find that coherent accounts of traumatic events prevent recurring uncontrollable emotional responses to the memories of trauma. Causal coherence is one of several strategies proposed in Herman’s (2003a) discussion of the role of narratives in the organization of experience. Herman follows Bruner (1991) in viewing narratives as tools for thinking. The specific strategies Herman discusses are: chunking experience, imputing causal relations among events, framing events as instances of a certain type of occurrence, and construction of event sequences. Jointly, the strategies allow one to construct events as bounded concepts appearing in a sequence and linked causally (‘he proposed, she accepted, they got married’). One might add that these strategies are supported by a number of linguistically available frames (such as, what counts as a marriage proposal), scenarios (marriage requires prior consent of both parties), and attribution of intentionality (his proposal counts as the expression of his wishes), but the fact that such conceptualizations in fact rely on each other makes it difficult to determine which of the prompts is the crucial one. It seems that such instances of ‘organization of experience’ are in fact a kind of feedback loop: the emergence of narrative structure may crucially rely on such lower-level cognitive operations, while at the same time these operations support categorizations of behavior which make narrative structure possible. Furthermore, events can be categorized as participating in a narrative chain only when they can naturally be construed as having causal force – otherwise, they do not participate in narrative structure. It is thus clear that causality plays a crucial role in the concept of narrativity, but the way in which it participates in narrative structure is more difficult to determine. In any case, these low-level cognitive processes are an important aspect of the construal of narratives. To sum up, the resulting picture of art-based cognition requires that the properties of the brain and our core skills of symbolic representation be put in the context both of a social network which gives validity and purpose to the creative acts, and of the lower-level cognitive operations such as chunking and
16
Language and literary narratives
adding causal connections. Moreover, while art relies on our mimetic imagination, it becomes more complex as humans’ technological, cultural and communicative abilities are honed in the context of a human culture understood as a cognitive distributed network. It is perhaps possible now to see how literature owes its existence to our general cognitive abilities, to our ‘artful’ mind, and to the communicative needs of a group, all of which gave rise to more and more sophisticated use of language (see Dancygier [in press a] for a broader discussion). The combination of human cognitive set-up, the cognitive roots of culture, and the cognitive roots of language is what might be an account of literature which uses science without simply repeating its claims in a context which is not reducible to the concerns of scientific research. As more and more researchers notice, being able to see the human brain at work through fMRIs and scans will not tell a researcher in the humanities much beyond the obvious observation that something is happening in the brain as we listen to Mozart or read a novel. The humanities need their own tools to represent the ultimate forms of human language and cognition, though these tools cannot ignore what science has to offer. In this book, I will attempt to clarify one end of the spectrum – the contribution of language as a cognitive tool to the formal and interpretive complexities of literary texts. 1.4
Literary texts and communication
If questions of ‘evolution’ are relevant to this enquiry, it is to the degree to which we can attempt to answer the question of how literature has emerged in its present form. I am not at all interested in speculating whether literature is an adaptation or an evolutionary by-product,7 because the way such questions are posed assumes that once it has emerged it continues playing its evolutionary role and whatever changes in literary form is no longer relevant to the question. I believe that the questions which can be considered profitably are different ones: what form of communication has fiction evolved out of? what can the emergence of different genres and poetic forms teach us about both literature and language? did language have to change to accommodate literature, and if so, how? how do literary conventions depend on our folk understanding of the human mind and human emotions? These questions cannot be answered through evolutionary theory, but they are, in my view, directly pertinent to the question which seems to be the evolutionary Holy Grail: why do humans create and read fiction? In what follows I will assume that in order to understand the emergence of literature we have to place it in the context of communication. Broadly speaking, I find the picture emerging out of Tomasello’s (1999, 2008) argument about the origins of human communication to be a good starting point, especially in its interesting elaboration of the concepts proposed by Donald and its reliance on
1.4 Literary texts and communication
17
the concept of joint attention, recently broadly discussed by Oakley (2009). Tomasello sees the emergence of narrative as relying on the human basic need to share information and to arrive at social, and also linguistic, norms in the process. Further, Tomasello claims, telling stories gives rise to specialized linguistic constructions which allow tellers and listeners to connect events across different temporal planes, to keep track of individuals, and to take another person’s perspective. Indeed, the emergence of elaborate pronominal systems and of constructions reporting thoughts and feelings while clearly keeping track of whose thoughts and feelings they are suggests some form of feedback loop between narrative and linguistic forms. Tomasello’s model can also be useful as a starting point in an attempt to track the emergence of fiction as a step in the emergence of forms of sharing information. An attempt to establish social norms through storytelling requires at some point either that the narratives used to establish those norms become more salient than what is available in the immediate environment, or that the actual true stories are told in an exaggerated form. Perfect examples of this can be found in folk myths and stories. For example, Scalise Sugiyama’s (2005; Scalise Sugiyama and Sugiyama in press) work, including her account of stories of cannibalism told in various oral traditions of foraging peoples, poses some crucial questions regarding the history of storytelling as an information-sharing device. Do the vivid and horrifying descriptions of cannibalism she discusses actually represent the past of all of those peoples? In other words, did it actually happen that children could be cooked and eaten by their parents in time of famine? Or are these stories the products of generations of inventive storytellers who made the stories more and more vivid to make their messages more memorable? Have children actually been abducted and devoured when they left the safety of the settlement, or did they need to be discouraged in exaggerated ways from leaving the safety of the group? From the point of view of Tomasello’s account the question is as interesting as it is unanswerable, since it is impossible in these cases to separate the account of events which occurred in the past from the narrative selection and presentation of events which play the normative cultural role better. Since the question about the fictionality or truthfulness of early narratives cannot be resolved, it is probably more useful to consider the ways in which their messages may have been structured. One such way relies on linguistic markers of evidentiality and hearsay (similar to English adverbs like reportedly, but often taking inflectional form, as in Central Pomo, Makah, or Pawnee). While evidential markers appear in some languages to distinguish first-hand from second-hand accounts of events, other languages develop special storytelling forms such as the ‘mythic’ or ‘ancient’ past (in Upper Chinook, or Kiksht’). While evidential markers are primarily available in spoken communication, the ‘mythic past’ has developed specifically for narrative purposes. In other words, the more salient the cultural role of the narrative, the less salient
18
Language and literary narratives
the question of its truthfulness or the specific source of information. Once the text becomes a shared cultural artifact, it participates in the system of distributed cognition and is no longer treated as one speaker’s communicative contribution. Presumably, culturally salient stories are tools of distributed cognition and as such may develop a linguistic status of their own. It is also important to note that such stories are subject to the process which blending theory refers to as ‘compression’ – the primary process leading to new blends. An example of compression could be a very simple expression, such as the generations of tomorrow, where the deictic form referring to the day following the current one (from the point of view of the speaker) is used to refer to all such days in our future. The compression of all specific ‘tomorrows’ into one downplays their temporal or deictic location, and highlights the speaker’s generic intent. In the stories under discussion here, various potentially factual events are compressed into the one the story actually tells, moving again from the specific to the generic. In the process of compression, the story may undergo various embellishments, so that the generic aspects (the ones prompting ‘narrative transport,’ as Gerrig 1993 defines it) are more salient. Thus the gruesome image quoted by Scalise Sugiyama and Sugiyama, with the mother stirring the pot in which the children’s little feet and hands are still visible, is too strikingly detailed to be an actual description of an actual event from distant past. Its salience and descriptive detail seem to serve the purposes of the generic message and effective compression, not the accuracy of the storyteller’s account. One can thus conclude that such narratives are true and not quite true at the same time, because the singularity of the event recounted is not important from the point of view of the cognitively distributed goals of the group. Looking at oral storytelling from this perspective may help us see the connection between early narrative forms and highly elaborate contemporary ones. First, oral stories are limited by the capacity of human memory, but, at the same time, they expand that capacity by relying on the storyline as the glue which holds various events, times, people, and locations together. Written narratives capitalize on these narrative constructions, but they are gradually freed from the worry about anyone’s capacity to retain the story in its entirety, and narrative form becomes more and more complex. Second, the stories such as the ones discussed by Scalise Sugiyama (also in her earlier work) are clearly intended to achieve a certain effect in the listeners – compassion with victims, fear of predators, increased sense of dependence on social and family structures. But these effects are achieved as a kind of ‘narrative uptake,’ not because they are directly communicated. The mechanism whereby an elaborate linguistic construct can prompt ‘narrative transport’ (cf. Gerrig 1993) must have been present in the minds of the foraging peoples and it is present, in an enhanced form, in our minds. Still, regardless of the form and context of the story, the ‘point’ of the story is only available through constructing its meaning – not
1.4 Literary texts and communication
19
only by tracking the sequence of events, but, more important, by exploring all the ‘what-if’ scenarios which could have prevented overly confident children from being abducted or might help us understand ourselves and our world better. There is no limit to what a narrative can do, but it requires a mind which uses the story as it is told or written to construct its meaning. Apart from this broad context in which stories achieve their status of meaningful artifacts there is also the question of how a literary text is similar to or different from any other text, spoken or written. I will assume that language and its basic communicative context are the basis on which literariness is built. I treat the basic communicative context similarly to what cognitive linguists call the Ground (Langacker 1991, 2001; Coulson and Oakley 2005; Verhagen 2005) – the immediate surroundings, the presence of the speaker and the hearer, shared knowledge, shared visual and aural field, shared understanding of bodily experience, and, last but not least, the surrounding discourse context. The Ground makes various aspects of communication possible, since meaning can be derived from ostensive aspects of communication, from objects manifest in the environment, and from the interlocutors’ body language, gesture, facial expression, tone of voice, et cetera. In my understanding, the Ground conceived in this way also provides access to shared cultural and conceptual framing, so that the interlocutors can assume access to culturally salient concepts. The difference between the Ground of any specific exchange and the default communicative context is that the latter does not have to profile any individual interlocutors, but rather a generic concept of what engaging in communication is about. Similarly, what is required is a concept of a shared visual field, and not a concrete visual field accessible in a specific exchange. Another aspect of what is required for literature to build on is the assumption that linguistic expressions prompt conceptualizations. As Verhagen (2005) argues, any exchange may involve two aspects of conceptualization: the ‘subject’ of conceptualization (the speaker, the hearer, other participants, and salient aspects of the Ground) and the ‘object’ of conceptualization (the actual subject matter of the exchange). While the potential disagreement between the speaker and the hearer as to the actual conceptualization of some subject matter is always a possibility, what Verhagen also shows is how much of the actual linguistic structure is addressed at negotiating those conceptualizations. If a child visiting a zoo calls to the parent “Look, Dad! A dragon!” this counts as an attempt to conceptualize a phenomenon manifest in the Ground and in the shared visual field, but when the father answers “No, Jimmy, it’s a lizard/a crocodile/a snake . . .” this counts as negotiating a different conceptualization, with its attendant cultural consequences (‘we cannot start pretend play whereby you are a knight and I’m a wizard and we slay the dragon, but we can learn more about reptiles’). Naturally, there is no such potential exchange between the participants in the act of reading (reader versus author/narrator/text), but the
20
Language and literary narratives
negotiation of conceptualizations is built into the narrative structure. This is, I think, what Tomasello talks about when treating narrative as an engine to produce more complex linguistic constructions. These structures not only maintain temporal, referential or viewpoint links, but allow the construction of narrative meaning to happen. The greater part of this book is devoted to revealing the various ways in which narratives perform this task. It is important to be specific about how narrative context deviates from the default communicative context. First, ordinary communication features two interlocutors, the speaker (I) and the hearer (you), who, in the most standard set-up, have visual and aural access to each other and exchange their deictic roles in every turn of the conversation. In this default set-up what counts as communicated has to be said or done (or, from the point of view of the receiver, heard or seen). Naturally, this would never suffice, and that is how languages develop various forms of establishing categorizations, encouraging inferences, reaching beyond the available Ground, reporting past situations, et cetera. But the default, as I will try to show in chapters 6 and 7, remains surprisingly persistent in structuring different literary forms. Deviations from the default are common, and in most cases discourse is maintained via appropriate constructions, such as back-channeling in phone conversations (all the yeah and hmm expressions signaling continued presence), emoticons in e-mail, page layout and paragraphing in journalistic discourse, intonation patterns, rhetorical devices and repetitions in public speaking, et cetera. More interesting, however, there are instances of discourse which seem to assume the default in obvious absence of its crucial ingredients. People routinely speak to animals and plants, and also to cars, computers, and other gadgets (mostly phrases like Don’t do this to me!), while animals and objects commonly speak in stories, comics, and poems. Not only does this deviation not cause incomprehension, but most of the time it also goes unnoticed. It seems that the very fact of something being said evokes the default set-up, and thus ostensible communication is treated (linguistically) as bona fide communication. Such a ‘constructed’ communicative set-up is just an example of the fact that something being communicated suggests intentional communicative agents. Dennett (1997) talks about our ‘intentional stance,’ whereby we attribute agency to objects and animals, but it seems to be only a part of what Deacon calls a ‘representational stance,’ whereby we tend to treat things occurring around us as meaningful. In the context of a written narrative, both stances are involved. The presence of a text is treated by default as interpretable via our symbol-reading abilities, and as presented to us in its form by an intentional agent. Reading is thus a communicative act which has an addresser and an addressee, even though the communicative intent is available only through the form of the text in front of us. It is interesting to note how natural it is for readers
1.5 Why is fiction special?
21
to talk about ‘books’ or other texts as ‘saying’ something, and attributing intentions to authors (‘the author is saying X,’ ‘the author meant X,’ et cetera). These ‘fallacies’ seem to rely on the default understanding of communication. This seems also to be the case with extended narratives. There used to be a teller of an oral story, and the listeners may have assumed that the communication mostly goes one way, but these assumptions did not affect the communicative impact. Contemporary fiction is mostly written, and so the ‘communicator’ is hidden behind the text, but the concept of a narrator has substituted that of a teller. As I will argue throughout this book, the narrator (but not the author) may be no more than a linguistic construction (or constructions, since there are many kinds of narrators), and yet a written work of fiction is naturally seen as conforming to the communicative default. Also, earlier narrative forms are more preoccupied with mimicking the ‘somebody told me’ or ‘I saw it’ pretense, while less narrator-mediated forms emerge much later. Furthermore, visual prompts participate in storytelling alongside words, and often in ways which not only support words, but sometimes give them different meanings. 1.5
Why is fiction special?
Or is it? There are several recent attempts to address the question. Zunshine (2006) sees reading fiction as practice for our theory of mind, assuming, with Palmer (2004), that the access to characters’ minds is more important than the story itself. Oatley and his co-workers,8 for comparison, argue for some kind of cognitive gain, especially if the material read is sophisticated and has some psychological depth. Similar claims, though based in an evolutionary view, have been made in Boyd (2009), who argues for the development of fiction as driven by cognitive play. Researchers in the cognition of art look at fiction as a form of art, and try to view it alongside the path leading from cave drawings to Van Gogh. Others seek the value of fiction in entertainment, in crossing the boundaries of time and space, in exercising compassion, et cetera. However, there are many questions which are not addressed through these enquiries. First, why do people write fiction? The view from neuroscience, lucidly presented by Flaherty (2004), suggests that creative urges to write are to a degree a part of our biology. From the cultural perspective, there has never been a problem with low supply of aspiring writers, though we do not seem to question the fact that only some writers make a difference to our view of the world and have pushed the boundaries of narrative and poetic skills to new heights. There is plenty of fiction which seems no more than chewing gum for the eyes, and yet finds plenty of committed readers, and definitely counts as literature. There is also the new phenomenon of writing as a job. When Haruki Murakami, a very prolific and popular author, talks about his career, he simply
22
Language and literary narratives
says that he thought at some point in his life that he could make a living by being a writer (Murakami and Gabriel 2008). My point is that while we tend to focus on the writers who make a difference and build on that, a lot of storytelling seems to have much less sophisticated forms or less high-flown goals. It is a form of linguistic expression which builds on culturally supported cognitive needs. Furthermore, we ‘talk fiction’ much more than we are willing to admit, and our languages have elaborate structures allowing speakers to explore different scenarios. Conditionals and other hypothetical constructions in languages are varied and complex, no less complex than a fictional story, if not more, simply because of the compactness of actual constructions. There are many ways for speakers to mark evidentiality, epistemic stance, or emotional stance, and most of these linguistic forms are either lexicalized, or grammaticalized, or constructionally determined. Linguistic forms are typically understood to mean more than the facts described (consider implicature, linguistic politeness, indirect speech acts, and the like), and open themselves to reasoning more often than just stating the facts (hence so many attempts to treat language as based on some kind of logic, however misguided some of these attempts may be). My point is that the idea that colloquial language use is about facts while fiction is not is simply wrong. Language is very well equipped to deal with ‘what ifs’ and hidden meanings and is usually very clear in representing the speaker’s attitude, knowledge and the source of information. Contrary to common belief, statements that logicians love to work with, such as Grass is green or Mary is in London, are much less common than ones starting with if, I think, he said, or I wish. The question is, then, why treat fiction as . . ., well, pure fiction? In his delightful personal account of narrative techniques, Vargas Llosa (1997/2003) talks about the author as resembling “the Catoblepas” – a mythical creature which consumes itself, starting from the feet. As he argues, the themes an author will explore have to emerge from his own experience, though he can then give them any narrative form he chooses. There is thus a grain of truth in any fictional narrative, however weird it may be. Setting the case of realism in fiction aside, genres departing from ‘realism’ also are real to a degree. Even a cursory look at the ‘ultimately’ fictional genre of science fiction shows that while trying to construct storyworlds as remote from planet Earth as they can, authors cannot but rely on our reality in many of its aspects. At the other end of the spectrum, texts which are treated as novels but have been described (by their authors as well) as primarily autobiographical are versions of actual happenings. A perfect example of this phenomenon is J. G. Ballard’s acclaimed novel Empire of the Sun, which is a blend of fact and fiction. As a child, Ballard did spend three years in an internment camp in China, and the novel is based on his story. In his autobiography, The Miracles of
1.6 Narrative and grounding
23
Life, Ballard claims that most of the events described in the book happened in reality, though not to the people actually portrayed in the novel. He also explains the reasons for his choices, for example, for placing his character, Jim, in the camp without his parents, while in fact they were there with him. Ballard wanted to focus on a child’s experience without the protective layer which could only blur the message – a choice as much related to compression and narrative salience as the one in the cannibalism story. Fiction thus cannot be divorced from facts, some as basic as the existence of oceans, dogs, and bacteria, and some as specific as the conditions in internment camps in China during World War II. Furthermore, the need to categorize any given story as either true or not true seems to be important for cultural rather than linguistic reasons. It is important to know whether, let’s say, a politician accepted a bribe or not, because it is a legal issue. But whether a man named Darcy actually married a woman named Elizabeth Bennett, or whether they never met, or never even existed, is not relevant to the story. Pride and Prejudice is not a true account of what happened (as far as we know), but it can be read as such and may not ‘feel’ different in terms of factual plausibility from a story about distant cousins one can hear from one’s grandmother. Crucially, the latter story would have a different form: it would be shorter and use colloquial linguistic constructions. It thus matters little whether the story presents facts, but it is crucially important how it is told, and here is where the issue of literary discourse becomes interesting. To conclude, fiction has to be based on reality to have its effect. Reading biographies is often as moving and interesting as reading fiction, because the textual matter and the attendant constructed meaning or emotional response is triggered through the same medium and the same conceptual mechanisms. Uncovering the linguistic specificity of the medium and the conceptual mechanisms leading to narrative uptake is the goal of this book. 1.6
Narrative and grounding
One of the themes in the discussion of narratives which this book engages with is the cognitive status of narrativity. While definitions of narratives vary, they consistently stress temporality, causality, representation of minds, relationship to spatially and temporally defined storyworlds, problem solving, and the construction/resolution of some conflict (see Ryan 2007 for an interesting discussion). From the cognitive perspective, Herman (2003a) argues further that narratives are ‘tools for thinking,’ which help in the organization of experience, through the establishment of chunks of experience and causal links among them, recognition of types of events, et cetera. These lists include some more general concepts such as causality, and some which are narrativespecific, such as the ‘storyworld’ or ‘conflict.’ And yet there are also attempts
24
Language and literary narratives
to reach to the very conceptual core of the narrative and establish the role of narrative structures in thought processes in general. One such attempt is the book by Turner (1996), which sees basic narrative organization as a primitive concept and the source of the emergence of both linguistic and literary forms of expression. As Turner argues, perception of basic ‘spatial stories’ such as opening the door or throwing a rock is the starting point of higher levels of conceptualization. In this section, my goal is not to arrive at some definition of narrative which would reconcile all these descriptions, but to consider some problems in relating narrativity to conceptualization. Herman’s approach (2002, 2003a) identifies aspects of narrative which participate in conceptualization of events as linked into a story. Generally, Herman sees the story as a composite concept which provides a link between raw experience and its ‘storylike’ conceptualization. Turner, for comparison, views ‘storiness’ as a basic perceptual tool, leading to conceptualization of all kinds of experience as well as to projections of one ‘story’ onto another, leading to metaphor, parable, and more complex narrative forms. Essentially, Turner’s concept of a primitive narrative is a configuration of image-schematic concepts rooted in sensorimotor experience in which (causal) force or sequence plays a central role. Turner’s approach to narrative, and, by the same token, the entire theory of conceptual metaphor and blending, was extensively criticized in van Oort (2003), primarily on the grounds of what he perceives as the assumed direct link between perception and conceptualization. Since the issue, which I will refer to as the ‘grounding problem’ (after Harnad 1990), is clearly at the core of much of the current discussion of the narrative, I will attempt to review the questions briefly. In much of the work in cognitive linguistics it is assumed that meaning is grounded in perception and sensorimotor experience.9 Basic spatial configurations called image schemas (such as containment, force, balance, boundedness, et cetera) lead to the emergence of linguistic expressions describing concrete aspects of experience, but can then be projected into abstract domains of human activity. Thus most concrete categories as well as many abstract meanings are grounded in human perception and experience, via levels of projections which also distinguish multiple meanings of lexical items. For example, the preposition in can be used to describe concrete cases of containment, but also abstract states such as being in despair. Van Oort criticizes this explanation by arguing that it assumes a direct connection between perception and linguistic categorization, while what is needed is an explicit inclusion of a separate intervening level of symbolic representation. Van Oort refers extensively to Deacon (1997), who sees symbolic representation as the core of human cognition. He also evokes the claim that words, as symbols, are primarily in relationship to other words, and not to the speaker’s perceptions or actions. Finally, van Oort reads Turner’s discussion of the basic role of
1.6 Narrative and grounding
25
narrative as a hidden claim of the ‘story’ being an innate concept, very much like Chomsky’s universal grammar. It might be helpful to put this discussion in the context of recent experimental work in cognitive psychology. Surely, the idea of representation and categorization is at the centre of the discipline’s interest, and some of the research done in the last few years explicitly addresses the grounding problem – what does it mean for perceptions to drive conceptualization? how are categories formed in the brain? how do abstract concepts emerge? First of all, research starting as early as the 1990s (Barsalou 1999; Harnad 1990) suggests that treating meaning in terms of links within a network of abstract symbols in the mind runs into problems. One, the approach does not explain the observed emergence of arbitrary symbols from experience and perception, and, two, it offers no solution to the commonly asked question about the ways in which symbols connect to reality, since without some connection to the real world, symbols cannot perform any cognitive function. Experiments which ensue suggest that perception and action are central to higher cognition, without postulating intervening mental representations.10 Moreover, one of the claims is that experience can be used in the conceptualization of situations in other times and places, and the idea led to further investigation of how basic perceptions and actions can participate in conceptualizations without necessarily going through the configurations of arbitrary symbols. There seem to be many reasons not to treat discrete symbolic representation of categories as the basis on which meaning emerges. Recent work by Prinz (2002, 2004, 2005) makes a very strong case for the experientially grounded emergence of abstract concepts, including emotions. First, Prinz shows how even complex concepts such as ownership can be naturally explained through configurations of basic perceptual and sensorimotor concepts. Second, he sees concepts as necessary for action, and not for static categorizing, so that conceptual representations are variable and context-sensitive rather than participating in a stable system of symbols. Furthermore, experiments reported in Spivey, Richardson, and Gonzalez-Marquez (2005) show the smooth and pervasive interface between on-line language use and sensorimotor processes. Broadly speaking, research reported, among others, in Spivey (2007) and Pecher and Zwaan (2005a, 2005b) supports some early suggestions made in Gibbs and Matlock (1999) that tracing complete conceptualizations in the brain may not be possible because of the dynamic and continuous nature of such processes. At the same time, the research gives experimental support to two major claims of cognitive linguistic theories – one, that meanings emerge out of the basic levels of embodiment, and, two, that meaning is constructed on-line, for the needs of the specific act in which the conceptualizer is immersed, which does not preclude conceptualizations of situations not directly available in embodied experience. In other words, meaning is embodied and dynamic, even when what
26
Language and literary narratives
is being conceptualized is abstract, decoupled from the circumstances of the act of conceptualization, or leading primarily to an emotional response. Crucially, theories of embodied meaning do not require that arguing for a connection between experience and conceptualization puts one automatically on the side of extreme empiricism. While the authors who argue for embodied roots of abstract thought or emotions from a philosophical (rather than psychological) perspective (Johnson 1987, 2007; Lakoff and Johnson 1980, 1999; Damasio 1999, 2003; and Prinz 2002, 2004) agree that mind/body dualism is a convenient fiction, they are centrally interested in the emergence of conceptualizations, rather than in defending some form of empiricism. Even when experimental work confirms the role of image schemas in higher cognition, what is properly being described is the role of basic spatial configurations emerging from experience in concept formation. In other words, the upshot of the research is explaining the reason why an abstract word like status evokes the upper end of a vertical scale and a spatial expression like up top does too. Here is where van Oort appears to misinterpret the cognitive paradigm, by assuming that what is being said is that a use of a spatial preposition to describe a spatial configuration relies directly on perception without symbolic conceptualization, while its abstract meaning does not, whereas what is argued is that all kinds of meanings emerge on the basis of perception concepts. It is thus indeed possible to argue that basic spatially and temporally situated actions can serve as sources in conceptualization of abstract or displaced concepts, such as the ones underlying basic narratives – which has been Turner’s claim, formulated at a level closer to language than to cognition. The claim does not need to involve any hidden or explicit assumptions of innateness, because conceptualizations which emerge are not static or determinate enough to be candidates for such status. In general, claims of an innate mechanism related to language are not corroborated by experimental research. Second, while the symbolic nature of language is beyond dispute, the participation of symbols in various relationships with other symbols does not have to assume either discreteness or systematic correlations (such as binary oppositions, semantic features, et cetera). Also, our model of meaning has to capture the processes whereby new meanings emerge out of contextually prompted combinations of old meanings, and frameworks such as construction grammar and blending have made significant progress in that direction. Perhaps most important, though, treating symbols only in opposition to indexes and icons misses the point, for it creates an impression that only indexes and icons are linked directly to experience. While they have their own roles to play, symbolic concepts, as has been amply demonstrated, rely even more specifically on embodied experience. In other words, symbols are not detached from the real world; on the contrary, they are immersed in it via their reliance on image schemas and action. Categories and symbolic concepts are thus related to other
1.6 Narrative and grounding
27
categories and concepts not on the basis of the system of related meanings, but on the basis of underlying shared experiential concepts. To give just one example, if Spivey’s eye-tracking experiments reveal minute eye-movement upwards as a shared and consistent response to words such as up, tall, and respect (in clear contrast to down or submit), the connection thus revealed is among various applications of the basic image schema of verticality, from movement or location, through scalar measurement specific to cultural norms (a tall pygmy is a short Masai), to a social relationship. As an additional bonus, the proposed experiential treatment of meaning does not exclude cultural differences from the actual conceptualizations. Returning now to the concept of ‘narrative,’ we might say that in the context of the discussion above, it is not substantially different from any other form of conceptualization in that it relies on the dynamic emergence of representations prompted by the text (in terms of both the events portrayed and the emotional response to those events), and that it relies on embodied cognition. The specificity of how narrative chains of events are held together has much to do with the processes identified by Herman (2003a), but these processes are also essentially supported by experiential concepts. If I think of a portion of a story as ‘giving offence,’ I have to rely on the force-dynamic interaction between the participants, the emotional impact of their words and actions, the need to restore balance, et cetera. The fact that the events to be conceptualized are fictitious is not an obstacle to conceptualization. Naturally, narratives make heavy demands on cultural knowledge, memory, empathy, and other aspects of cognition, but they are not encumbered by the need to refer to discrete symbolic representations as independent steps in the process. Turner’s claim about the centrality of narrative thought and projection to conceptualization of any kind should not be read to suggest some evolutionary process which led to the emergence of the narrative module in the mind. In fact, claims about the evolutionary history of specific human abilities are becoming common. Since Fodor’s early work (1983), followed by the Chomskyan paradigm in linguistics and Pinker’s (1994) ‘language instinct,’ the search for modules in the brain to explain essentially human abilities seems to continue, even if jocularly. For example, a recent book by Dutton (2009), The Art Instinct, adds to the ‘instinct’ metaphor by considering art as a specifically human trait, doubtless originating in some evolutionary event or other. While any discussion of whether such innate modules in general are indeed a defensible explanation of the central aspects of human cognition is beyond the scope of this book, there is clearly no need to postulate a narrative module. Is there, then, any connection between basic narratives and evolution? While some recent work (Carroll 1995, 2005; Gottschall and Wilson 2005) connects evolution and literature, as well as evolution and art (Boyd 2005), these researchers make no claims about narrativity as such, and yet ‘narrative
28
Language and literary narratives
thought’ seems to pervade conceptualization. And as van Oort suggests, Turner’s idea of the ‘literary mind’ might be read as saying something similar. But while the pervasiveness of narrative is a fact, I hope to have shown that there is no need to postulate a narrative module or ‘instinct.’ Narrative-based conceptualizations may be indispensable (so I tentatively agree with Turner here), but they are readily explicable in terms of lower levels of embodied cognition. Also, while language is naturally treated as the basic environment in which stories exist, it is not indispensable to narrativity, since a story can be ‘told’ through visual means. The concepts required for a narrative to emerge (sequentiality, causation, chunks of experience, cultural framing of such chunks, imageschematic force-gestalts of conflict and restored balance, et cetera) are the same concepts which are required for other language conceptualizations to emerge, so we should perhaps advocate a stronger claim, such that narrative form relies on a specific cluster of such concepts. Moreover, recent work in both cognitive psychology and evolutionary theory allows one to talk in a sufficiently rigorous manner about human abilities which are not only cognitively central, but also culture-based. The work by Donald and Deacon mentioned above speaks of ‘coevolution,’ and the idea is given a more specific meaning in the theory of gene-culture co-evolution.11 The focus on cognitive sources of culture can at least suggest ways in which a centrally cultural non-primitive concept such as the ‘narrative’ may have become crucial to higher cognition, while remaining reliant on more primitive image schemas of force, sequence, or balance, on specialized language forms (extended structures which connect sentences in a variety of complex ways), and on the culture’s need to represent events and cultural norms in forms which guarantee their temporally extended availability. This discussion has not yet addressed a related but largely independent question concerning the cultural and linguistic emergence of narrative fiction. Zunshine’s work (2006) locates the explanation in evolutionary psychology and more specifically in the so-called ‘theory of mind’.12 The term is best known in the context of Baron-Cohen’s (1995) generally accepted explanation of communicative deficiencies summarized in the term ‘Autism Spectrum Disorder.’ At least some of the difficulties of autistic speakers are attributed to their lack of an understanding of how language and behavior can be read in order to grasp people’s mental and emotional states and adjust one’s language and behavior to meet their expectations. Zunshine builds on this ability to ‘read’ other minds to argue that fiction offers a form of ‘exercise’ in using it. There are some problems in such a use of the concept. First of all, ‘mindreading’ includes (and is tested with the use of) understanding of communicative components such as eye contact, facial expression, or body posture, and not exclusively language. It is also being used in explaining a variety of behaviors, including appreciation of humor, irony, and figurative language.
1.7 Approaching narratives
29
Reading cannot test all these skills, or serve as practice, because it relies on a narrative rendering, not lived experience, and is mediated entirely through language. If there is any disposition commonly associated with good mindreading which can also be applied to fiction, it is the ability to grasp the emotional impact of people’s words and actions. Interestingly, appreciation of mental and emotional acts has already been claimed to be the single most salient feature distinguishing narrative fiction from other narratives. Palmer (2003) argues that ‘fictional minds’ are the core of narrative discourse, even if narration presents mostly actions – this is what fiction is about, and the plot is in fact secondary to this primary focus; a similar point has also been made in the philosophical context by Currie (2007). I agree with Palmer because his observations elucidate the role of much of what needs to be narrated in a novel, but also because they explain how fiction stimulates an emotional response in the reader without relying on the whole constellation of abilities roughly summarized under the rubric of ‘theory of mind.’ Palmer’s claims can also be supported with recent research on the role of embodiment in representing mental states (cf. Zlatev et al. 2008), and I will return to the question in Chapter 4. However, accepting Palmer’s argument does not require a sharp distinction between, let’s say, narratives in oral traditions and narrative fiction as we know it. Any narrative form assumes the readers’ or listeners’ ability to feel for the characters and interpret the events from the characters’ point of view – to feel sympathy, envy, disgust, anger, et cetera. While some of these emotional responses may be complex, or even conflicted, they are better explained through our ability to feel empathy, and also to attribute emotional meaning to linguistic expressions – even simple experiential concepts such as up (alert, conscious, confident, in good mood) or down (tired, inactive, sick, unconscious, dejected). To conclude, I will treat extended fictional narratives – the primary focus of the project – as relying on our abilities to construct meaning and respond to it emotionally. But I will also assume that these abilities emerge, more or less in the ways described by Tomasello (2008), from the simplest linguistic forms based in embodied cognition, through more and more complex narrative forms, all the way to the complexities of contemporary stories. I will also argue, mainly in Chapter 6, that narrative ways of representing mental and emotional states of characters have evolved through literary periods and genres towards a more and more direct, and thus narratively more complex, access to fictional minds. 1.7
Approaching narratives
As the discussion above has indicated, narratives and narrative fiction have been subject to enquiry in a number of different disciplines, from psychology and cognitive science, through philosophy, literary criticism, stylistics and discourse analysis, to linguistics and narratology. Each of the disciplines contributes
30
Language and literary narratives
much to the discussion, while also having its own goals and methods. Where does this project locate itself then? The simplest answer would be – towards the linguistic and cognitive end of the spectrum. I want to propose an analysis which is informed by current research in cognitive science and which clearly addresses the linguistic phenomena specific to extended narrative texts. I will refer repeatedly to the concepts proposed by narratologists, but I also want to put this discussion in a context which is more open and more restricted at the same time, as it seeks to explain processes of conceptualization by looking at the fine-grained linguistic prompts for these conceptualizations. At the same time, the tools proposed should be useful in a discussion within stylistics, where language is discussed in terms of specific textual choices and their impact on the special quality of the individual text or genre. Like some previous work in cognitive discourse analysis (Emmott 1997), this project will also attempt to look at features of textual matter which contribute to text-wide aspects of interpretation. The hardest relationship to define is between this work and literary criticism. While I hope to open the linguistic tools to the role of cultural values or historical facts, I cannot make them central to the discussion, and thus much of the interests of literary critics will be downplayed. However, my interest in paths of interpretation is what critics should recognize as relevant to their field. There is also the question of pedagogy. In my work with students of literature, I found that introducing them to the concept of a frame or an image schema often gave rise to new and innovative interpretations, not because the concept could be found in one text or another (it would probably be found in any), but because of the special salience of some basic cognitive configuration in a text. In other words, they noticed things that required interpreting because they were highlighted once a cognitive concept became available. So cognitive analysis did not seem to restrict the playing field, it in fact opened it up. This book will thus focus on the link between language and cognition and the ways in which such correlations affect the construction of narrative meaning. First, I will outline the ways in which tools developed in the theory of conceptual integration can be applied to longer texts. I will start the discussion of narratives with situating the text-wide construal of narratorship in the conceptual integration model. The central part of the book will focus on discussion of viewpoint, looking at some ways in which sentence-level lexical choices and referential expressions affect narrative viewpoint. I will then move on to the role of representation of fictional minds in various storytelling genres, including the discussion of various types of discourse in the narrative, with special focus on the role of direct speech and thought. Finally, I will resume more general discussion in the concluding chapter.
2
Blending, narrative spaces, and the emergent story
When he woke up, the dinosaur was still there. (The Dinosaur, Augusto Monterroso)
Stories can, naturally, be treated as linguistic constructions. They are attested examples of natural written discourse, and, while it is possible to treat linguistic choices in the narrative as primarily rhetorical, there is also crucial reliance on very basic facts of language structure. Stories should also be talked about in terms of meaning. They are particularly challenging in this respect, if only because of the sheer length of the text and the variety of narrative options available. This project proposes a framework which makes the discussion of the meaning of stories possible. It relies heavily on analytic tools established within the discipline of cognitive linguistics (mental spaces theory, construction grammar, and frame theory), but it is primarily focused on the theory of blending, or conceptual integration. Fictional narratives rely on a variety of linguistic phenomena, but adapt them to their needs. To give just one example, they use personal pronouns and deixis in ways which exploit the default spoken setting, but set it against a different kind of deictic ground. Thus a pronoun like I is often used in novels to represent the narrator – a narrative construct, rather than an actual communicator. The discourse setting is that of a communicative act occurring across a spatial and temporal distance, with no clear addressee available (other than some generic reader). The consequences of such a set-up are farreaching, as they create the need for a different treatment of referential expressions designating characters and put the concepts such as the ‘present’ or the ‘past’ in a different context. These peculiarities will be discussed throughout this book. With respect to meaning complexities of fictional narratives, blending (cf. Fauconnier and Turner 1998a, 1998b, 2002) seems to be particularly well-suited to proposing an account of narrative structure. The concepts and mechanisms proposed in blending work translate naturally into the needs of an account of narratives, but also open the door to the discussion of blending mechanisms and properties which seem specific to the emergence of meaning 31
32
Blending, narrative spaces, and the emergent story
of longer texts. In what follows, I will elaborate the blending framework in ways which best respond to the needs of accounting for narrative discourse.
2.1
Applying blending to fictional narratives
The essential claim of blending is that conceptual packets (mental spaces or frames, here called inputs) can be activated in a person’s mind by verbal as well as nonverbal prompts and integrated into a new conceptual configuration (called a blend) to construct new meanings as discourse develops. The conceptual structure of the blend, while relying on projection from the inputs, achieves its own coherence by selecting only the relevant parts of the inputs and by compressing vital relations, such as time, causation, or identity. For example, while the inputs are two mental spaces one of which is present and the other past, the temporal distance between them may be compressed to yield a blend which is either past or present. The inputs remain activated and retain their separate conceptual status, but a specific blend may also project structure back to the inputs (backward projection), and thus reconstrue them for the purposes of the ongoing exchange. The blend is characterized by its own structure (emergent structure), but can then become an input to another blend, or series of blends. While blending can account for ongoing processes of meaning construction (as in following the course of a conversation or reading a comic strip and accumulating content until the joke gels in the final frame), they can also become entrenched as new expressions, such as compounds or idioms. Blending also accounts naturally for multimodal contexts, where visual, linguistic, and aural prompts work in combination, as in film. The way the blend emerges is largely dependent on the topology (internal structure) of the inputs. While in a number of cases inputs are mental spaces, they may also be structured by frames (as introduced in Fillmore 1985, 2006; Lehrer and Kittay 1992). For example, a blended construction such as When I learn to surf, I will move to California assumes the availability of two mental spaces: the future space in which I am a surfer and I move to California, and the present reality space where it is not the case. Each of these constructs has the structure of a mental space, with participants, temporal and spatial features, and internal structure (such as a connection between California coast and surfing opportunities), and they yield a blended space wherein being a surfer in California also retains other aspects of my identity: family ties, life story, et cetera. Crucially, the imagined future situation remains a mental space precisely because it has been constructed on-line, that is, within ongoing discourse, to indulge in an ‘alternative life’ for the duration of the conversation about surfing, and it will be deactivated when the discussion of perfecting surfing skills stops or when the topic is changed.
2.1 Applying blending to fictional narratives
33
Frames are significantly more stable, in that they structure our understanding of reality beyond the conversational context and are often associated with specific lexical items. While we all have the general concept of dying, we apply very rich cultural framing to concepts such as killing, murder, infant mortality, accidental death, or genocide. Each of these includes the fact of a person or persons dying, but they also include complex scenarios specifying the circumstances, causes, the degree of agency, cultural and social realities, et cetera. We can thus say that they participate in different frames, and that these frames are associated with assignment of guilt, legal consequences of guilt, moral values, medical problems, et cetera. Using any of the expressions listed above evokes all the associated framing, and, by the same token, appropriate social actions. The most important aspect of framing is the possibility of accessing the entire frame when only one aspect of it is mentioned. In the most obvious case, the discussion of buying a new car automatically implies the aspects of the frame which are not explicitly mentioned: selling, transfer, object transferred, price of transfer, et cetera. In the less obvious cases, when an aspect of the frame is chosen in a less predictable manner, we can also talk about frame metonymy – linguistic usage whereby an expression is used to stand for an associated aspect of the frame, while also evoking the entire frame. Cases of such pragmatically established metonymies were discussed, among others, in Fauconnier (1994 [1985], 1997). Fauconnier analyzes examples such as Room 4 has a visitor, which can be said by a member of hospital staff to refer to the patient in room 4 and announce that a visitor has arrived. In Dancygier and Sweetser (2005), the concept of frame metonymy has been expanded to include constructional frames, so that, for example, a specific use of the present tense to mark an unpredicted premise in a predictive reasoning signals predictive meaning even if no other constructional features are marked. Thus, a sentence such as You miss one more meeting, you’ll be fired counts as a prediction of dire consequences if an as-yet-unpredicted event of being late occurs, without using a conditional predictive construction in its entirety – the verb forms are enough for the frame to be evoked. In recent work (among others, Dancygier 2009), I argued that frame metonymy at the lexical level, when paired with frame metonymy at the constructional level (termed constructional compositionality), creates blended constructional patterns in which meaning emerges based on the specific blend between the evoked lexical frame and the constructional frame. For example, a sentence such as Iraq is a new Vietnam seems to follow the format of a copular construction in which the subject is a specified entity and the predicate is a non-specific name of a category (as in John is a teacher), but both proper names are used to evoke metonymically frames of US interventions and extended military conflicts (rather than ‘countries’), and thus the adjective new signals the
34
Blending, narrative spaces, and the emergent story
transfer of some of the framing of Vietnam (high cost, lack of success) into the framing of Iraq. As a result, Vietnam does not become a common noun (even though the frame is detached from its standard referential meaning and an indefinite article is used), while Iraq counts as metonymic for the Iraq war, rather than the country. The interaction among the frames evoked by both proper names, the adjective new, and the construction yields a blended meaning not available outside of the construction. I will discuss framing and reference in more detail in Chapter 5, but the connection between frames and blends has thus been briefly illustrated, and will be relied on in further discussion. A simple example of a blend is a compound expression such as computer literacy. The two inputs are frames of being (one) a computer user and (two) a person able to read and write. The frames are rich culturally. While literacy used to be considered the cornerstone of participation in contemporary society, computer use until recently was not considered an indispensable skill. The compound expression blends the two concepts to refer to a set of computer skills which are now considered essential. They are not restricted to reading and writing, but involve also an understanding of basic computer functions. While the ‘computer’ input contributes the understanding of which skills are basic, the ‘literacy’ input further qualifies those skills as required of anyone who wants to be considered minimally educated. The blend also informs both inputs, by defining computer skills as basic, and expanding the scope of literacy to match contemporary needs. Interestingly, the expression is then cleverly reframed in the name of a well-known Silicon Valley institution – Computer Literacy Bookshop, where the expression computer literacy refers to interaction with written language again, even though only computer professionals are the intended readers. It is thus a good example of how subsequent blends manipulate both the inputs and the earlier emergent meanings. A more complex blend is prompted by a clausal construction such as If I were your mother, I would teach you how to cook. It sets up an imagined situation which is a blend of the if-space (the speaker is the hearer’s mother) and the reality in which the speaker and the hearer participate (the hearer cannot cook, her mother never taught her, et cetera.). The emergent structure also creates a new imagined identity blend, wherein (under one of the possible interpretations) the hearer’s mother shares the commitment of the speaker to teach the hearer to cook. The blend is complex, but the mechanism that allows it to emerge is the same simple mechanism, and, crucially to the subject of this book, operates on inputs which are rich in contextual information, feature various participants, and present events and human dispositions as causing other events and dispositions. We can assume that to the speaker and hearer of the conditional all the contextual implications are known – the desires and personalities of the participants, the reasons why knowing how to cook is useful, and the reasons why the speaker cares about the hearer’s skills.
2.2 Narrative spaces as mental spaces
35
Similar meaning construction processes occur in longer narratives. Frames and mental spaces structure inputs, which then become integrated, possibly in ways specific to a reader, into the emergent blend. The process continues throughout reading, until the complete blend of the story emerges. In what follows, I will consider some details of the process. 2.2
Narrative spaces as mental spaces
Mental spaces were defined (Fauconnier 1994[1985], 1997) as conceptual packets set up primarily by linguistic expressions (though visual or gestural prompts are also common), and being used in the on-line mental processes of meaning construction. Thus a mental space may be a recall of a past event in the course of a conversation about the current situation. For example, if the speaker says I was really tired last night, but I’m better now so do tell me what you think about my book she calls up a past space in the first clause, describes her present state in the next, and evokes the spaces of the content of the book, the content of the hearer’s thoughts regarding the book, and the hearer’s expected conversational turn. All these are mental spaces, activated for the duration of this part of the conversation, and then becoming latent until evoked again, when, let’s say, the speaker later complains about the hearer’s bitter criticism to someone else. It is clear even from this example that mental spaces are constructs of varying complexity (from tiredness to the content of an entire book), attributable to various conceptualizers (here, the speaker or the hearer), and including mental representations of concrete objects, conversational moves, opinions, et cetera. It is often asked whether it is possible to confirm the presence of mental spaces in the brain, but the question seems to misconstrue the usefulness of the concept. Even if we could monitor someone’s spontaneous engagement in a conversation and discover which areas of the brain are involved (some have to be!), we could at best capture some of the brain structure relevant to the content of the conversation, but apparently not the thought processes prompted by the discourse. Also, as Spivey argues (2007), the flow of thought, when traceable in the brain, creates a trajectory which maintains a number of open options throughout – very much in agreement with what mental space theorists claim. Mental space analyses have shown how mental space networks can represent the on-line correlation between choice of expression and the resulting meaning.1 But mental spaces are not a strictly linguistic phenomenon, as recent work on gesture, art, comics, theatre, or film shows;2 rather, the theory is a useful tool in teasing out the details of various forms of meaningful interaction and discovering the correlations between such prompts and meaning. The above discussion of mental spaces and blending makes it clear that the framework has been mostly used to elucidate the emergence of meaning of expressions, discourse fragments, or pictorial/verbal artifacts. Fictional
36
Blending, narrative spaces, and the emergent story
narratives, for comparison, are much more complex and also simply longer. It does not seem possible to provide a clear explanation of how individual sentences may be subjected to the overall storytelling goals. To address this concern, I will argue that narrative construction proceeds through several levels of narrative organization, involves an emergence of partial narrative structures called narrative spaces, and is primarily organized by various forms of narrative viewpoint. A narrative space has most of the features of a mental space.3 It is set up through language expressions, often ones not unlike ordinary space builders – such as Once upon a time . . . or In a hole in the ground there lived a hobbit . . . and further structured by grammatical choices such as person or tense. It has its topology, such as space, time, cultural norms, or language spoken. It also has participants, some directly involved in the telling, called narrators, some being directly affected by the events, called characters, though there are also participants (intradiegetic narrators) who play both roles at the same time. Like mental spaces, narrative spaces can be elaborated in much detail, while others are barely mentioned. Both types of spaces are structured by viewpoint phenomena, though narrative spaces seem to feature more elaborate constructions of viewpoint, to be discussed throughout this book. To sum up, a narrative space is a mental construct participating in the emergence of the story, having distinctive topology and narrative status, and linked to other narrative spaces in ways which prompt story construction. Narrative spaces are different from other ‘worldlike’ concepts such as possible worlds (as applied to narrative) which are discussed in relation to the ‘real’ world as its alternatives.4 Narrative spaces are defined in parallel to and in correlation with lower-level linguistic phenomena participating in meaning construction, and, while they may apply to ‘worldlike’ constructs, they do not apply to texts in their entirety. An emergent story results from the blending of all of the text’s narrative spaces. The structure of narrative spaces is thus determined by the specific organization of the text, and no claims are made about their necessary completeness or coherence. The concept of a mental or narrative space also shares some features with the construct which has been termed a ‘text world’ (introduced by Werth 1999, and developed by Gavins 2003, 2007). Text world theory shares some of its assumptions with the mental spaces framework – for example, it also talks about linguistic forms being used to prompt conceptual constructs and accumulating as discourse progresses. It is primarily a theory of the interpretation of discourse, and has been applied to a number of discourse contexts – verbal and nonverbal – including the context of the narrative. The primary difference is possibly that the text world theory is much more text-driven than the mentalspaces-and-blending framework, while also being more focused on the analysis of discourse, rather than the elucidation of linguistic concepts. Mental spaces
2.2 Narrative spaces as mental spaces
37
theory has originally been conceived as a theory of meaning, and, while it is developing its textual applications, it still relies more heavily on semantics than on discourse. Also, mental spaces analyses are naturally extended into work on blending. I will be relying on mental spaces specifically, but I assume most claims can be ‘translated’ into the text world approach. A narrative space is thus a construct which is set up through linguistic means and continues being elaborated through some parts of the text (possibly all). It is also subsequently enriched through blending and gradually starts functioning in the network leading to an emergent story. In these respects it is thus similar to a mental space, which participates in extended discourse. What constitutes a crucial difference, however, is the nature of the discourse, since a narrative requires that its primary spaces be maintained and elaborated until the completion of the reading process, until their role in the text is fulfilled. To give a simple example, if a flashback space is set up, it can then play different narrative roles, not necessarily dependent on time, though in each case it would constitute an independent, temporally distinguished space. Setting up such a space can constitute background to the main story, or the beginning of the story proper, or a digression which explains some form of behavior. In any such instance the space will be incorporated into the story, though not only temporally. 2.2.1
Sentential level versus textual level
It may seem counterintuitive to claim that a narrative space may in some cases take the form of a sentence, but there is no reason to draw a specific line at any level of linguistic structure, because what matters is the narrative structure. For example, the text which won the contest for the shortest story is also a sentence. It was composed by Augusto Monterroso, and it is entitled The Dinosaur: (1) When he woke up, the dinosaur was still there.
The story has all the ingredients necessary, but asks the reader to do much of the story-construction work. There is a narrator (probably omniscient), and a character, plus a dinosaur. The when clause profiles a temporal location wherein the character awakens, which suggests that he was asleep before (in Spanish it might have also been a ‘she’). The word still in the main clause further suggests that before the character fell asleep, the dinosaur was also present. There is thus a sequence of events, there are characters in some relationship, there is a narrator, and there is suspense – how come there is a dinosaur in the presence of a human being? What has often been called ‘suspense,’ in my view, is the need to engage with the text and have the intention to complete the gaps in the story. In other words, the reader is required to use extra-textual and linguistic knowledge to complete the emergent structure of the blend being prompted by the narrative inputs.
38
Blending, narrative spaces, and the emergent story
In the case of Monterroso’s story, there are in fact several options. The dinosaur could be a real T. rex or triceratops, and then the story qualifies as science fiction, and the character is possibly in grave danger (hence the fact that he could sleep in the presence of a monster is intriguing indeed). Or it is a fantasy about a friendly dinosaur, who guards the character when he sleeps. If the spaces set up are not intended to represent the story’s reality but the character’s delusions (think ‘crazy scientist’), then the story might be a variety of ‘stream-of-consciousness,’ following the inner visions of the character. If the dinosaur appeared in a dream, when the character slept, then the reader is faced with a mystery – did the dinosaur cross over from dream to reality? or is the character still asleep, but dreams that he is waking? In fact, there is an impressive number of possible readings of the story, all dependent on how the reader (not the text) construes the spaces set up, and how the spaces are then linked by, for example, the presence of the same characters (‘he’ and dinosaur in the same ‘reality’ or not, dinosaur projected from dream to ‘reality’, et cetera). All the interpretations above are blends – a reality/fantasy blend, a reality/hallucination blend, or a reality/dream blend. What is more, as the ‘dream’ version suggests, it may not be possible or even interesting to resolve what kind of blend it is, so long as the story construction process engages the reader. Even in a longer narrative, the sentential level of the text may crucially contribute to the structure of narrative spaces. Consider example (2), where the main character in Eggers’s novel A Heartbreaking Work of Staggering Genius, Dave, is in his car, struggling with guilty thoughts about having left his small brother with a babysitter: (2) I will come home and the door will be open, wide. The baby-sitter will be gone and there will be silence . . . Blood on the walls, handprints soaked in blood. A note to me, from Stephen, taunting; . . . There will be a hearing, a trial, a show trial – How did you come to meet this man, this baby-sitter? We found a posting, on a bulletin board. And how long did your interview of him take? Ten, twenty minutes. (AHWOSG, p. 126)
The narrator is assigned to an independent space (story-viewpoint space; SVspace), which has the entire story in its scope. The story itself is contained in the main narrative space (MN-space), which consists of a number of narrative spaces (NS). The details of this set-up will be explained in Chapter 3, but example (2) illustrates what it means for lower-level expressions to participate in narrative space construction. The main narrative space in (2) is that of Dave’s life with his brother, Toph. As Dave is driving across Bay Bridge, his mind starts conjuring up an elaborate scenario wherein the babysitter kills Toph, and Dave is then put on trial for negligence. The sentence I will come home . . . thus sets up an imagined future
2.2 Narrative spaces as mental spaces
SV-space
39
Narrator: Dave present
MN-space
Narrator: Dave / ‘I’ present tense Dave’s life
NS1 Narrator: Dave / ‘I’ driving to San Francisco NS2 – crime Narrator: Dave / ‘I’ imagined future babysitter (killer) Toph (victim)
NS3 – trial Narrator: Dave / ‘I’ NS3 – Discourse space S1: Dave S2: prosecutor
Figure 2.1 Narrative spaces (example 2)
space, then developed with the gory details of the crime. Then, from the point of view of the imagined ‘crime’-space, another space is built, that of an imagined trial. Each one of these spaces is set up at the sentence level, and even though their activation does not extend beyond several sentences, they do play a role in the narrative construction (see Figure 2.1). First, the scenario developing in Dave’s mind is one of numerous instances in the novel where Dave’s unrealistic fears and his lack of confidence lead to erratic and exaggerated responses. The ‘crime-and-trial’ sequence of spaces is thus a part of a broader narrative construct representing Dave’s life. While seemingly local, it participates in the higher-level story construction process, which does not have to proceed linearly, in the way the text does, but elaborates spaces incrementally, throughout the text. Second, the example represents one of the key issues of story construction – the management of viewpoint. In terms of mental spaces theory, the spaces just described represent familiar patterns of viewpoint and focus. The space of Dave’s life is the viewpoint space from which the ‘crime’ space is being set up and elaborated, as Dave’s fantasy – it is thus the focus space. As discourse progresses, the ‘crime’ space takes the role of viewpoint space, and the ‘trial’ space is now in focus, and being elaborated. This viewpoint/focus structure explains why the trial is construed as the result of the imaginary crime, and not, say, of Dave’s driving habits.
40
Blending, narrative spaces, and the emergent story
The viewpoint/focus configuration explains how spaces form chains, such that a higher-level space is the viewpoint, and a lower-level space is the focus. However, in order to explain the processing of narratives as complex as novels, we also need an explanation of the mechanism which links various lower-level spaces into clusters contributing to the construal of higher-level spaces. I will discuss this mechanism as cross-input projection. At the same time, the content of the lower-level spaces contributes to the construal of the higher-level space, so that Dave’s fantasy is not only projected downwards from the highest space of his narrative, but also used in the construal of the narrative itself – for example, it gives the reader access to an understanding of what kind of narrator Dave is, and how his narrative as a whole relies on fact and fantasy alike. These higher-level construals result from what I will refer to as viewpoint compression, an integration mechanism which allows lower-level viewpoint to contribute to the higher level. The phenomenon also has its linguistic correlates, such that lower-level spaces are presented without linguistic indications of their being embedded in higher-level spaces. The entire fragment partially quoted in (2) is narrated alongside the description of Dave’s drive, and thus, stylistically at least, it is presented as belonging to the core of the story being narrated. The effect of engaging multiple levels of spaces and presenting them as one is the primary sign of viewpoint compression. I will discuss the phenomenon in more detail throughout the book. Beyond the sentential level, fictional narratives set up various configurations of spaces, distinguished by basic space topology (such as time and space, or participants). Some contemporary narratives rely on an even more fragmented structure, wherein major narrative spaces are presented as independent texts – there can be a novel within a novel, as in Atwood’s The Blind Assassin, different narrators, or stories ostensibly consisting of separated narratives, to be linked at the end, as in Alex Garland’s The Tesseract or in the movie Babel. Accounting for the astounding variety of such narrative choices would not be possible in any analysis, but I will discuss one text, The Blind Assassin, in more detail, to illustrate the processes involved. 2.3
Narrative spaces – an example
Margaret Atwood’s novel The Blind Assassin is a good example of how the multiplicity of narrative space operates in a narrative. The novel’s framing narrative is a memoir written by a now aged and lonely woman, Iris Griffen (née Chase). She lives in the small town where she grew up and writes the story of her life, to be read after her death by her estranged granddaughter. The main narrative space thus has a first-person narrator, Iris, remembering the past. Consequently, large parts of what we are reading are descriptions of Iris’s daily struggle to go on writing, but the main story of the novel is reconstructing
2.3 Narrative spaces – an example
41
Iris’s entire life. These two narrative spaces of ‘Iris’s present’ and ‘Iris’s past’ are the core of the narrative structure. Also embedded in the text is a novel, titled “The Blind Assassin,” ostensibly written by Iris’s sister, Laura. It is written in the present tense, with just two characters never identified by name – just a ‘he’ and a ‘she.’ These grammatical choices (no I narrator, present tense) help the reader maintain the embedded novel’s relative independence, even as it is gradually incorporated into the story. In the embedded novel, the two anonymous characters compose a pulp sciencefiction story, taking place on the planet Zycron, and featuring a blind assassin as one of the (again nameless) characters. This story is never completed, and thus its role in the entire text is difficult to see. Finally, the text is sometimes interrupted with short newspaper clippings, giving an official version of some of the events. The clippings often contradict the information provided by Iris’s memoir. They jointly represent a narrative space telling the official version of some of the events, reappearing throughout the text. The space is anchored by the format of the clippings and is reactivated as the need arises. Taken together, the clippings construct a narrative space which tells a story different from the one narrated in the rest of the text. All these major spaces are presented as independent texts, with different authors and characters, occurring in different locations and times. As Figure 2.2 shows, they are initially set up as independent spaces – and the set-up changes
SV-space
Narrator: Iris present
MN-space
Narrator: Iris Author: Iris past tense Memoir Narrator: Iris past tense Laura Alex Richard Father
Clippings Author: journalist Mrs. Richard E. Griffen Embedded novel Author: Laura Narrator: 0 present tense he she
Sci-fi story Author: he Narrator: 0 present tense mute girl (she) blind assassin (he) Lord of the Underworld King
Figure 2.2 Narrative spaces in The Blind Assassin
42
Blending, narrative spaces, and the emergent story
as the text progresses. Crucially, they all have to be treated as subordinate to the space of Iris narrating (they are within the scope of the viewpoint of the SVspace), and are gradually identified as complementary to the memoir she writes. Coherence links are suggested throughout and the fabric of the story emerges through various links across the narrative spaces. Even more important, all the viewpoints of all the spaces are finally compressed with the viewpoint of the initial set-up with Iris as the narrator. The case of The Blind Assassin illustrates all the central processes of story construction. I will attempt to elaborate them below, starting with the concept of narrative anchors. 2.3.1
Narrative anchors
I have introduced the concept of a narrative anchor in earlier work, focusing mainly on The Blind Assassin and F. Scott Fitzgerald’s The Great Gatsby (Dancygier 2007, 2008a). In both of these cases I argued that the fragmented nature of most narratives requires specific narrative mechanisms which provide coherence links across different narrative spaces. I defined narrative anchors as expressions which set up or suggest the availability of narrative spaces, but do not elaborate them right away. Such ‘place-holders’ may activate new narrative spaces and allow them to remain active, but the spaces are elaborated gradually as the text unfolds, and often contribute to the topology of other spaces constituting the story. The second function of anchors is ‘link-building’. The links they set up may prompt what I have called cross-input projection – spreading of topology from one input to another and building the coherence and completeness of the emergent story. To refer to the ‘clippings’ again, they are constantly crosslinked with other spaces, adding topology, foreshadowing events to be narrated later, but, first of all, presenting a viewpoint not available in other narrative spaces. The usefulness of anchors becomes clear when we realize that it is difficult to imagine a fictional narrative which would explicitly provide the entire content of its narrative spaces. Narrative coherence crucially depends on selection of the content needed at a given point in the story, and on suppression of all information which would reveal too much or drown the important narrative elements in a chaos of unbounded verbosity. Narrative anchors are difficult to define formally because they are not restricted to any specific aspect of narrativity which is typically discussed (plot, characterization, et cetera), but they may participate in all. They are best defined in terms of the processes they trigger which lead to the construction of the story. Various expressions can play the role of anchors, and the choice depends on the type of cross-space connection to be established. One can thus expect that well-selected referring expressions would lead to connections in terms of
2.3 Narrative spaces – an example
43
finding counterparts in different spaces and establishing identity (see 2.3.2 for more discussion), but descriptive expressions may play the same role. For example, the man named Alex in the main narrative space of The Blind Assassin has to be recognized as being the same as the one referred to only as ‘he’ in the embedded novel space, even though the two counterparts are characters in different stories, taking place in different space and time: (3) He stubs out his cigarette, reaches for another, thinks better of it. He’s still smoking ready-mades, a luxury for him. (The Blind Assassin, p. 133, embedded novel) (4) He produced a packet of cigarettes from his shirt pocket – Craven A’s, as I recall. He tapped one out for himself. I was a little surprised that he smoked ready-mades – it didn’t go with his shirt. (The Blind Assassin, p. 222, Iris’s memoir)
Making the connection is prompted by careful deployment of the same bits of information in both narratives, and by repetition of some of that information. Bits of characterization, such as (3) and (4), are spread throughout the narrative spaces via cross-input projections and jointly create the character’s identity which fits coherently in all the narrative spaces. Crucially for this example, as the page numbers show, the reader first sees the man in the embedded novel smoking ready-mades, but is also alerted to this trait by the use of still – which indicates a time prior to the scene when the same trait was observed. So when Alex Thomas is first introduced in Iris’s memoir (in a scene temporally prior to that of [3]) and the same information is presented as surprising, the reader can make the connection between the two characters. In fact, guessing Alex’s identity is made easy, but the identity of the woman remains elusive till almost the end of the novel. Anchors can also rely heavily on frames. As shown by Lermitte (2010), anchors may also be images which form an entire network of concepts and jointly give meaning to an abstract and difficult text. In Annie Dillard’s nonfiction text Holy the Firm the meaning is not so much a story as a reflection on religious and artistic passion and a life centered around such deeply felt emotions. The crucial image is that of a moth being attracted to the flame of a candle, dying, and acting as the candle’s wick as a result, allowing it to burn longer: She burned for two hours without changing, without bending or leaning – only glowing within, like a building fire glimpsed through silhouetted walls, like a hollow saint, like a flame-faced virgin gone to God. Passages such as this recur through the text, highlighting various dimensions of the emerging blends. The recurring images of burning as a sacrifice which in the end prolongs life allow Dillard to tell a poignant story of creative endeavors and religious devotions as acts in which an individual needs to ‘die’ in order to ‘keep the flame burning’ and go on living through what has been created. The anchors do
44
Blending, narrative spaces, and the emergent story
not simply ‘construct the story’ here, they construct its meaning through a network of blends and frame metonymies. Without them, the text achieves no narrative coherence. These various examples of anchors pose questions as to the actual cognitive nature of anchoring. It seems that it relies on a variety of possible construals and mappings. In the work done so far, I have distinguished several types. First, there are anchors which are quite naturally space builders. When the text of The Blind Assassin repeatedly refers to ‘notebooks’ a now dead character left in hiding, the content of the notebooks constitutes a narrative space which promises to fill huge gaps in the story. Naturally, when the notebooks are found, they do reveal crucial facts. Another type of anchor consists in evocation and repeated re-activation. A mention of Alex as wearing a blue, worker’s shirt but smoking ready-made cigarettes is a salient descriptive detail, and its reactivation in a different narrative space prompts the cross-input projection linking Alex to ‘him’ and making this cross-space identity available in both spaces. This type of anchoring relies as much on the reader’s attentiveness and memory as on the salience of the frames thus constructed and on the sheer number of anchors establishing and reestablishing the same cross-input links. Crucially, such anchors not only link the spaces, but also prompt projections of topology from input to input (narrative space to narrative space). The identity of the character named Alex in one input is cross-linked to the identity of an anonymous man in another, but the information from one input is thus also available in the other. Below I will present an extended analysis of a pair of anchors in The Blind Assassin which are central to the narrative meaning of the entire novel. Throughout the novel, these anchors play all of the roles described above – narrative space building, prompting cross-input mappings, establishing identity, metonymy, and metaphor. While this may be a special case, it does show the potential of anchoring in a complex story.
2.3.2
Anchoring: representation blends and frames
The main narrative space of the novel is the life of two sisters, Iris and Laura, and their involvement with a man, Alex Thomas. Iris is the narrator, while Laura, who has committed suicide, is not given a narrative voice. Even though one of the narrative spaces is a novel she presumably wrote, it turns out in the end that the novel was also written by Iris and published under Laura’s name. Throughout the reading, the text playfully alternates between suggesting Laura or Iris, leaving it to the reader to construct the true identity of the narrator of ‘Laura’s novel’ (I will refer to it as ‘embedded’). The authorship, in turn, also identifies the writer as Alex’s secret lover. While there are many salient anchors
2.3 Narrative spaces – an example
45
leading the reader to a conclusion, the main anchors are a photograph and a hand. They are initially mentioned in the first lines of the embedded novel: (5) She has a single photograph of him . . . the photo is of the two of them together, her and this man, on a picnic . . . he’s holding up his hand, as if to fend her off in play, or else to protect himself from the camera . . . Over to one side – you wouldn’t see it at first – there’s a hand, cut by the margin. (The Blind Assassin, p. 7)
The fragment is full of unanswered questions – who are the people? whose hand is visible at the edge? what happened to the man? and many more. They all emerge because of the fact that the concept of a photograph (similarly to picture, painting, drawing, et cetera) prompts a blend connecting two mental spaces: the representation space (distinguished by its form, colour, materiality, boundaries, et cetera) and the represented space (the actual reality it captures, with its temporal and spatial features, identity of participants, as well as, crucially, topology available beyond what the representation shows). In the fragment, only the representation space is provided, while none of the features of the represented space are clarified – hence all the unanswered questions which frame the entire ensuing narrative. It is thus a perfect example of a narrative anchor. It opens up a narrative (represented) space which is necessary to complete the representation blend prompted by the description of the representation space, and the anchor remains active until the blend is complete – until the end of the novel. The narrative of The Blind Assassin uses this anchor throughout to keep the reader guessing. It is reactivated regularly, and each time a new piece of the represented narrative space becomes available. The next mention, in Iris’s memoir, is as follows: (6) Then he took a picture for the paper with his camera . . . Alex Thomas raised his hand as if to fend him off. (The Blind Assassin, p. 223)
This appears in a fragment describing a factory picnic when the sisters first met Alex. The identity of the man in the photograph is suggested by the second mention of his gesture, but since both women were present at the picnic, the woman telling the story in the embedded novel is still mysterious. Finally, the newspaper report of the factory picnic is published: (7) One of the pictures was of Alex Thomas, with the two of us – me to the left of him, Laura to the right, like bookends. Both of us were looking at him and smiling too, but he’d thrust his hand up in front of him . . . The caption was, “Miss Chase and Miss Laura Chase Entertain an Out-of-Town Visitor.” (The Blind Assassin, p. 241)
The anchor of Alex’s raised hand is repeated again, but the identity of the woman in the photograph in (5) becomes even harder to discover. The fragment makes it clear that while the represented space is now known (Laura, Iris, and
46
Blending, narrative spaces, and the emergent story
Alex at the picnic) the representation in the photograph mentioned in (5) is incomplete. One of the women – Laura or Iris – is missing from the photograph, and the hand at the edge is thus a token of the presence of ‘the other woman.’ At this point, the identity of the narrator of the embedded novel has to match the woman in the photograph, and would also point to the one whose hand is visible. It is clear that the writer is playing this ‘now you see it, now you don’t’ game on purpose. In fact, the remaining (very rich) narrative is playfully providing ‘helpful’ information which alternatively points to Laura or Iris, and thus the text offers an abundance of anchors which do not materialize into complete spaces or unambiguous cross-input links until the end of the novel. In this context the ‘hand’ becomes another important anchor, and carries with it a rich array of meanings. At the most basic level, the hand in the photograph is simply an indication that there was a person sitting next to Alex – it is thus an anchor to the undisclosed identity of the other woman at the picnic. At the same time, it suggests that someone (the narrator of the embedded novel?) did not want the other sister in the picture – in the literal and figurative sense. The ‘hand’ is not only the hand of someone who was present in the represented space and time, but the hand that was left after someone tried to erase that person from the memory of the events. There are several levels of frame evocation here: ‘hand’ evokes the entire body and an entire person, the deliberate distortion of the original scene evokes someone’s intent, and the incomplete photograph is also an anchor to the missing part of the narrative, where the character forcefully removed from the story tells what had really happened. The double anchor of the photograph and the hand is a token of what the narrative needs in order to achieve coherence – the missing part of the story connected to the other woman. Later in the text we learn that Laura had access to the negative and produced two prints from it, each with one sister absent (save the ‘hand’); what is more, following a childhood habit of colouring black and white drawings, she stained each of the ‘hands’ a different color: Laura’s hand is yellow, Iris’s hand is blue. (8) But she’d cut herself out of it – only her hand remained. [yellow-Laura] ... “I have another one [photograph], for me.” “And I’m not in yours?” “No,” she said. “You’re not. None of you but your hand.” [blue-Iris] This was the closest she ever came, in my hearing, to a confession of love for Alex Thomas. (The Blind Assassin, p. 277)
This explains how there could be two photographs in the possession of two sisters – evoking two stories of love and loss related to Alex Thomas. However, both stories are in a sense written into both pictures, for even though the narrative might focus on one of the women, it will not be complete without the other.
2.3 Narrative spaces – an example
47
The hand/photograph anchor begins to interact with other anchors later in the text. In the embedded novel the photograph is kept in a gardening book, which also appears in the main narrative space (Iris’s autobiographical story), where she describes ways of occupying herself in the difficult period of her married life. “Perennials” is thus another anchor, identifying the narrator of the embedded novel as Iris: (9) She has a single photograph of him. She tucked it into a brown envelope . . . between the pages of Perennials for the Rock Garden, where no one else would ever look. (Prologue: Perennials for the Rock Garden) (The Blind Assassin, p. 7)
In the final chapters, Iris finally acquires her sister’s notebooks, mentioned in the very first paragraphs of the novel as a source of crucial clues to Laura’s suicide. (10) I riffed through the other notebooks. History was blank, except for the photograph Laura had glued into it – herself and Alex Thomas at the button factory picnic, both of them now colored light yellow, with my detached blue hand crawling towards them across the lawn. (The Blind Assassin, p. 627)
The hand’s suggested motion (crawl) evokes intended interference in the lovers’ encounter, as the body image of someone crawling evokes bent posture, such that the couple might not even notice anyone approaching; this in turn could be read as signaling evil intentions. The ‘crawling hand’ is narratively equal to a crawling person, someone who might then hurt the couple on the lawn. There are several important points to note here. First, this very low-level narrative choice has text-wide consequences, precisely because it appears in connection with the novel’s primary anchor. One of the points this analysis makes is that a narrative achieves its meaning through the compounded effect of meanings emerging at the lowest lexical and grammatical level. Second, it is important to note that the view of Iris as a destructive intruder which (10) suggests comes at the end of the novel, when Iris realizes that she unknowingly caused her sister to commit suicide. For Iris, the narrative voice of the story, to describe herself as my detached blue hand crawling towards them across the lawn, is to complete the untold story of the photograph. The representation blend has not only been fully integrated, through a match of the representation space and the represented space, but it achieves its narratively satisfying status through the interpretive viewpoint of one of the participants. The facts of the picnic as such would not tell the story that led to Laura’s death, but Iris’s realization of her own role in her sister’s life is the story being told throughout the book. In the end, ironically, the true story of the novel is told through the truncated photograph, not the photograph published in the paper, with all three of the main characters sitting together. In a sense, the way this anchor shapes the story in The Blind Assassin is by adding viewpoint and
48
Blending, narrative spaces, and the emergent story
causation to the events which otherwise would not have been framed to achieve any psychological effect. The role of the ‘hand’ anchor does not end there. While it structures much of what happens in the novel, it also participates in the construal of narratorship and narrative viewpoint. Narratorship here is not only the question of an epistemic viewpoint (who knew enough to tell the story in this way, a crucial consideration with respect to the embedded novel). Iris is ostensibly telling her own story, but it becomes clear as the text progresses that she is, perhaps primarily, telling the story of her sister’s life, as she finally understands her own role in it. In the last sections of the novel she talks about herself as the ‘hand’: Laura’s odd, extra hand, attached to no body – the hand that passed her on. The frame of the ‘hand’ now refers to a common understanding of the ‘hand’ as the body part capable of manipulating and doing things. The detached hand in the picture now stands for Iris’s decision to attribute the novel to Laura. The embedded novel tells the story of Iris’s secret relationship with Alex, but granting Laura the authorship has ironic consequences in view of the fact that the characters in the embedded novel are never referred to by proper names, only as ‘he’ and ‘she.’ By giving Laura the authorship, Iris’s ‘hand’ in a way also gives the participation in the story to her sister – something that could have been the case and which was clearly Laura’s deepest desire. In fact, people in Iris’s family understood the novel as Laura’s confession of a secret affair, so from their point of view, Laura was Alex’s lover. By referring to herself as Laura’s ‘extra hand,’ Iris removes herself from the story as its full-bodied participant – again, something the anchoring photograph supports. The anchoring image of the severed hand is reframed as the narrative voice. Iris explains this further: (11) I can’t say Laura didn’t write a word. Technically that’s accurate, but in another sense – what Laura would have called the spiritual sense – you could say she was my collaborator. The real author was neither one of us: a fist is more than the sum of its fingers. . . . Laura was my left hand, and I was hers. We wrote the book together. It’s a left-handed book. That’s why one of us is always out of sight, whichever way you look at it. (The Blind Assassin, p. 644)
These fragments provide crucial metanarrative comments. The tricky structure whereby a myriad of anchors keep on pointing to first one sister and then the other is now described in its full meaningfulness. Throughout the reading, one is tempted to keep on guessing which one of the sisters is the ‘she’ in the embedded novel, and the reader’s detective skills are rewarded when Iris admits that she wrote it and she was Alex’s lover. But the final comments quoted in (11) suggest that the shimmering narrative has a different purpose. The story is now seen as having two emergent structures, not just one. There is the narrative Iris constructs to set the record straight and explain her sister’s suicide, but there is also the narrative where she and her sister become one narrative agency
2.3 Narrative spaces – an example
49
and thus share in the experience of the story as well as the telling of it. The narrative thus reconstrues Laura’s life, by giving her a chance to experience what could not happen in the novel’s reality and to become an agent in the way the story is told. In the final sections the double photograph/hand anchor plays its last narrative trick. The embedded novel is spread chapter by chapter throughout the text, with other parts of Iris’s narrative taking much space in between. The Epilogue starts very similarly to how the embedded novel begins – see the fragments quoted in (5) and (9) for comparison: (12) The photo has been cut; a third of it has been cut off. In the lower left corner there’s a hand, scissored off at the wrist, resting on the grass. It’s the hand of the other one, the one who is always in the picture, whether seen or not. The hand that will set things down. (Epilogue: The other hand) (The Blind Assassin, pp. 649–650)
There is one striking difference, though. In the Prologue ([9]), the woman is clearly identified as Iris through the Perennials anchor (available later in the text), and the description of the photograph does not specify whether it is the left or the right hand that is seen next to the man. In the Epilogue, the hand is identified as the left hand, which clearly suggests that it belongs to Iris (when we imagine looking at the photo as described in [7], with Iris seen to the left of Alex). So the woman looking at the photo in the Prologue is Iris, but the woman looking at it in the Epilogue is Laura. This also explains why the ‘hand’ in (12) is the one that will ‘set things down,’ that is, Iris. What is more, the Epilogue also reconstructs the embedded novel to match the entire story of the photograph. The Prologue introduces Iris’s viewpoint, the text in the middle focuses on Alex and his lover, whoever she is, and the Epilogue adds Laura’s viewpoint. To make the analogy clearer, the women in the Prologue and the Epilogue are construed as ‘viewing’ their respective photographs, so that the concept of narrative viewpoint is clearly evoked. The anchor has now achieved its full potential and it explains what Iris means by saying that both sisters had their ‘hand’ in the embedded novel. The anchor described here requires a careful reader. Interestingly, critics (Roy 2006; Perrakis 2006; Wilson 2006) treat The Blind Assassin as just a postmodern exercise in instability of meaning. But the acknowledgement of the anchors, as I hope to have shown, does not trivialize the novel – on the contrary, it helps reveal its double bottom and its poignant view of what narrative fiction does. Just telling the story of what happened is not good enough, it seems to be saying, because the narrative may be the story that should have happened, but never did. The crucial role of the anchors discussed above depends largely on the nature of the concepts they metonymically evoke. The idea of a photograph is in itself a rich source of mental space topology. The other anchor, the hand, is also craftily
50
Blending, narrative spaces, and the emergent story
chosen. Being associated with the body but also with the human ability to manipulate, make new things, destroy things, et cetera, the ‘hand’ is easily seen as metonymic for a number of actions the person thus represented can undertake and perform. Moreover, all the actions associated with the hand also require a will and a subjectivity, thus in spite of not bearing any agency in itself, this body part can naturally stand for the subjectivity of a person. Atwood’s use of these anchors capitalizes on all of these aspects of the framing of photos and hands. The naturalness of the story construction processes triggered by the anchors supports Deacon’s (2006) discussion of how art works. He argues that in a broader context a single word such as hawk is much more than a symbol for an actual species of a bird, but rather a way to activate a network of concepts in the mind of the hearer/reader (bird of prey, powerful beak and claws, ruthless hunter, linguistic form, emotional response, et cetera). Furthermore, activating the image of the claws and predation may simultaneously evoke the concept of a victim, fear, and pain, or, for comparison, power, courage, attack, and pride. I’m reading Deacon’s claim as an acknowledgement of the aspects of the process of reading which rely not only on the constructing power of the words alone, but also on the richness of experiential frames already present as the background to reading. The anchors in Atwood’s novel seem to support this. While some are relatively simple (for example, Alex’s ready-mades), others interact with seemingly unrelated frames present in the reader’s mind. Also, all such frames evoke emotional responses needed in processing texts. Relying further on Deacon’s work, one can claim that anchors are an ultimate way of building on what he refers to as our ‘representational stance’ – the natural assumption of human cognition that something stands for something else. It is a more straightforward (though not simple) process in our use of language as a system of symbolic representations, but it is more complex in the narrative. Naturally, the narrative expressions are symbolic in the ordinary sense of the word, but they may be deployed in the text in a way which poses a challenge to our representational stance. We do know what a photograph is and we expect it to represent a recognizable reality. If that reality is not provided, it can still be constructed. We can thus say that narrative anchors are narratively salient expressions which rely metonymically on frames and exercise our representational abilities, but create suspense by providing necessary information only bit by bit. To sum up, the cognitive impact of anchors on the reading of narratives is multidimensional. At the simplest levels, anchors are frame metonymies (photograph for the representation blend it constructs, fire for pain, destruction, purification, light, et cetera). Moreover, the textual situation of anchors guarantees the emotional response Deacon sees as indispensable in the appreciation of literature. Perhaps more important, the patterns according to which anchors are reactivated and gradually enriched with new framing are also the patterns of
2.3 Narrative spaces – an example
51
narrative construction at every level. They participate in establishing temporal sequences of events, in revealing identities of characters, and in supplying missing elements of the plot or prompting their reconstruction. Overall, they provide the skeletal system of interacting frames which, as they are enhanced with new topology, point to crucial narrative spaces and connections between them, and thus lead to the emergent story. Narrative anchors capture more than the progressive accumulation of content and context, as described, for example, in Werth’s study of text worlds (1999) or Emmott’s discussion of ‘contextual frames’ (1997). They are also a more specific mechanism than what literary critics refer to as ‘motifs’, especially in the mental-space setting function. They exploit the mechanisms of frame metonymy, frame evocation, mental space set-up or evocation, and crossspace projection. At the same time, they capture the complex interaction between specific expressions used, emergent connections across various parts of the text, and the reader’s processing of it. 2.3.3
Reference and story construction
The discussion above shows how narrative anchors of various types play a role in prompting cross-input projections and the emerging cohesion of the story. Here, I will consider the ways in which referential expressions contribute to the emergence of the story. The central piece in the narrative-space structure of The Blind Assassin is the embedded novel, featuring only two characters, referred to through pronouns only. The pronouns he and she let the reader assume that these forms refer to subjectivities identified elsewhere in the novel, and thus a large part of the reading process consists in attempts to establish cross-space identity links which would explain who the characters are. He is soon identified as Alex Thomas, while the decision about the identity of the she oscillates between two sisters, Iris and Laura Chase. However, another narrative space of the novel, a science fiction pulp story Alex is composing, also features no names, using only descriptions of roles instead. The roles are specific to the text and not very inventive (‘blind assassin,’ ‘mute girl to be sacrificed,’ ‘king of the underworld’), but their presence in the overall novel and the metaphorical understanding of their characteristics (blind, mute, hidden) provide important clues to the understanding of the story (cf. Schnepf 2006 for more discussion). The story these characters participate in explains the facts from their lives which are not represented anywhere in the text, but the role of these facts is appreciated only through the matching of the roles with characters in other narrative spaces. Crucially, the entire construal of the sci-fi story represents the viewpoint of the in-novel writer, Alex, who is thus revealing the nature of his participation in the story. He was not given his own
52
Blending, narrative spaces, and the emergent story
voice anywhere else in the narrative, and thus the pulp sci-fi space is where his viewpoint is construed. Without the reliance on the role of referring expressions The Blind Assassin could not be imagined in its actual form. It forces the reader to construct dynamic frames to represent the identities signaled by proper names – otherwise, the entire reasoning of the choice between Iris and Laura as the ‘she’ of the embedded novel could not take place. It builds on the assumption that personal pronouns do have fully framed counterparts within the spaces of the text. It constructs roles which allow the reader to place characters in the appropriate frames. And it also relies heavily on the construal of the SV-space as profiling the narrator matched with a character (Iris), so that the spaces construed from another viewpoint are constantly matched with the SV-narrative. Crucially to the point of this project, the establishment of cross-input projections based on the links between names, role-descriptors, and pronouns is an exploitation of low-level linguistic mechanisms on the level of the extended discourse, such as a narrative. Knowing that a pronoun requires an antecedent in some mental space or other is what prompts readers into the attempts to seek referents of the expressions in the embedded novel in other parts of the narrative. Similarly, the frames set up in the sci-fi text, with their attendant roles, create the need for further referential links. All these referential connections are not simply available in the text of the novel: they need to be constructed (Figure 2.3 shows the final resulting set of referential links). In a sense,
SV-space
Narrator: Iris present
MN-space
Narrator: Iris Author: Iris past tense Memoir Narrator: Iris past tense Laura Richard Father Alex
Clippings Author: journalist Mrs. Richard E. Griffen Embedded novel Author: Iris Narrator: 0 present tense he she
Sci-fi story Author: Alex Narrator: 0 present tense Lord of the Underworld King blind assassin (he) mute girl (she)
Figure 2.3 Referential connections across spaces in The Blind Assassin
2.4 Emergent story
53
referential frames (profiling slots for a name, a pronoun, and a role) serve as prompts for the filling in of all the slots, even though there are no clear anaphoric or cataphoric patterns to rely on. Such a text-wide deployment of reference seems to be a characteristic of narrative discourse. This form of reliance on referential expressions is used in cross-input connections and the resulting enrichment of input narrative spaces with content. But we should also note that in the process of reading, the distance between the referential slots in individual spaces needs to be compressed, so that the final complete identity of characters emerges as a result. Blending and compression are thus participating in story construction, even at the level as basic as identity of characters and reference. Referential expressions in the narrative pose a complex and multidimensional problem. Starting with the distinction between first- and third-person narration, continuing to the less usual forms of second-person narratives and an occasional use of we, the range of usage of pronouns in narratives poses complex questions about the internal coherence within a text. In Chapter 3, I will consider pronouns used as text-wide narrative choices; Chapter 5 will look further at the question of proper names and roles, while Chapter 7 will consider the use of pronouns in Speech and Thought Representation (STR).
2.4
Emergent story
In much of the narratological literature the concept of the story is defined in relation to text.5 The text is the form (some also mention discourse as an additional element, cf. Toolan 2001) while the story is what the text represents. While the text may be fragmented, incoherent, temporally disorganized, the story is a temporal sequence of causally linked events leading to a resolution of some conflict or problem. The contrasting concepts seem very much like an elaboration of the form/meaning pair, but there is an additional claim, made with special clarity by Chatman (1978, 1990), arguing that the story as the abstracted chain of events remains the same when it is told through a different text or a different medium, such as a film. This may be related to the linguistic tradition of maintaining that propositional meaning remains constant under various formal changes, so that it is natural to see the ‘story’ as the meaning of the ‘text.’ Regardless of the source of the distinction, the ‘story’ is one of the most salient concepts in narratology, but it is typically not discussed in terms of its cognitive status emerging in reading. While Herman (2003a) talks about cognitive aspects of story construction such as chunking of experience or attributing causation, the concepts he evokes refer to the general concept of a story as a way to structure experience. The ‘story’ under discussion here, on the other hand, is the semantic correlate of a specific text.
54
Blending, narrative spaces, and the emergent story
The text, however, should not be seen as simply the words and syntactic structures, as it also relies on cultural availability of frames structuring the events described via the choice of words. The text is not just a collection of linguistic forms, it is also a rich resource of framing. To explain the emergence of the construct called a ‘story,’ I will treat textual choices as prompts guiding layers of blends leading to the sequential, causally structured configuration. Finally, the story is the final result of several modes of interaction with the text – reading the words, activating the frames, searching for correlates in one’s experience, making cross-space connections, blending narrative spaces, establishing identities, constructing tentative scenarios, storing them in memory, revising them as new events are narrated, responding emotionally, et cetera, et cetera, et cetera. The richness of the story as a construct can best be explained if we attempt to discover the processes that lead to the final result. Much of this work is already being done in various contexts – consider Gerrig’s work in psychology (1993), Emmott’s studies combining cognitive psychology and discourse analysis (1997, 2002, 2003; Emmott et al. 2007), or Herman’s renewed vision of narratology (2002, 2007). The goal of this study is to give the view afforded by cognitive tools used in the analyses of linguistic meaning. 2.4.1
Sequence of events and the story
One of the most contentious aspects of ‘storiness’ is its idealized status which, as narratologists argue, remains the same in different tellings. The claim calls for an explanation of where such a construct resides. It cannot be an overall interpretation arising in a single reader’s mind, since such a construct would be too specific. If, on the other hand, the concept is expected to remain constant across choices of a medium, across readers, or even across readings (as readers may change their perception in time), then it has to be ‘boiled down’ to the sequence of events and some conflict to be resolved. It would then be a very minimalist view of the story. The sequentiality of the story raises questions about the temporal discontinuities in texts, which are a norm rather than an exception. In fact, the need to disrupt temporal sequence appeared important in the most celebrated of epics – the Odyssey. It starts in the middle (with the episodes when Telemachus attempts to take control of his father’s household, while Odysseus is held captive on Calypso’s island), changes narrators, and does all the ‘modern’ narrating besides. But knowing why Odysseus is missed on Ithaca puts his entire journey in a different light and leads into the conclusion where the point is not that the journey ends, but that balance returns at last. Also, temporal infractions serve specific narrative purposes. Even in classical texts, not as experimental as contemporary ones, flashbacks are used for specific reasons. The story in Pride and Prejudice starts just before the ‘eligible
2.4 Emergent story
55
bachelors’ appear on the scene and proceeds, through various complications, until the two couples are happily married (it is unfair to Austen’s talent to summarize her masterpiece in this way, but time is all that is being considered here). In fact, there are important flashbacks in the novel: one when Wickham tells a false story of his youth which puts Darcy in a bad light, and the other when Darcy tells the true story of what happened between them. So it is by far not enough to describe these stories as flashbacks, since they present two opposite images of the main character, and entirely change the course of the story – the first makes marriage with Darcy impossible for a person with Elizabeth’s sensibilities, the other makes it very desirable. They do so not because of what ‘really’ happened in the story’s past, but because the versions of the past are crucial to characterization. The little deviation from sequentiality is thus not just acceptable, but central to the construction of characters’ dispositions. Broadly speaking, the list of temporal infractions in narratives is so long and varied that insisting on the centrality of this dimension of cognitive organization seems questionable. What is more, it is also questionable on experiential grounds. First of all, sequential interpretations of events profiled in linguistic constructions often emerge from construal, rather than from straightforward meaning of linguistic forms. They had a baby after they got married or They first got married and then had a baby clearly indicate the sequence of events with after, first, and then, but They got married and had a baby or They had a baby and got married only allow one to construe the sequence one way or the other based on assumptions of the clause order reflecting the order of events. But in real situations we only follow the sequence to the degree to which we experience the events ourselves or are exposed to their results. Temporal sequence is rarely relevant to our understanding of events, but knowing their consequences is crucial. The sequence is a questionable criterion even in the most sequential of stories, but epistemic stance and understanding of causation seem to matter much more (see Dancygier in press b for more discussion). The popularity of the ‘sequential’ core of narratives can be explained if we consider the cognitive construction of time. Stories develop through time and our conceptualization of time assumes something like a linear path progressing from prior events to the events that follow. The discussion of metaphors of time and their instantiations (Lakoff and Johnson 1999, Evans 2004, Nuñez and Sweetser 2006) postulates the emergence of our temporal concepts from our experience of motion through space (which not only changes our position but also takes time). Motion through time also assumes a specific orientation, such that the past is ‘behind’ us and the future ‘ahead’ of us. Stories also assume a direction, following a path from the story’s past to its future and its conclusion, among other reasons because the characters’ experience of the story’s events is modeled after our experience of time, but primarily because a sequential
56
Blending, narrative spaces, and the emergent story
organization is best suited to the representation of the causal organization, which seems more central to narrativity. In other words, even if the events do not appear in a sequence, the causal chains suggest such a sequence, and, as a result, the entirety of the story is best viewed as sequential, even though its telling may not follow the temporal chain. Also, even the spontaneous oral stories, such as the ones analyzed by Labov (1972) and Labov and Waletzky (1967), include components such as evaluation, where events from the story’s past or even future can be evoked to justify the telling of the main line of ‘complicating action.’ To conclude, sequentiality is useful as the final construal, and as such it is also the most salient one.
2.4.2
Story construction, vital relations, and optimality constraints
A ‘story’ can thus be discussed as a cognitive construct, a blend, emerging through the process of meaning construction triggered by reading. The role of the text is crucial in providing such prompts, but the emergence of the story relies to a comparable degree on the frames evoked in the reader’s mind, and on the construction of double-scope blends, integrated into the mega-blend. The story is the mega-blend arrived at in the interaction with the text. The processes of integration, as described above, rely on all the primary aspects of blending. Narrative spaces, such as the ones in The Blind Assassin, are set up as independent inputs, but then selective projection utilizes the elements needed in the blend. For example, the story of The Blind Assassin does retain all the hesitation about the authorship of the embedded novel, and retains Iris as the actual author (Figure 2.3). Also, subsequent blends throughout the process affect the understanding of the inputs, via backward projection. Finally, the story-blend itself undergoes the required integration and elaboration (so that, for example, Alex and he are fused into one participant, but the understanding of his identity relies on setting up a coherent view of his persona). Vital relations play a role in the construction of the story as well. Because of the two inputs, Alex and he are not really connected at all at first, then become connected by the relation of Analogy, until they fuse into the Identity relation in the emergent story, and in the final stage he is also linked to the ‘blind assassin’ via a Role–Value link. Similarly, the relation of the role of the author of the embedded novel is not clearly attributed to any of the Chase sisters, but in the final blend Iris is linked to that role, as well as the role of the lover in the embedded space and the role of the mute girl in the sci-fi story (she cannot tell anyone either about her relationship with Alex, or about her sister). The vital relation of Time also affects the blend, as temporally unconnected spaces (the memoir and the embedded novel) are compressed into the sequential organization of the story; the relation of Change becomes the relation of Causation as we
2.4 Emergent story
57
discover why Laura committed suicide, et cetera. Enumerating all the vital relation changes would take far too much space, but these examples support the claim that story construction progresses along the lines proposed in the blending framework. Construction processes described above also satisfy all the optimality constraints, as described by Fauconnnier and Turner (2002). The integration principle allows representations in the blend to be manipulated as a single unit, and that is what appears necessary in the story of The Blind Assassin, after all the identity relations are established. The blend maintains the inner structure of the inputs (topology and web principles), while tightening metonymies (as in the case of the hand and the photograph discussed above). Finally, the understanding of the entire story can be unpacked into an explanation of how the blend integrates the inputs, and the blend is necessary (good reason principle) to explain the events and the role of characters. Although The Blind Assassin seems to be an unusual example of the difficulties of story construction, it is not more than a useful one. One of the reasons to argue that the story is an emergent structure is to consider differences among readers. The emergent construct is not the same for all readers as different aspects of integration can come into focus on different occasions – even the critics’ responses to novels show that foregrounding certain aspects of the story yields a different blend in the reader’s mind. Thus the stability of the story in various media (such as film versus text) is an idea requiring further investigation. While it may not matter much if some bits of plot are omitted from the movie or if secondary characters are fused, the process of arriving at the story through visual means may highlight or downplay frames which play different roles in the reading. There is enough similarity in most cases, but because the process of arriving at the final blend is different, the results may also not be fully compatible.
3
Stories and their tellers
I am the man who comes and goes between the bar and the telephone booth. Or, rather: that man is called “I” and you know nothing else about him . . . (If on a Winter’s Night a Traveller, Italo Calvino)
The preceding chapter has outlined the concept of a narrative space as a textually prompted construct used in story construction. Narrative spaces can thus be distinguished through their temporal dimension, a consistent subplot, or the construction of a specific epistemic point of view. I have also explained how narratives achieve coherence through levels of blending leading to an emergent story. However, reading a work of fiction invariably assumes a text-mediated contact with the constructed fictional subjectivity often referred to as ‘the teller’ or ‘the narrator.’1 The concept is textually constructed, which has been stressed repeatedly, primarily in order to avoid simplifications whereby the narrator is identified with the author. At the same time, the illusion is compelling in many narratives, and the reason why the author is so naturally understood to be responsible for the way the story is told has not been answered in sufficient depth. While the ‘death of the author’ (as a viable narrative concept) has been widely publicized, the news may be exaggerated, as some flesh-and-blood authors begin to claim their right to be considered legitimate participants in the narrative exchange.2 Moreover, the ‘demotion’ or backgrounding of the author also avoids any interest in the question that is just as important as the question of the reader response and narrative transport, namely, the ability of a writer to construct a textually available subjectivity. What one might call ‘authorial transport’ is a cognitively unique ability whereby a writer can construct a mind other than his or her own and use it as a medium in making a story accessible to the reader. While the interest in the theory of mind has highlighted the psychological aspects of reader response (Iser 1974; Zunshine 2006), these approaches do not cover the mechanisms of story creation. My tentative suggestion is that creative storytelling requires the construction of very elaborate blends in which actual experience is embodied in a different context, after a process of selection and compression. Selection is necessary in order for the most salient aspects of 58
Stories and their tellers
59
experience (whether physical, historical, or psychological, first-hand or observed) to be used as the conceptual skeleton of the emerging blend. The cognitive ‘padding’ which adds specific character traits, circumstances, events, et cetera is the matter of the emerging narrative spaces. However, the crucial mechanism is that of compression (Fauconnier and Turner 2002; Turner 2006a; Dancygier 2004b). I argued in Chapter 1 that compression is the mechanism whereby specific stories of the oral tradition acquire their generic power to shape behaviors and maintain distributed cognitive networks. The crucial effect emerging out of the compression is also an enhancement of the central themes and increased salience of the intended ‘narrative uptake.’ In the written fictional narratives these tasks are equally important, but are also subject to the author’s choices. Naturally, any narrative selects the facts to be narrated and leaves much of the assumed story untold, but the kinds of selection would depend on the authorial choice of themes. The central compression mechanism, however, is the one whereby the entire story is packaged into a text where only some of the story’s events are told, and the specific compression choices lead to narratives of different kinds. At the very basic level, a fictional story affords a compressed view of the events, of the kind which is typically not available in actual experience. Only a novel gives one access to a picture as complete as is required in a form which is manageable, because it is compressed, and providing the required multitude of viewpoints. The kinds of compressions will affect all aspects of the narrative – time, viewpoint, characterization – but without the compressions the text could not achieve its salience. The processes of selection and compression play an important role in narrative form. In the actual narrative organization they are often subordinated to the choice of some subjectivity as the source of the narration, whichever term is chosen to describe it. While it is clear that such a subjectivity is a narrative construct, the reader has the impression that there is an intentionality and epistemic status behind the discourse, which is typically talked about as a narrator (Gibbs 1994, 1999). The intentionality is crucial in that the very act of storytelling assumes the intention to use the story in its proposed form to communicate some content, even if various forms of narrative experimentation disrupt the impression of consistency and purpose. The epistemic status is equally important as a constitutive feature of ‘narratorship,’ since, regardless of the scope of knowledge the narrator displays (ranging from omniscient, constituting a narrator who knows everything including characters’ thoughts, to unreliable, with a narrator who has a very limited view of the facts), a narrator gives the reader access to crucial narrative facts. In this chapter, I will focus on the aspects of ‘narratorship’ that involve specific configurations of narrative spaces, and also consider the macro-level narrative viewpoint in its relation to the textual profiling of storytellers. As
60
Stories and their tellers
Chapter 2 made clear, narrative spaces as cognitive constructs depend on various kinds of choices, and may not always take the form of narratives, not even skeletal ones. Here, however, I will focus on how various conceptualizations of the story’s teller(s) affect the narrative as a whole. 3.1
Narrators, narrative spaces, and viewpoint
While the focus of this book is not primarily aligned with the interests in narratology, nevertheless the distinctions of types of narration introduced by Genette (1980) and later developed in numerous studies (Rimmon-Kenan 1983; Bal 1985; Simpson 1993; Herman 2002) should briefly be mentioned here. The two questions which seem to be most pertinent to both the original distinctions proposed by Genette and their later expansions seem to seek a positioning of the narrating fictional subjectivity within the story or outside, and aligning it with that of a character or not. While these are indeed crucial distinctions, they fit the framework proposed here only to some degree. First of all, the dual contrast between ‘inside’ or ‘outside’ of the story cannot naturally be maintained in a number of narratives where there are in fact different storyworlds set up, possibly narrated through different types of narrator. Furthermore, as in the case of the ‘Clippings’ space in The Blind Assassin discussed in Chapter 2, there may be no single ‘narrator’ present in any form, while the coherence of the narrative space is achieved through an inexplicit but shared epistemic and moral viewpoint. While it is true that even the most fragmented narrative builds a conceptual structure which I have termed ‘emergent story,’ the boundedness of this construct is difficult to establish in a way which would clearly determine the position of the narrating voice. It seems intuitively convincing to claim that if a narrator is not at the same time a character, then he/she/it must have some knowledgeable position in a different domain, but it is interesting to ask also what that domain is and how the narrator gets access to the ‘story’ if he/she/it is positioned outside of it. I will thus attempt to talk about the ‘type of narration’ or ‘narrator’ as a feature of the narrative space configuration in any given case, and show how the spaces are linked via blending and viewpoint compression. As regards the cases where a narrating voice is aligned with that of a character within the bounds of the storyworld, the status of that voice also depends on the role of the specific character in the text. If the character/narrator tells his or her own story, in the first person, it is still typically the case that he/she relies on facts reported by others as well, and the balance between ‘narrating-the-self’ or ‘narrating-the-story’ is different in different texts. For example, the narrator/ character in Eggers’s autobiographical novel is almost solipsistic in his focus on his own consciousness (though even he relies on others in some fragments of the story), while the narrator/character of Chatwin’s Utz is in fact telling the story of another character, very much as an omniscient narrator would. While
3.1 Narrators, narrative spaces, and viewpoint
61
narratologists have made all the required distinctions, I will assume here that they can be elaborated and clarified in a broader context. The categories to be discussed below emerge out of the understanding of the concept of narrative space, as outlined above. Primarily, I will rely on the concept of narrative space as a source of narrative structure (among others, the features of the so-called storyworld), but, more important, as a source of viewpoint configurations. I will understand narrative space viewpoint as a particular use of the structure of the space. While the narrative may profile a number of characters, and temporal and spatial structures, and involve knowledge of certain facts, et cetera, the specific viewpoint, whether local or global, is a matter of selecting an aspect of narrative structure within the space as foregrounded or specifically profiled. In other words, viewpoint does not necessarily rely on a viewing subject, so that it is not necessarily ‘someone’s’ point of view, but it relies crucially on a selection of an aspect of space topology which provides a filter through which the events are narrated. Viewpoint may thus be spatial or temporal in the basic sense of aligning the narration with a given spatial or temporal dimension of the narrative space. But it can also be epistemic, when access to certain facts, but not others, is assumed (which includes any range of narrating voices, from omniscience to unreliable narrator), or ideological, when the very selection of facts presents the story in a specific light.3 In this approach, trying to come up with a classification of viewpoint types will not lead to a better understanding of the concept, though it may add to an appreciation of its complexity. In fact, narrative spaces, even more than mental spaces, can, and indeed have to, profile a number of possible viewpoint options in order to construct a topology suitable for narrative processing. While the ‘narrating fictional subjectivity’ may represent aspects of viewpoint which remain constant during most of the narrative, there are also local shifts of viewpoint which may affect expressions as small as individual sentences or phrases. As I will try to show in this chapter and the next one, viewpoint is best talked about as a configuration of viewpoint aspects, ranging across narrative spaces involved, and not as a uniform feature of a narrative fragment or the whole text. Even in a basic act of spoken communication one constantly operates in a network of viewpoint dimensions – the speaker’s and the hearer’s (with their epistemic and emotional stances), the time of conversation, the time of events talked about, the spatial locations experienced and mentioned, epistemic or other viewpoints available in the conversational record, in the broad context, or in the artifacts available, et cetera. Choosing to foreground one of these aspects in a construction does not make other viewpoint angles disappear, and shifts are common. If a speaker says in the course of a gossiplike exchange I’m not saying I expect he will propose, but he might, she is manipulating a number of viewpoints in just one sentence, even if we disregard the time and place of the conversation for the sake of clarity. There is her epistemic viewpoint
62
Stories and their tellers
(I expect), the hearer’s viewpoint (you are saying he will propose), the speaker’s rejection of attributed viewpoint (I’m not saying it), the man’s emotional viewpoint (he might be in love enough to propose), the distanced viewpoint marked by the past tense might (instead of may), etc. There is no reason to assume that narratives are any less complex in this respect. However, narratives can be talked about in terms of levels of viewpoint structure, which I will refer to as ‘macro-level’ and ‘micro-level.’ First, there seems to be a crucial difference between the consistent, narrative-wide choice of the way in which the narrating subjectivity is profiled, and the sentence-level shifts and manipulations of viewpoint. While the former emerges out of specific kinds of narrative space configurations, the latter elaborate the topology of individual spaces to give depth to the story. Second, macro-level phenomena reflect the text’s assumed epistemic viewpoint and contribute directly to the construction of the emergent story, while micro-level forms have more to do with the construction of specific events and conceptualizations. The two levels, however, are not independent, as the aggregate effects of micro-level shifts affect the macro level all the time, especially with respect to the epistemic viewpoint of a narrative space. Let me illustrate the interaction between levels with an example, before moving on to the general discussion of macro options (in this chapter) and micro phenomena (in Chapter 4). Example (1) comes from the autobiographical novel by Dave Eggers A Heartbreaking Work of Staggering Genius, where he tells the story of the years when he was bringing up his young brother, after their parents had died. The text follows the inner thoughts of the narrator/main character throughout. In the text preceding (1) he talks about a night spent with a woman, while Toph, the brother, slept at a friend’s house. (1) I wanted to be home in case he came back early . . . made it in time . . . The house was empty, and I dove into bed, fell back asleep, and when he came back home his brother was there, of course had been there the whole time, of course had never left. (AHWOSG, p. 112)
The first sentences, as the rest of the novel, refer to Dave (the sole narrator of the entire text) as I, while he refers to Toph. But when reference to Dave shifts from I to his brother, both levels of viewpoint are affected. At the micro level, the viewpoint no longer focuses on Dave’s intentions (make Toph believe that he spent the night at home), and shifts to the perception attributed to Toph (that indeed Dave never left the house). But it clearly involves the macro level as well, since the expression his brother assumes Toph’s view of Dave’s family role (available throughout the text), and Dave’s view of himself as ‘Toph’s/his’ brother (if the viewpoint totally shifted to Toph, my, not his, would be the required pronoun). The fragment thus manipulates viewpoint to let the reader appreciate both Dave’s intention to mislead, and the fact that Toph was convinced by the
3.1 Narrators, narrative spaces, and viewpoint
63
situation Dave had staged, both of which are relevant to the macro-level narrative, even though the effect is achieved through a minute shift at the micro level. Discussing viewpoint as the specific presentation of the content of a given narrative space makes it possible to acknowledge the reasons why a text may profile several spaces. In the case of The Blind Assassin, macro-level viewpoint varies significantly from space to space, while in other narratives the choice of macro viewpoint may be more consistent. Even though in all cases the narrative space is structured in terms of time, space, and the epistemic or emotional disposition of the teller, et cetera, macro viewpoint foregrounds some of the aspects of space topology and backgrounds others. For example, the consistent first-person narration of Eggers’s novel allows him to focus on the main character’s emotional states and obsessions, rather than on the consistent presentation of events; thus the actual story, even though basically sequential, has to be pieced together in the reader’s mind. In other texts the narration of events may take priority, but then a specific temporal focus will be chosen while other pieces of the timeline will be rendered through flashbacks or omitted altogether. In what follows, I will argue that the macro-level viewpoint is determined by various configurations of three types of spaces. First, there is story-viewpoint (SV) space, the space where the ‘narrator,’ or some features of narratorship, are located. The overarching narrative structure (the primary plot) can be referred to as main narrative (MN) space, which may profile its own narrator(s), but may rely entirely on the SV-space structure. Finally, a character in the MN-space can be selected as the narrator, and we can then talk about Ego-viewpoint, such that the character’s knowledge or intentionality provides the primary focus to the narrative mode. I will consider various examples of these categories below. I will also rely on two additional concepts introduced in cognitive grammar (Langacker 1990a, 1991) to clarify the kinds of possible correlations between construals prompted and the linguistic forms prompting them. While the verb may require a construal including an agent, a selected grammatical form may not profile the agent specifically. In an active voice construction, such as Tom broke the window, both the agent (Tom) and the patient (window) are on-stage participants, explicitly profiled by the form of the sentence; in the passive form The window was broken, however, the agent is off-stage, but nevertheless remains a part of the construal, since the act of breaking requires someone doing it. Various constructions can manipulate the ways in which construals are prompted by expressions, and narrative constructions can be claimed to do the same, with different elements of the space topology on-stage or off-stage for different lengths of time. I will describe some examples of various narrative constructions of tellers in terms of their relationship to the MN-space of the story being constructed. My goal will not be to cover exhaustively all available types, but rather to outline ways in which such constructions can be talked about. I will also attempt to rely on the linguistic correlates of these narratorial constructions, primarily tense and
64
Stories and their tellers
person, and refer to certain general ways in which these categories influence the narrative. For example, the use of past tense as the primary form of narration creates some distance between the narrator and the story, thus often opening more room for various degrees of omniscience. The use of present tense, for comparison, suggests a narration following the events as they develop and more immediate access to the minutiae of the story, but leaves less room for reflection and commentary. Furthermore, the choice of first person reinforces the effect of immediacy and direct access, while the choice of third person is often correlated with some distance and various degrees of ‘omniscience.’ These general observations will be further substantiated through the examples discussed below. In what follows, I will look at some of the common narratological constructions rendered by these choices and discuss examples which relate these concepts to configurations of narrative spaces. 3.2
Types of teller and epistemic viewpoint
The so-called omniscient narrator is a common choice of narrative form – from Henry Fielding’s Tom Jones to most Victorian novels, and all the way to contemporary literature. In many earlier texts, as in Tom Jones, the narrator is also ‘intrusive’ (addresses the reader directly), but as the novel form develops this form is used less and less. ‘Omniscience,’ however, remains a common feature of narrators. Considering that the concept in fact applies to both Henry Fielding’s and Virginia Woolf’s work, it does not seem to be a category salient enough to distinguish narrative techniques with any reliability. What it does refer to, though, seems to be the narrative space set-up whereby the content of all the narrative spaces of the text is accessible from an independent space in which all other spaces are embedded. This ‘outer’ narrative space, which I introduced above as ‘story-viewpoint space’ (SV-space), usually has little or no topology of its own in terms of time, setting, events, or subjectivities, but it houses a vantage point with the rest of the narrative in its scope. Because of the SV-space, the story is presented as if there were a deictic ground for the overall structure. I argue that the deictic elements such as time, speaker/hearer, or location are profiled even if the narrator is not. The vantage point is cognitively available in all cases, but the degree to which such a vantage point is associated with a narrating subject or not accounts for the various choices made.4 3.2.1
On-stage narrators
For example, an intrusive narrator, as in Tom Jones, speaks as follows: (2) Reader, I think proper, before we proceed any farther together, to acquaint thee that I intend to digress, through this whole history, as often as I see occasion, of which I am myself a better judge than any pitiful critic whatever. (Tom Jones, p. 17)
3.2 Types of teller and epistemic viewpoint
SV-space
65
Narrator: ‘I’ Reader: ‘you’
Time: present
Character: he/she
MN-space
Tense: past
Character: he/she
Character: he/she
NS1
NS2
NS3
NS4
Metanarrative space
Figure 3.1 Intrusive narrator (example 2)
Here is an on-stage (fully profiled) narrator who addresses the reader directly and comments on expected reactions to the narrative (see Figure 3.1). The SVspace thus has a certain topology – the narrator is on-stage, the time is later than story-time. All characters are referred to in the third person. One of the results of this set-up is that there is clear deictic distance between the SV-space and the MN-space, including each of the narrative spaces (NS) profiled. While perhaps artificial, this narrative set-up constructs its SV-space as close as possible to the default discourse space, with the subjectivities of the speaker and the hearer, and as such is also the closest to what might be considered the most salient narrative set-up. In fact, the narrator’s mention of critics also sets up a metanarrative space, where the entire novel and its discourse are available for independent evaluation. This kind of set-up is also interesting with respect to its epistemic viewpoint. The source of all the information the story provides is localized in one space. There is no need for the story’s narrative spaces to be evaluated from the epistemic perspective or matched with appropriate subjectivities – all is in the hands and mind of the narrator. This narrative form is not often used now, although it is worth noting that it was also used by Bulgakov (to be discussed below), with a highly humorous effect.
66
Stories and their tellers
The narrator may further attune his/her presence to the needs of the narrative construction. In the first pages of Henry James’s novel The Portrait of a Lady the narrative starts with an “I” narrator, commenting on what he intends to render to the readers. The “I” voice introduces the scene through a fine and extended description of the sunlit lawn where tea is served, but then the narrator’s presence gradually blends away as the characters begin to speak and take their roles in the story. The narrator’s voice in the first paragraphs is only used to give the setting and the atmosphere of the first scene, but the narrator’s presence is thus defined for the duration of the novel. Though this narrative choice has little influence over the text as a whole, it is an interesting confirmation of the need for the SV-space as a separate cognitive construct available.
3.2.2
Off-stage narrators
It is interesting to compare these examples with the more contemporary form of omniscience, as in Anne Tyler’s novel Ladder of Years. The main character, Delia, comes to a job interview, and is waiting in the office: (3) She crossed to the desk, which was bare except for a telephone and a typewriter. She lifted a corner of the typewriter’s grey rubber hood. Manual; not even electric. (She had worried she would find a computer.) She gave a small, testing spin to the swivel chair behind it. Good afternoon, she would say. I’m here to ask if . . . No, not ask. Ask was too tentative. (Ladder of Years, p. 94)
This is a more common type of omniscient narrator, who knows Delia’s spatial and temporal location and follows her actions (she crossed to the desk, lifted a corner of the hood, et cetera). More important, the narrator knows what Delia registers consciously in the environment (bare desk, manual typewriter), and what her thoughts are. The narrator is thus off-stage (does not speak as an independent subjectivity), but all that is said about Delia is relayed through the SV-space, which is located temporally later with respect to the main narrative space. The only glimpse of a narrator’s subjectivity is the comment in parentheses: She had worried she would find a computer. It is less important that the sentence flashes back to Delia’s thoughts before her arrival, but it is crucial that it is an explanation of the reasons why Delia checked the kind of typing equipment she would be using if she were to get the job, and why the phrase not even electric is to be interpreted as an expression of relief rather than dismay. The almost invisible narrator is thus present here and guides the reader in interpreting Delia’s words and actions. In other words, while off-stage, the narrator’s subjectivity is still present as a source of coherence and organization of the story.
3.2 Types of teller and epistemic viewpoint
67
The next example comes from Virginia Woolf’s modernist novel Mrs. Dalloway: (4) But – but – why did she suddenly feel, for no reason that she could discover, desperately unhappy? . . . no, it was not Sally Seton saying that Richard would never be in the Cabinet because he had a second-class brain (it came back to her); no, she did not mind that; nor was it to do with Elizabeth either and Doris Kilman. (Mrs. Dalloway, pp. 120–121)
Here is an ostensibly similar third-person, past-tense narration (which Humphrey [1954] describes as ‘indirect internal monologue’), but the narrator does not say much about the character’s actions and focuses on the flow of thought in the character’s mind, with just a minimal level of commentary – it came back to her. Throughout the novel, narration of events is minimal, and transitions between episodes are primarily transitions from one consciousness to another: (5) “The time, Septimus,” Rezia repeated. “What is the time?” He was talking, he was starting, this man must notice him. He was looking at them. “I will tell you the time,” said Septimus, very slowly, very drowsily, smiling mysteriously at the dead man in the grey suit. As he sat smiling, the quarter struck – the quarter to twelve. And that is being young, Peter Walsh thought as he passed them. (Mrs. Dalloway, p. 70)
At first, the reader is given access to Septimus’s visions, occasionally interrupted by scraps of conversation between Septimus and his wife. But in the last sentence, the viewpoint abruptly shifts to Peter Walsh, and it remains in his thoughts for some paragraphs. These are but a few examples of the way in which a narrative of this type includes much more than the specific viewpoint of a specific character at any given point in the story. There are many Ego-viewpoints represented (Clarissa, Rezia, Septimus, Lady Bruton, Miss Kilman, et cetera) through reports of the characters’ acts and thoughts, but the narrating voice is minimal and the distance between the SV-space and the MN-space is also much reduced. The set-up with Ego-viewpoints is represented in Figure 3.2. While the case of example (3) is different in that one Ego has been selected (Delia), and (4) and (5) represent a multiplicity of parallel Ego-viewpoints, the general narrative space structure is similar. Also, the dotted lines which link characters to the zero narrator represent the fact that the viewpoint of a character becomes the narrative viewpoint for so long as that character’s viewpoint structures the narrative – the entire novel, as in (3), or short fragments, as in (4) and (5). This ‘uploading’ of viewpoint to a higher level constitutes the narrative effect which will be discussed throughout the text – that of viewpoint compression. It is this compression to SV-space which yields the effects wherein the narrator is off-stage and yields the storyviewpoint to a character.
68
Stories and their tellers
SV-space
Narrator: 0 Reader: 0 Time: present
MN-space Character2: he/she Character1: he/she Character3: he/she
NS1
NS2
NS3
Tense: past
NS4
Figure 3.2 Ego-viewpoints
3.2.3
Omniscience and narratorship
There are thus many ways in which the omniscient narrator participates in the story construction. The narrator may be on-stage, and thus be given an independent voice, or off-stage, orchestrating the narrative. But in both cases the subjectivity talked about as a narrator is part of the topology of the SV-space. The third person in all of the instances discussed above is a device setting up a deictic centre of the narrating subjectivity outside of the main narrative space. Also the choice of the past tense serves similar purposes – it distances the main narrative space from the SV-space. The concept of ‘omniscience’ relies on the setting up of a deictic center independent of the story’s main space, which then provides a bird’s-eye view of the structure of the story. The choice of the subjectivity being on- or off-stage has stylistic consequences, but exerts much less influence over the way the story is read. At the same time, the narrating subjectivity, whether intrusive or completely backgrounded, is a source of a variety of blends, involving both the characters and the reader. The intrusive narrator requires that the reader or ‘narratee’ blend him/herself with the position of the addressee in the fictional deictic set-up of the SV-space. The interaction still proceeds only in one direction (from the
3.2 Types of teller and epistemic viewpoint
69
narrator to the reader, but not the other way round), but the reader is given the illusion of a personalized narrative account which addresses the reader’s storyconstruction needs directly and guides the reading process. In a sense, the intrusive narrator style is an interesting (even if now rare) narrative form which openly acknowledges that the story told is not necessarily sufficient for the reader to grasp all that is required. It is also interesting to point out that even the most intrusive of narrators is not restricted to the SV-space as if it were an ivory tower, and may actually signal some degree of involvement in the main narrative space. (6) He now lived, for the most part, retired in the country, with one sister, for whom he had a very tender affection . . . This lady was now somewhat past the age of thirty. (Tom Jones, p. 17)
In (6), the narrator tells a story which is presumably past with respect to the SVspace, and yet occasionally uses the adverb now in the way which is mostly found in free indirect discourse (see Nikiforidou 2010; also Chapter 7). In the case of (6) such constructions are found in narration, and not in represented speech and thought fragments. This is a result of the narrative set-up in which the narrator is explicitly communicating things to the reader, so that narration in this case turns into a form of represented speech, but of the narrator, not of a character.5 The narrator of Tom Jones does not act as a character as well, and yet needs to mark a presence both in the MN-space and in the SV-space – as a witness-like participant in the former and as a narrator in the other. Crucially, the use of now here is referring to the time of storytelling, not the time of the story’s events, so it helps establish the state of affairs in the MN-space with respect to the moment when the narration begins. It negotiates the relationship between the SV-space and the MN-space.6 The omniscient narrator in Tyler’s Ladder of Years is different. The reader is not an overt addressee, the narrator is off-stage, and the narration does not explicitly signal the SV-space. The narrator can naturally blend with the character and report the thoughts (Manual; not even electric.), but also comments on these same thoughts by evoking another moment in the story, whether previously narrated or not (She had worried she would find a computer.). Such narrative moments suggest that the narrator can access any point in the MNspace at any point in the narration. In a way, the narrator’s viewpoint alternately blends with the character-viewpoint in MN-space, or returns to the SV-space viewpoint, and this alternation between spaces constructs a specific form of omniscience. But the blend favors one character, so that the viewpoint alternation affects two stable viewpoints in two spaces. The difference between Tyler’s narrator and that of Mrs Dalloway is mainly in the proportion between the SV-space viewpoint and Ego-viewpoint. While Tyler’s narrator reports various actions and movements, Woolf’s alternates from character
70
Stories and their tellers
to character and intervenes only occasionally. There are no transitions from one Ego-viewpoint to another, or between the inner and outer states of characters, so that narrative spaces are naturally compressed into one flow of the narrative. The primary blend of this kind of narration is compressing the SV-space viewpoint with subsequent Ego-viewpoints in MN-space, and backgrounding the storyviewpoint. To sum up, these ‘omniscient’ blends are varieties of a similar initial set-up, and there are probably other varieties possible. What remains constant is some degree of profiling of the SV-space, while differences result from various degrees of compression of the distance between the narrator viewpoint in SVspace and any specific character viewpoint in the MN-space. The higher the degree of compression, the less narrating the narrator does and the closer the narrative gets to some variety of the ‘stream-of-consciousness’ technique. 3.2.4
Tense, person, and distancing
It seems that the distance between the SV-space and any of the narrative spaces is most saliently marked by the choice of the third person, not of past tense. Some texts, such as Neal Stephenson’s Snow Crash, maintain the third person and shift to the present tense (see Figure 3.3), and the effect may be similar to some of the narration modes described above: (7) The Oldsmobile keeps popping. Jason can’t think of what it is until he realizes that people are shooting at him. Good thing he let his uncle talk him into springing for full armor! When he figures that one out, he actually gets psyched. This is the real thing, man! He’s driving around in his Olds and the bastards are shooting at him, and it just don’t matter. (Snow Crash, p. 147)
The Ego-viewpoint here is maintained for some length of the narrative, and the characters whose viewpoints are thus represented change very often. The use of the present tense signals temporal viewpoint compression up to the SV-space, and the effect is that the narrative now favors keeping record of the on-line character’s thoughts and perceptions (The Oldsmobile keeps popping; This is the real thing, man!). In spite of the present tense, on-line form, the narrator still uses varieties of STR (he realizes that people are shooting at him), and narration, but the consistency with Ego-perspective is achieved differently. Even in the narrated parts (When he figures that one out, he actually gets psyched.) the narrator chooses expressions which would seem more appropriate to the character whose Ego-viewpoint is used here. It is thus an interesting blend of the narrator’s epistemic viewpoint and the character’s discourse style, which makes the subsequent Ego-narrator choices less conspicuous and gives coherence to longer narrative fragments. At the same time, viewpoint compression here affects not only the character viewpoint (as in Figure 3.2), but also temporal viewpoint.
3.2 Types of teller and epistemic viewpoint SV-space
71
Narrator: 0 Reader: 0 Time: present
MN-space Character2: he/she Character1: he/she
Tense: present Character3: he/she
NS1
NS2
NS3
NS4
Figure 3.3 Present time and present tense (example 7)
For comparison, first-person/present-tense narratives (typically not considered ‘omniscient’) go still further in compressing the SV-space and the MNspace, so that in effect the SV-space viewpoint and the MN viewpoint become the same (see Figure 3.4): (8) I open the window, turn up the volume. I pass two cars at once and get on the highway . . . Through the toll, the light, onto a ramp, onto the bridge. Now I can’t turn back. The Oakland shipyards to the left, a billboard encouraging the saving of water. (AHWOSG, Eggers, p. 126)
In (8), the narrator and the character subjectivities and viewpoints are fully blended, but the SV-space does not disappear. The record of events such as opening the window or passing other cars narrates the driver’s actions and measures the passage of time since Dave (the Ego-narrator) left home, independently of his on-line record of perceptions (shipyards, billboard), and thoughts (I can’t turn back). What seems to be the most salient characteristic of this narrative mode is the completeness of the narrator/character blend, so that all parts of the story, including the events which took place prior to the storyline in the MN-space, are told with the same voice.
72
Stories and their tellers SV-space
Narrator: ‘I’ Reader: 0 Time: present
MN-space Main Character: ‘I’ Character2: he/she
Tense: present Character3: he/she
NS1
NS2
NS3
NS4
Figure 3.4 Ego-viewpoint in the present tense (example 8)
There are small parts of the story told in the past, but the difference is that the ‘past’ parts are not intended to be an on-line account of one person’s participation in the events, but are a more ordinary distanced narrative. An example from the same novel is given in (9): (9) When my father was in intensive care, about a day and a half from throwing in the towel, a priest was sent, presumably to administer last rites. After meeting him and ascertaining the purpose of the visit, my father quickly dismissed him, sent him out. (AHWOSG, p. 40)
Characteristically, the style of passage (9) is markedly different from (8), as the events are rendered in distanced form, including past tense, markers of epistemic distance (presumably), passive voice (was sent) and formal vocabulary (ascertaining the purpose of the visit). These brief moments when Dave leaves his frantic on-line persona aside and assumes a narrating role with all seriousness of purpose mark viewpoint shifts, but also reconfigure the type of narrator Dave is and can be. While most of the text is in the frantic present-tense, those composed past-tense parts may be read not just as shifts of temporal perspective, but also as glimpses of a different ‘narratorship.’ While both (8) and (9) belong to the book’s MN-space, their style is adjusted to what is being narrated – past
3.2 Types of teller and epistemic viewpoint
73
family events can be presented in a reflective and serious way, but the ongoing roller-coaster of bringing up a young brother calls for a different narrative style. The multiplicity of smaller spaces set-up is also represented in Figure 3.4 by the smaller circles around more major narrative spaces. The types of narration sketched above are distinguished, via constructional means, as reflecting a specific relationship between the SV-space and the MNspace. The compressions along the lines of temporal distance (past or present) or the distance between a deictically determined communicator in the SVspace and a character in the MN-space (first or third person) yield several types, even though they are not always clearly distinguishable. The important point is that in most cases the specific viewpoint configuration remains as the default set-up throughout a novel, so that various specific parts of the narrative structure are subjected to the specific SV/MN relationship. For example, all the numerous Ego-viewpoint blends in Mrs. Dalloway are naturally assumed to be subjected to the MN-space of the day of Clarissa’s party. Similarly, the narrative of Snow Crash blurs the Ego/narrator boundaries through consistent stylistic form. Also Eggers’s narrative, which is less stable in its viewpoint setup, remains within the bounds of an Ego-narration. Features of style, however, contribute significantly to how the narrative space set-up will be framed. Eggers’s narration is scattered and shifts narrative-space-viewpoint all the time, and thus, by comparison, the characters in Mrs. Dalloway seem to ‘think’ in voices too similar and stylistically colorless to represent abrupt viewpoint shifts, even though the Ego-viewpoint changes all the time. The stylistic dimension (choice of register, degree of distance) thus seems to create the quality of narration which obliterates or highlights the viewpoints profiled. The smoother the transition from narrator to Ego, the more cohesive the narrative seems. The scattered quality of Eggers’s story persists even when the Egoviewpoint is maintained. Example (10),7 which follows (8) immediately, without even starting a new paragraph, has to be read from the point of view established in (8) (Dave, driving to San Francisco, after he left his brother with a babysitter): (10) I will come home and the door will be open, wide. The baby-sitter will be gone and there will be silence . . . Blood on the walls, handprints soaked in blood . . . There will be a hearing, a trial, a show trial – How did you come to meet this man, this baby-sitter? We found a posting, on a bulletin board. And how long did your interview of him take? Ten, twenty minutes. (AHWOSG, p. 126)
Most readers would agree that Dave’s worries about his brother are exaggerated, so that he could be viewed as an ‘unreliable’ narrator. The concept was
74
Stories and their tellers
introduced by Booth (1961), and was recently discussed by Semino (2007) as related to what stylisticians refer to as ‘mind style.’ Whichever term is used, the point seems to be that some narratives are relayed through an Ego who is either unaware of or unable to grasp the nature of the events (for example, Benjy in Faulkner’s The Sound and the Fury). As Semino rightly points out, there are a variety of ways in which the specificity of the constructed mind influences the narrated content. The so-called unreliable narrator is reliable enough for the reader to construct the story, but the narrative viewpoint is indeed expressed somewhat differently. Primarily, unreliable narrators are presented as only capable of a very specific epistemic viewpoint, so that the narrowness of the understanding of events strikes the reader as inaccurate. Whether such a narrator is accessible through the first-person or third-person style, other viewpoints are not directly available. It seems that such cases constitute another blend, such that the SV-space is entirely blended with the MN-space relayed through one specific Ego. Such a narrative space naturally calls for the reader to construct alternative viewpoints, prompted only by the bizarre consequences of following the viewpoint available. In the case of (10), the readers understand that a parent would most likely not be put on trial for leaving a child with a babysitter, but does it make Dave ‘unreliable’? He is, after all, narrating his obsessions and worries, which are a crucial part of the story told. It seems that the viewpoint is entirely subjected to the goals of the narrative, and such goals may not include giving a reliable account of events. As the examples above suggest, the goals are more often focused on representing the character’s epistemic and emotional perspective. The examples above all point to the specificity of distance in narrative discourse. Dancygier and Vandelanotte (2009) and Vandelanotte (2010) discuss the varieties of distancing devices in literary texts – specific forms of represented speech and thought, temporal distance, metalinguistic distance, epistemic and emotional distance – mostly available through textual access to more than one viewpoint-marked narrative space. In the fragments considered above, the concept of narratorship itself seems to emerge from the varieties of temporal and epistemic distance between the major spaces involved (SVspace, MN-space, Ego-space). While the temporal distance (e.g., present SV/past MN) and epistemic distance (SV access to story/Ego access to story) seem to be the most common devices, the stylistic features marking emotional, evidential or social distance contribute substantially to the impact of the text, including perception of viewpoint. Ironically, Mrs. Dalloway may seem more coherent, in spite of its galaxy of narrating Egos, while Eggers’s text feels scattered, even though there is only one Ego narrator, because its emotional temperature keeps jumping up and down and his mind wanders in most unlikely directions. Overall, the viewpoint texture of a narrative as a whole depends as much on the general space set-up as on the stylistic profiling
3.2 Types of teller and epistemic viewpoint
75
of various aspects of viewpoint in those spaces and the foregrounding or backgrounding of various types of distance. 3.2.5
Constructional compositionality
Viewing the narrative options described above as constructions highlights important facts about interpretive consequences. There seem to be four constructional parameters involved: narrator on- or off-stage, person, tense, and the profiling of Ego-viewpoint (or several). All these parameters represent interactions between the SV-space and the topology of the MN-space, suggesting various degrees of distance and compression. The degree of profiling of the SV-space correlates with the degree of distance between the SV-space and the MN-space. If the narrator is an onstage part of the SV-space, the degree of compression with the MN-space is rather low, as the storyworld and the character are not seen as participating in the narration. For comparison, an off-stage narrator invites less distance and more compression, especially with many Ego-viewpoints profiled. The yielding of narrative voice to Egos (as in the fragment from Woolf) downplays the role of the SV-space, reducing the distance to the temporal dimension, and yielding a narrative where the character perspective dominates and the organizational role of the narrator is not foregrounded. Intrusive narration and multi-Ego narration are thus two extremes along the spectrum of narratorial involvement. In this context, the choice of first-person or third-person narration is also dictated by the degree of distance involved – with first-person narration being fully compressed into the SV-level (so that the independent narrating voice is not construed by the set-up), and the third person giving the narrator opportunities to intervene and comment. The choice of tense is less consequential. I have assumed that the SV-space is to be treated as ‘present,’ since the narrating is always done in the narration’s present. This yields a rather straightforward set-up in the case of narration in past tense, as it only underscores the distance between the SV-space and the MN-space. Narration in the present tense, however, does not suggest that the time of the story’s events is the same as the time of the story-telling. Rather, it diminishes the distance between the spaces, and allows the story’s time to be compressed into the SV-space – thus, even if the SV-time is present time, it does not mean it is the same as the story’s present tense, which will still be past time, wih respect to SV. There is still conceptual distance between the SV- and MN-spaces. This brief overview suggests that the actual form of any given narrative emerges out of a combination of these constructional parameters, and is primarily defined by the degree of remaining distance between SV- and MNspaces. Crucially, not many narratives are entirely consistent in the choice of the narrative space configuration, and such shifts can be explained only against
76
Stories and their tellers
the template of the SV-space. The SV-space thus needs to remain separate from other narrative spaces, so that the overall coherence of the story can be established at some stable narrative level. 3.3
Second-person narratives
Some of the issues raised above can further be clarified with a look at the less common form of narratives – with second-person pronoun as the primary SVspace-related choice. These instances are related to the SV set-up described above, but exploit it in different ways. In fact, there are more significant differences across examples than, let’s say, among third-person narratives. One of the texts where the second person dominates is Oriana Fallaci’s novel The Man, describing the life of Alekos Panagoulis, her close friend. She dedicates the book to him (For you), then, in the introduction, promises to tell your tale and describes Panagoulis as my only possible interlocutor. The story follows, with Panagoulis consistently referred to as you. (11) You thought about the amnesty all night, sometimes believing it, sometimes not, and when you didn’t believe it, you felt serene, when you did believe it, your conscience was split in two. (The Man, p. 132)
Interestingly, the second-person form is used here to represent the inner thought of the main character, so that the text contains all the narrative elements the reader expects. There is also an I in the novel, referring to Fallaci herself, also framed as the narrator: (12) As if stung by a wasp I stood up. I said I had to leave you and go find a hotel. You answered categorically: “You’re not going anywhere. You stay here.” (The Man, p. 150)
The novel thus explicitly sets up a deictic frame, with the writer as an equivalent of the speaker, and the main character as the addressee (see Figure 3.5). The addressee position here is not that of the reader or narratee, who remains a standard recipient of what the text communicates. To accommodate the you, we need an extra space of the reality, in which Fallaci and Panagoulis were friends, and we need to establish a cross-mapping between the two of them as people and as narrator/character and character/addressee respectively. The novel’s set-up thus explicitly comments on reality and fiction, while also blending the two grounds – the imagined deictic ground wherein Pangoulis is an addressee, and the novel’s narrative ground, where the addressee is the generic reader. The set-up is different in another novel which relies heavily on you – Calvino’s If on a Winter’s Night a Traveller. The structure of the text is very interesting, with the novel itself partially embedded in it, and with various excursions into other (presumably nonexistent) novels. You in the text refers
3.3 Second-person narratives SV-space
77
Discourse space Speaker: Oriana Hearer: Alekos
Narrator: ‘I’ Reader: 0 Time: present
Reality-space Oriana Alekos MN-space
Tense: past Character: Oriana
Character: Alekos NS1
NS2
NS3
NS4
Figure 3.5 Second-person narrative (examples 11 and 12)
to the reader, but the reader is present not only in the SV-space, as is usually the case, but also as a character whose experience of reading and interacting with the text is the primary theme of the novel: (13) The novel you are reading wants to present to you a corporeal world, thick, detailed. Immersed in your reading, you move the paper knife mechanically in the depth of the volume. (If on a Winter’s Night a Traveller, p. 42)
The present tense and the progressive aspect in (13) also suggest that the intended construal is that of the ongoing process of reading (with the exception of fragments from embedded imaginary novels, which are differently framed). The MN-space is initially structured as taking place in a railway station, with the reader as one of the characters and the narrator as the other. The text also implies that the scene is developing within the reader’s field of vision: (14) I am the man who comes and goes between the bar and the telephone booth. Or, rather: that man is called “I” and you know nothing else about him . . . (If on a Winter’s Night a Traveller, p. 11)
However, the reader’s observations are relayed through the narrator’s voice, as in (15). (15) It is obvious that I am a subordinate, I do not seem the sort of man who is travelling for personal reasons or who is in business for himself. (If on a Winter’s Night a Traveller, p. 14)
78
Stories and their tellers
SV-space
Discourse space Speaker: Narrator Hearer: Reader
Tense: past
MN-space
Character-Reader: ‘you’
NS1
Narrator: ‘I’ Reader: 0 Time: present
NS2
Character-Narrator: ‘I’
NS3
NS4
Figure 3.6 Second-person narrative (examples 13–15)
Similarly to The Man, the novel profiles the narrator and the addressee as characters, but neither of them is correlated with reality, so that the crossmappings reach across MN-space and SV-space (see Figure 3.6). The two characters are thus compressed up to the SV-space, but the narrator retains his ability to reach into the mind of the character, even when the character’s thoughts are about the narrator as character. The identity of the I character is not developed to include an independent subjectivity, and he is reduced to his narratorial function and the perception of the reader-character. Second-person narratives (cf. Fludernik 1994) thus rely on mechanisms described above for first- and third-person texts. The presence of the narrator and reader slots in the SV-space is exploited for cross-mappings. However, while the reader is only profiled in the text in intrusive narration, in the examples above other cross-mappings are set up – with a character in the MN-space, or with an addressee in the reality space. Furthermore, the viewpoint configuration such as in (14) and (15) includes a blend wherein the I-narrator does not report his or her own story, but another character’s, cross-mapped with the reader. Finally, such a set-up does not invalidate the added layer of narrative space structure wherein there is an actual reader reading Calvino’s novel.
3.4 The teller, the author, and the character
79
This discussion has not mentioned the narrative use of you which is very similar to the colloquial discourse usage – the generic you, as found in suggestions, how-to descriptions and other contexts where a general pattern of behavior is being described. Because this generic form is not specific to narratives, I will not discuss it any further. Thus far, I have tried to trace some of the constructional correlates of general viewpoint mechanisms in a fictional story. The tense and pronoun choices discussed here are specifically those which can be naturally explained in the overall narrative space set-up. There are also more local, micro-level pronominal choices, sometimes affecting the concept of narratorship, but these will be discussed in Chapter 5. At the same time, there are also text-wide consequences of how the teller of the story is construed, especially with respect to its author, as tellers can be discussed from the point of view of the overall coherence of the text. The question of how the conceptualization of the teller affects the story will be considered in the remainder of this chapter. 3.4
The teller, the author, and the character
Recent work by J. M. Coetzee explicitly addresses the complexities of the role of the teller in the understanding of a text (see Dancygier [2010] for broader discussion). Coetzee himself is an author of fiction as well as nonfiction, and his recent novel, Elizabeth Costello, combines the two. In the text, he includes various articles and lectures he wrote or delivered as J. M. Coetzee and attributes them to a fictional author, his alter ego Elizabeth Costello. The resulting blend represents authorship from two perspectives: on the one hand there is the subjectivity named on the cover, who only speaks through the authored text, while on the other hand there is the actual person, having reasons to write what she does. Costello is different from any other author we know only from a vignette on the dustjacket in that she also represents the desire to be understood and heard, strong personal beliefs, and tiredness with fame. It is natural to read her as a blend of the fictional subjectivity and that of the actual author. In the next novel, Slow Man, Costello appears again. She walks into the apartment (and life) of a character, Paul, who loses his will to act after a crippling accident. This narrative construction, whereby there is an author inside the narrative space, gives a view of authorship which questions a number of default assumptions. While other characters in the novel as well as its critics suggest that Costello’s interventions are those of an author writing the novel who wants to put Paul in her book, this seems to be an oversimplification. True, Costello tries to prompt Paul to action, but not in order to use the events as narrative material, but rather in order for the narrative to ‘take place.’ As she (or Coetzee) sees it, it is the character’s responsibility to ‘create’ the story by living it. The story of Slow Man does not exist in Costello’s (or anybody else’s) mind,
80
Stories and their tellers
and the reader thus reads a novel as existing in a narrative space independent of the author’s or narrator’s or character’s subjectivities, but driven directly by their agency as characters and writers. The presence of Costello as an embodied narrative choice (to live a life worth writing about and let the novel emerge from it) constantly reminds Paul that a different story could be lived and told, but if we recall that she is in fact Coetzee’s alter ego, the choices she offers are choices Coetzee might be associated with. In effect, the character is not ‘put in a book,’ but the author is. In the next novel, A Diary of a Bad Year, Coetzee’s next alter ego, Señor C., speaks more openly of his role and claims his right not to be seen as ‘merely’ a set of rhetorical tricks. Coetzee himself is a critic, but in his own case, he speaks through his novels on his own authorial viewpoint and his understanding of what fiction is. He seems to find the pronouncement of the author’s death (even if it is meant just to be provocative) to be an insult to his personal and intellectual involvement in the process of writing and, while he is willing to yield most of the power to the character and the logic of the story, he does not like the idea of giving it all to the reader. To sum up, Coetzee’s recent work suggests a re-evaluation of the narrative which goes in two directions. On the one hand, fiction needs to be freed from the overpowering subjectivity of a teller, and given its own place as a representation of a story. On the other hand, stories and characters need their authors to take shape and be convincing. While the author cannot write a story that a character refuses to live, life is full of stories which are lived in a way worthy of an author’s attention. The viewpoints which thus emerge are all subjected to the viewpoint of the story as such. 3.5
Multiple tellers
The question of narratorship and the cognitive status of a teller becomes particularly interesting in the cases where the text as a whole profiles more than one teller, while ostensibly telling the same story. The case of narratives which rely on various focalizers is not the one under scrutiny here; instead, I will focus on cases where the text designates different characters or different identifiable narrators to tell different partial stories. I have discussed one such case in an earlier work (Dancygier 2008a), Jan Potocki’s The Manuscript Found in Saragossa. The text has an interesting history, and a still more unusual composition. It was written in French by Jan Potocki, a Polish aristocrat living and writing in the late eighteenth and early nineteenth century, whose life story could make a fascinating novel of its own. The text was first published in two different parts, so that even reconstituting the entire book in its intended form was a historical and critical challenge. The ‘novel’ consists of 165 stories (or so one story-counter claims), connected in different
3.5 Multiple tellers
81
ways through embedding and digression, told by a number of different tellers, and yet, in some way, constituting a coherent narrative of the story of the main character, Alphonse van Worden, and of the Moorish family of the Gomelez, to whom he is related through his mother. The intricacies of the narrative in this case challenge the reader in many ways, as even keeping track of which level of embedding a given story belongs to is not a straightforward task. As I argued elsewhere (2008a), the book does form a coherent whole not because the stories naturally form one overarching storyline (they don’t), but because various stories contribute to the establishment and maintenance of viewpoints. In other words, tellers are not necessarily responsible for a piece of narrative structure, but their stories represent ways of interpreting the overall story being told, and thus the whole is not only a fascinating narrative, but also a reflection on honor, religion, tolerance, the value of wealth, as well as mathematics, philosophy, and history. One of the most salient questions is the choice between religion and faith on the one hand (this is the time of the Inquisition!) and honor on the other hand, as principles guiding a moral man. Indeed, these are the choices the main character, Alphonse, faces throughout the text, but the options are presented through different stories, told by different people in different places and times. Thus there are a number of stories whose conclusions suggest that faith is the best foundation, while others could be read as mocking religious devotion, and still others reinforce the idea that honor is the best guide. Within his own story, Alphonse chooses honor, and gains respect as well as riches. The power of the Manuscript, then, is in the setting up, elaboration, and reinforcement of viewpoints, while their continued presence and salience in the text are constructed through a variety of smaller narratives, none of which can sustain the subtlety and multidimensionality of the moral and philosophical debate. Thus the number of ‘tellers’ and narrative contexts is the source of variety and interest, but the main viewpoints guiding the story and its main character are established over and above the various narrators, through the construction and selection of the stories themselves. There are also cases where the story profiles more than one teller but in the end remains unspecific as to the subjectivity which the story is to be attributed to. There are numerous examples of this in contemporary literature, but I will discuss one in detail here – Mikhail Bulgakov’s The Master and Margarita (M&M) – to propose also an analysis of how the misleading structure fits the viewpoints represented. The novel has two major narrative spaces: the story taking place in Moscow, in the late 1920s, told by an omniscient intrusive narrator (M&M1), and a novel about Pontius Pilate and a wandering philosopher Yeshua who is crucified at the end (PP&Y). The narrator of the PP&Y novel is also omniscient. M&M1 is a story of frightening, but also hilarious events which occur in Moscow as a result of the visit by Satan (traveling as Mr. Woland) and his team. Among other
82
Stories and their tellers
things, Woland organizes a ball where all the sinners and souls committed to eternal damnation have a day’s break. He needs a lady to accompany him, and an appropriate woman is soon found. Her name is Margarita. The PP&Y novel was written by a man living in Moscow, who only appears in the text as the Master (his name is unknown). In the novel, the story involving Pontius Pilate, Jesus (here Yeshua), and Judas, and leading to the Crucifixion is told in an entirely realist mode, with no apparent allusions to the religious context of the story. The writing of the novel and the reasons why it was never published (in fact, the manuscript was burned) form one of the narrative themes in M&M1, where the Master is one of the characters. From the narrative perspective, the most interesting aspect of the PP&Y novel is how it is given to the reader. The first installment is told by Woland (Satan) to two incredulous men – the critic Berlioz and the poet Ivan Bezdomny. However, it is presented as an eyewitness account – Woland describes what he saw and heard in Pilate’s palace. Ivan the poet is shocked by this encounter and eventually ends up in a lunatic asylum, where the next fragment of the novel appears to him in a dream. Finally, Margarita recovers the entire manuscript (Satan gives it back to her, adding that ‘manuscripts don’t burn’) and reads it (throughout the remainder of the M&M1 story), which is how we get to the end of the text. The last words, as announced earlier by Margarita, are “procurator of Judea, the equestrian Pontius Pilate.” The text of the PP&Y novel is thus presented at once as a novel on a subject entirely divorced from the reality of Moscow in the post-revolution years, and as an account of Pilate’s encounter with Jesus and of the crucifixion, which is fictional (it is a novel), but also factual (eyewitness report). On yet another level, Ivan and the Master are both insane, and Margarita, by usual standards, loses her sanity as well – so perhaps these are just ravings of lunatics – but then, how come they give a very sane account of events which are typically seen in more otherworldly terms? The choice of narrators is the only source of such questions. The search for authorship within the storyworld is equally elusive. PP&Y is the novel written by the Master (even though it exists beyond the manuscript or the author’s consciousness). But in the final scenes the Master is allowed to free Pontius Pilate and is thus still responsible for his characters and has the power to change their fate – as we expect a true author to do. But now he not only ‘writes them,’ he can change their lives in the reality the novel represents, which is now a part, though an invisible part, of the reality as we know it. Also the authorship of M&M in its entirety becomes questionable. The final words of the Master’s novel, “procurator of Judea, the equestrian Pontius Pilate,” also end the story of Ivan the poet, and the entire book. The words become a powerful narrative anchor which marks the limits of authorship as based on magical insight into the truth, not on the power to create fiction.
3.5 Multiple tellers
83
In this sense, the Master, the poet, and the actual writer Bulgakov are all candidates for authorship-as-insight, while the Master and Bulgakov are blended as possibly actual writers of the words. Indeed, there is much in Bulgakov’s biography that matches the misfortunes of politicized criticism which destroy the Master, but, first of all, the final words seem to be the ultimate wink of this self-ironic and at times amusing, at times deeply moving novel. It does not really matter who wrote what, just in case some eager critic will want to use it against the writer. The way in which Bulgakov plays with the concept of authorship and with the boundary between fact and fiction is of course deliberate. The truth or fictionality of the story of the Crucifixion is crucial to the worldviews which are competing in the historical context throughout Bulgakov’s life. The political impact of Bulgakov’s text is beyond the interest of this analysis, but there are important narrative lessons to be learned. There are things authors can talk about which are just for the enjoyment of their readers, but there are also things in socalled-fiction which are deeply true, even though there is no proof of their truth in any historical document. It is the truth not owned by the author, but perhaps shown to him in a moment of reflection, or in a dream. The blurring of all crucial narrative categories in this important novel is deliberate, and it is achieved through manipulation of narrative structure worthy of a true Master. From the point of view of the emergent story, the case of M&M is also interesting as a result of its choice of tellers. The two primary narrative spaces (M&M1 and PP&Y), which start out as separate (the ‘Woland-in-Moscow’ story and the Pilate/Yeshua story) go through interesting changes of status as the text develops. Initially, they are linked through Woland, who is a character in M&M1 and the first teller of the PP&Y part. Later, the link is further developed through Ivan’s dream. Then Ivan meets the Master and the Master’s novel first enters the narrative scene. The fact that the reader has thus far had access to PP&Y through major characters in M&M1, and can only then reconstrue the story as the Master’s novel has crucial importance for the narrative structure. The progression is from eyewitness report to a dream, to a novel, and thus the truth-or-fiction issue is set up from the start as impossible to resolve. Having different narrative spaces as homes for construals of the Crucifixion story makes it possible for these spaces to maintain three different viewpoints throughout the text, even though the sequence of events told is not disrupted at all – Woland starts the story, Ivan’s dream continues it, and then the Master’s novel takes care of the rest. This example supports the suggestion made earlier that the sequentiality of the emergent story is of secondary importance to the meaning of the narrative, while available viewpoints shape the emergent structure in substantial ways. To put it simply, if the Crucifixion story is true, then Satan exists and can interfere in contemporary affairs, while Ivan and the Master are sane and have been given access to the truth. If it is a delusion or fiction, then we have read a
84
Stories and their tellers
funny fantasy story. This choice is established textually at the beginning, since the first scene of the novel has Ivan and the critic, Berlioz, discussing the question of whether Jesus really existed or not. The whole conversation can thus be viewed as a narrative way of establishing the two viewpoints which drive the ensuing narrative – similarly to Potocki’s text. The viewpoints, not the facts told, decide on how this book is read and understood, and both viewpoints remain active till the very end. 3.6
Narrative space embedding
Thus far, I have looked at instances where the SV-space either has little structure or is only a source of temporal and epistemic viewpoint, but plays little role in the story. However, there are also cases of straightforward space embedding, where the SV-space contributes elements of the story and often changes the emergent story significantly.8 The role of embedding in emergent narrative structure can be illustrated with the narrative space configuration of Ian McEwan’s novel Atonement. The text, a typical third-person/past-tense omniscient narrative, consists of three parts. It starts when the main character, thirteen-year-old Briony Tallis, writes a play, The Trials of Arabella, a rhymed story full of kitsch emotions and childish ideas of love. Although the play does not participate in the story proper, it provides an opportunity to characterize Briony as an aspiring writer, a person prone to using big words, and treating fiction as seriously as she treats life. The play foreshadows the coming events, where Briony witnesses sexual encounters between her sister, Cecilia, and a man named Robbie. She interprets these events as Arabella probably would have, as signaling brutality rather than romance. After another character helpfully provides the term ‘maniac,’ Briony has her vocabulary to describe Robbie, and when another girl is raped, she blames it on him. This part of the narrative clearly defines Briony’s over-excited confusion of the adult world around her with the fiction she writes. In a sense, Briony tells two fictional stories – of Arabella and of the rape, both structured with the same authorial disposition. The story develops dramatically. Robbie is sentenced, then freed to join the army, while both sisters take part in the war effort as nurses. Later in the war, Robbie returns to England, and Briony meets with him and Cecilia to acknowledge that her accusations were wrong. This ends the main narrative space of the novel, and the text is signed “BT, London 1999.” This brief addition, however inconspicuous, changes the entire viewpoint structure of the novel. The narrator is no longer a creation of the (implied?) author, McEwan, but the main character, Briony. The novel we read is not the entire story constructed by McEwan, but only its part. And we also learn why the text had to be written, as the last words are: She knew what was required of her. Not simply a letter, but a new draft, an
3.6 Narrative space embedding
85
atonement, and she was ready to begin. A “new draft” of what? This narrative anchor, as well as that of the date, are crucial to the final shape of the emergent story. The fourth part of the novel, titled London, 1999, mentions various drafts of the novel (1940, 1947, 1999) and acknowledges that Briony has been changing the facts from draft to draft until she finally decided to tell the whole truth of her crime, but not say what really happened to Cecilia and Robbie. She suggests a possibility that they died in the war and never saw each other after the police took Robbie away, but we can only guess that that was what ‘really’ happened. Throughout the majority of the book the text constructs a viewpoint of a narrator who knows what happened and does not judge anyone, but in its last twenty pages it changes everything. An additional narrative space is set up where the true author writes a story which will count as ‘the truth,’ including the name of the real rapist, now married to the girl he raped. It is in fact an explicit construction of an SV-space. The ‘embedded novel’ was developing throughout the first 350 pages as an ordinary set-up with a narrator off-stage, and a temporal and epistemic viewpoint yielding a third-person/past-tense narrative. But the final revelations add a lot of topology to the SV-space, which now has a firstperson/present-tense narrator who also identifies herself as the author of the story told in the MN-space. The MN-space thus becomes an embedded space, in the sense typically used in mental spaces theory, which means, among other things, that the stance and viewpoint of the higher SV-space are projected into the lower MN-space. This means that the topology of the ‘embedded novel’ space now inherits the relevant bits of the topology of the SV-space, so that, among others, there is a cross-space connection linking Briony-the-character to Briony-the-author. But the epistemic viewpoint of the MN-space had to change, in that we now know that the ‘true’ story resides in the SV-space, along with all the previous drafts where different names were used and different circumstances were described. The new set-up also crucially changes the intended sense of atonement, which now has three different meanings. In the embedded MN-space, atonement for Briony’s guilt is achieved by clearing Robbie’s name and accepting the blame. The novel itself, as a text to be read, counts as atoning for the crime by changing its result and giving Robbie and Cecilia a life together. In effect, for a reader in the space in which the embedded novel exists as an artifact, there is a happy ending. For the author, however, atonement is only to be sought. The various drafts she alludes to, with different versions of what happened, are attempts, and she finally settles for the version which is the closest to the truth, but changes the ending – and that is how atonement might be achieved. To conclude, the establishment of the SV-space is consistently used to structure the story’s general viewpoint and to put the MN-space story in perspective, or even change its structure. Regardless of whether the set-up is
86
Stories and their tellers
textually obvious, or whether the embedding space is smoothly blended with the embedded space, as in the Tyler or Woolf examples, the narratives discussed here rely on the embedded set-up and its effects – primarily on the projection of SV-space viewpoint into the MN-space.
3.7
Narrative viewpoint and narrative spaces
What the examples above suggest is that one might consider narrative viewpoint in terms independent of the subjectivity of a specific teller, but instead as a feature of the narrative space. Major narrative spaces, being mental constructs and basic units of narrative structure, are equipped with rich topology, including spatial and temporal grounding, the record of events and/or conversations, and, first of all, the characters. In the various cases of profiled narrators, as described in Section 3.2, the viewpoint is at least partially relayed through the fictional subjectivity of the narrator. The need for the ‘narrator’ construct, whether independent of the characters in the space or blended with one of them, emerges out of the default communicative set-up, whereby content communicated should be attributed to a communicator. But the presence of the construct does not do much more than provide a focal point which gives the reader access to the topology of the space. What we have seen in the texts briefly described above is that narrative viewpoint can be complicated by creative manipulation of spaces, rather than reconstrual of narrators. There may be additional spaces set up, or viewpoint may be spread over a number of different narrative spaces and tellers, as in Potocki’s novel, or the space topology and epistemic viewpoint may alternatively match one character or another or remain shared across two subjectivities. Finally, as in Bulgakov’s text, distributing narratorship across different subjectivities can contribute to the establishment of multidimensional or competing viewpoints. In none of these cases is the viewpoint uniformly attached to a subjectivity of a narrator; rather, it is a characteristic of the narrative space. While it is perhaps more typical for narratives to profile tellers, they remain a structural solution to a conceptual problem – how does the reader get access to the facts s/he needs in order to comprehend the story and see its broader implications. At the level of the reader’s construal of the emergent story, the selection of an aspect of narrative topology to serve as the focal point is a way to direct the story-construction processes.
4
Viewpoint: representation and compression
There were two men in my father’s chair. (Coasting, Jonathan Raban)
The preceding chapters have outlined ways in which narrative viewpoint is construed and linguistically expressed at the level of major narrative spaces of a story. However, viewpoint phenomena of more partial nature pervade narrative discourse at every level. There are several terminological issues which require clarification at this point. ‘Viewpoint’ and ‘perspective’ are the broadest terms in use, with respect to both macro-level phenomena (such as narratorship, temporal and spatial anchoring of the narrative, et cetera) and micro-level phenomena, which are often talked about in terms of Genette’s (1980) concept of ‘focalization,’ or ‘who sees.’1 Much of this terminology evokes visual access to situations, though the nature of focalization may include other ways in which the narrative is aligned with a fictional subjectivity, whether of a narrator or of a character. To avoid the visual implications, Toolan (2001) proposes to talk about ‘orientation’ instead, which in turn evokes spatial configuration of some kind. All three terms, ‘focalization,’ ‘orientation,’ and ‘viewpoint,’ in fact cover a broad range of phenomena having to do with the specific use of narrative space topology – temporal, spatial, perceptual, and the like. Much of the relevant discussion is also complicated by the assumption that there is a difference in narrative discourse between stylistic forms which ‘tell’ (such as narration) and those which ‘show’ (those which are assumed to represent, rather than describe) (cf. Booth 1961; Rimmon-Kenan 1983). The distinction, while useful in many ways, calls for a clarification of other concepts. Narration, in this sense, is the domain of the narrator, who (which?) is a narrative construct only – so if ‘telling’ the story is what is indeed talked about, this involves rather strong assumptions at play about a narrating subjectivity or a ‘teller.’ In fact, ‘telling’ seems to be a more complex concept. In Chapter 3, I suggested that there are three sources of the conceptual set-up underlying textually determined construction of the narrator’s ability to tell the story: the viewpoint configuration involving the SV-space and the MN-space, the reliance on the default communicative set-up, and the choice of grammatical forms such as tense or person. In 87
88
Viewpoint: representation and compression
such a set-up, the ‘showing’ would also be the domain of the narrator, except that the ‘voice’ is temporarily given over to a character. But how can a character have a ‘voice’ independent of the narrative choice of the specific situation and wording? The so-called direct speech in narrative fiction should not be taken as a token of plausible conversational discourse, as any discourse analyst knows very well. Natural discourse is full of interruptions, hesitations, overlaps between speakers, unfinished sentences, discourse markers, et cetera, in sharp distinction to the narrative representation of discourse. The idealization does not consist only in smoothing out the performance, but also in very careful choice of which exchanges between characters will be represented and how. Direct discourse in fiction is as ‘narrative’ as narration or free indirect discourse, but it is designed to ‘tell’ (not really ‘show’) the story through a different conceptual medium (for further discussion of direct discourse see Chapter 7). ‘Viewpoint’ as a feature of a narrative space, rather than of a fictional mind, is also a generic concept which underlies much of the discussion of represented speech and thought, especially forms such as ‘free indirect discourse,’ which have been claimed to represent a ‘dual voice’ (Fludernik 1993). One of the ‘voices’ then would be that of a narrator, and the other of the character, but both types of fictional minds are primarily a source of viewpoint, rather than discourse as such. The difference between a fictional mind which is ‘seeing’ and one which is ‘speaking’ may be partly tenable with respect to represented speech, which would then be an instance of ‘speaking,’ but does not seem justified in application to thought, given that the unspoken content of characters’ minds definitely involves those characters’ viewpoints (or ‘seeing’) and is to a much lesser degree aligned with specific forms of expression. The forms of speech and thought representation (STR) are usually considered to be an independent narrative phenomenon, and are definitely distinguishable as linguistic constructions (see Vandelanotte 2005, 2009 for a very thorough analysis), but they do assume a viewpoint of a speaker or thinker. What is more, the indirect types of discourse described within STR are often (misleadingly) talked about as representing actual words or thoughts, while they can at best be claimed to be narrative-specific representations of their content.2 Yet another question is the distinction between ‘speech’ and ‘thought’ in STR, although in most cases both are assumed to be represented by almost identical constructions, whether described as direct or as indirect.3 In natural language use, the distinction is crucial, in terms of addressivity, accessibility, and the specificity of propositional content. In narrative fiction, for comparison, the distinction is almost nonexistent, and I will discuss more specific exmples in Chapter 7. Generally, I will assume that both have the role of representing conceptualizations, with little need to distinguish sharply between communicated and noncommunicated ones, since the overarching goal is to communicate them both to the reader.
Viewpoint: representation and compression
89
It seems that the treatment of ‘thought’ in narrative discourse emerged out of the possibility to use similar ‘reporting’ constructions to handle both speech and thought. In spoken discourse, it is equally possible to say Patsy said she was in love or Patsy thought she was in love, and Patsy’s disposition thus described is the same. What is different is the management of epistemic commitment of all the subjectivities involved. In the ‘say’ example Patsy went on record about her feelings, or the ‘narrator’ is going on record about correctly interpreting Patsy’s words (which may have been different). In the ‘think’ example there are more possibilities. Patsy may have said something suggesting she suspected her ‘symptoms’ indicated being in love (so the speaker is interpreting); or she may have communicated what her suspicions were (such as I think I am in love, though I’m not sure); or the speaker indicates that Patsy did think she was in love but no longer does; or the speaker means to suggest that Patsy was wrong, even if Patsy didn’t realize it at the time. All these interpretations arise out of various epistemic stance configurations prompted by the constructions, but the possibility which is in fact least likely is that the speaker is ‘reporting’ Patsy’s thoughts as they were, simply because in natural discourse we do not assume access to other minds, we can only construct our understanding of them and communicate that. Why, then, would narrative discourse be so different? It is only special in one respect – that it excludes the possibilities wherein the ‘narrator’ was in communication with Patsy, but otherwise it leaves the door open to options favoring ‘construal’ rather than ‘reporting.’ Thus narrative discourse inherits some of the constructional meanings available in spoken discourse, but it also adds the possibility that the narrator had access to Patsy’s mind without talking to her. This, in turn, is a viewpoint phenomenon, which can only be defined in terms of viewpoint configurations discussed in Chapter 3. Moreover, the idea of narrative access to minds calls for an explanation going beyond the usual treatment of STR (see Chapter 7 for more discussion). Before I move on to the sections detailing specific instances of microviewpoint besides STR, I will look briefly at the viewpoint potential afforded by seemingly straightforward representation of speech of characters. Example (1) illustrates some of the issues raised above. In the fragment, the Ego-narrator of Margaret Atwood’s The Blind Assassin, Iris, accounts for the information she received about the death of her sister, whose car went off the bridge: (1) I was informed of the accident by a policeman: the car was mine, and they’d traced the license. His tone was respectful: no doubt he recognized Richard’s name. He said the tires may have caught on streetcar track or the brakes may have failed, but he also felt bound to inform me that two witnesses – a retired lawyer and a bank teller, dependable people – had claimed to have seen the whole thing. They’d said Laura had turned the car sharply and deliberately, and had plunged off the bridge with no more fuss than stepping off a curb. They’d noticed her hands on the wheel because of the white gloves she’d been wearing. (The Blind Assassin, p. 3)
90
Viewpoint: representation and compression
Ostensibly, the fragment is an elaborate report of various stretches of speech and thought. There are thoughts of the narrator, as she guesses how the police got to her and why the policeman sounded respectful, there are the policeman’s words describing his interpretation of events and the information he had gathered from the witnesses, and there are the eyewitness reports which include descriptions of what happened, and in what manner. At this level, the fragment is saturated with STR constructions, with words and thoughts attributed to four different individuals. The fragment appears on the first page of the novel, and introduces some of the major narrative spaces and viewpoints to be elaborated. There are several levels determining how the fragment shapes the reading of the remaining parts of the text. First, there are several participants mentioned, but only three are given salient narrative roles: Iris (the Ego-narrator), Laura, her sister, whose death begins the story, and Richard (as we learn later, he was Iris’s husband). The actual speakers – the policeman and two witnesses – are not referred to by name and are never mentioned again. They are, in a sense, sources of information about the circumstances of Laura’s death, rather than characters proper. Primarily, their reports are used in establishing two possible interpretations of what happened – as an accident or as suicide (see Figure 4.1). Jointly, they construct the crucial point of the entire narrative, rather than having a specific role as characters whose speech and thought need to be represented. Second, the reported utterances, independently of their speakers, give different accounts of what happened. The policeman, who knows that Iris’s husband is a prominent businessman and politician, who wouldn’t appreciate an ugly case of suicide in his family, helpfully provides a possible explanation of Laura’s death as an accident (he reports it to Iris in this way). But he sounds speculative (may have caught, may have failed), and admits that the witnesses told a different story. The witnesses are very clear, report what they ‘saw’ and ‘noticed’, and use words like deliberately – clearly describing a suicide. Between the policeman’s hesitant epistemic stance and the witnesses’ strongly positive epistemic stance (supported by visually, and thus directly, acquired evidence), it is clear that suicide is the more likely explanation, but there may be reasons why it won’t be described that way. In effect, the reported words play no role other than the construction of two takes on the facts – suicide versus accident – and frame the rest of the novel as an attempt to explain both the suicide and the efforts to hide it. As such examples suggest, reporting utterances in narrative discourse may have goals pertaining to the overall structure of the story, and are not necessarily local attempts to render the ‘scene’ (RimmonKenan 1983). Contrary to the narrative form chosen, the reported discourse fragments here count as a kind of proleptic ‘summary,’ or, as I would prefer to call it, as narrative space builders. In order to consider them in this way, though, we have to assume that viewpoint spreads (in Fauconnier’s 1994 [1985], 1997
4.1 Viewpoint and representation SV-space
MN-space
Narrator: Iris present Narrator: Iris past tense
NS1 Laura dies DS 4 Clipping vpoint: accident
91
vpoint: accident
vpoint: suicide
vpoint: accident
vpoint: suicide
vpoint: accident
vpoint: suicide
DS 1 S: policeman H: Iris DS 2 S: policeman H: Iris vpoint: accident
DS 3 S: witnesses H: policeman vpoint: suicide
Figure 4.1 Reported voices and viewpoint (example 1); S = Speaker, H = Hearer, vpoint = viewpoint
sense) through compression across narrative levels, from the lower to higher levels of the story, and thus participates in story construction. This understanding of viewpoint will be explored throughout the chapter. The issue of STR and the various constructions involved will be taken up in Chapter 7. In what follows, I will attempt to clarify the conceptual underpinnings of micro-level viewpoint, starting with examples involving visual perspective and modes of visual representation, and leading towards understanding micro-level viewpoint as a representation of experiential phenomena and of conceptualization. I will then return briefly to the concept of focalization. 4.1
Viewpoint and representation
While the general tendency in narratology is to treat Genette’s ‘seeing’ metaphorically and move away from the visual understanding of viewpoint,4 I will start out here by showing that even literally visual viewpoint has specific discourse functions pertaining to narrative viewpoint in general, and helps in explaining viewpoint phenomena on the micro level of the narrative. The potential of visual viewpoint and visual representation is well illustrated by the extended fragment, from the first pages of Jonathan Raban’s travel narrative Old Glory, to be discussed here. The fragment, titled The River, shows a series
92
Viewpoint: representation and compression
of micro-level viewpoint shifts and compressions which use the narrative as a way of constructing a broad representation of a concept. Thus, what seems to be a non-narrative series of visual representations is a story of the emergence of a concept which is crucial in framing the entire text. The book is a travel story in which the narrator/traveler sails down the Mississippi in a small boat. It starts in the present tense, as a description: (2) It is as big and depthless as the sky itself. You can see the curve of the earth on its surface as it stretches away for miles to the far shore. Sunset has turned the water to the color of unripe peaches. There’s no wind. (Old Glory, p. 11)
The description suggests an observer taking in the view, even though it is phrased with the viewer off-stage, only as a conceptualizer. In the next sentences, the description becomes less literal: (3) A fish jumps. The river shatters for a moment, then glazes over. The forest which rims it is a long, loping smudge of charcoal. You could make it by running your thumb along the top edge of the water. (Old Glory, p. 11)
In the text, the forest is not compared to a “smudge of charcoal,” it is one. And in the very next sentence it is suggested that the description is not that of an actual view, but of a pictorial representation being created by someone. In fact, it is impossible to decide what the intended interpretation is, as the representation in the observer’s mind and the representation in a picture are blended, and thus the viewpoints are also blended. The compressed ‘representation/reality’ viewpoint introduces the story as one of both real travel and of narrativized aesthetic experience. However, the next paragraph changes the viewpoint (and the status of the “river”) once again: (4) It is called the Mississippi, but it is more an imaginary river than a real one. (Old Glory, p. 11)
“It” refers to the blended construct introduced in the first two paragraphs, but it also acknowledges its being mostly “imaginary” – so that the ‘objective’ existence of either the river or the painting is questioned. The next paragraphs shift into the past tense and profile an on-stage ‘I’ narrator, but the story is still not about events the narrator participated in, but about the gradual emergence of the concept of ‘the Mississippi’ in the conceptualizer’s mind. In the following paragraphs the reader is taken through various pictorial representations of the river which gradually shaped the traveler’s conceptualization and his decision to sail the length of the entire river and finally experience it beyond his imagination. First, there was the book he read as a child (I had first read Huckleberry Finn when I was seven. The picture on its cover, crudely drawn and coloured, supplied me with the raw material for an exquisite and recurrent daydream.), then a map (I found the Mississippi in the family atlas . . . I looked at the
4.1 Viewpoint and representation
93
Mississippi, wriggling down the middle of the page . . .), which both support childhood imaginary games played on the shore of a local stream in the traveler’s native Norfolk (. . . if I concentrated really hard, I could see the Mississippi there. First I had to think it twice as wide, then multiply by two, then again . . .). Further details are added as years go by – there was a picture in the National Geographic (very disappointing in its literalness), then a painting by George Caleb Bingham (It showed the Missouri, not the Mississippi, but I recognized it immediately as my river.), all of which add salience to the conceptualization which has started in a child’s mind. With one paragraph of text devoted to each of the subsequently available representations and its relevance to the childhood dream which remains the point of reference, the text is effectively constructing the representation in the writer’s mind, which is marked with a blended, and yet very specific viewpoint. The initial image of the calm river filling the entire field of vision is the image which has emerged out of years of transforming actually encountered representations into the private one. The whole fragment is an example of viewpoint construction through a series of specific representation spaces. The cumulative result, though, is that the represented object, the actual Mississippi river, never enters the stage, as the cross-space connections link various representations in paintings, atlases, photographs, and childhood dreams to the concept in the writer’s mind. The rest of the book is devoted to the real river, but it is to be viewed through the lens of the concept narratively constructed in the first paragraphs. The first sentence of the ensuing travel story is Now, thirty years later, the river was just a hundred miles ahead. Interestingly, the now marks a viewpoint shift from the story of the imaginary river, to the story of the trip, which is just about to start. The sentence, even though it is not part of the “imaginary” narrative, also completes the compression of all the representational ‘encounters’ mentioned earlier into the river, the concept originating in childhood, thirty years earlier. Even though reading the National Geographic or viewing Bingham’s art had to happen in adulthood, the time compression requires that its focal point remains in the childhood years in the English countryside. “The river” is the one which he had dreamed about, and the same one he will travel on. Here is, then, a case where what could presumably be talked about as reported perception, or aesthetic viewpoint, is an elaborate exercise in constructing viewpoint through blending and compression of a number of representation spaces into a narrative space which can naturally be viewed as the SV-space. When the MN-space of the trip is then introduced, the narrator there (very much on-stage and interestingly sarcastic and opinionated) is not only a traveler reporting the events, he is also the ‘owner’ of the dream concept of the river which propels him forward. The fact that the initial description is given in the present tense (even though the childhood fantasies had to be past, and even
94
Viewpoint: representation and compression
though the travel narrative itself is written in the past) further supports the suggestion that the initial image of the river is the SV-space of the entire narrative – it is the generic state of mind guiding the narrator. The ensuing paragraphs in the past tense construct the history of this narrative vantage point. In the fragments discussed above the question of perception and representation is put in an interesting context. On the one hand, the images described are presented from a visual perspective – looking at a picture on a book cover, a painting, or a little stream. On the other hand, as a result of all the compressions, the viewpoint is not strictly visual and would better be described in terms of a mental representation. As a result, these narrative renderings of pictorial representations rely on a triple, rather than double, configuration of mental spaces. First, any pictorial representation such as a photograph or a painting evokes two mental spaces – the represented space (the assumed reality portrayed) and the representation space (the visual form it is given, which includes selection of the field of vision, choice of focus, or color, the literalness level, etc.). This distinction is well established in cognitive literature (Jackendoff 1975; Fauconnier 1994 [1985], 1997), through the discussion of examples like In the picture, Mary is a strawberry blonde, not a redhead, where the phrase in the picture is a builder of a representation space, while the description of Mary as a redhead comes from the represented reality space. The aspect of this set-up which needs to be stressed here is the fact that the two spaces are two inputs to a representation blend, which then allows one to talk about either input or the whole configuration. Thus, as long as the picture is activated in the conversation, one can talk about Mary as a blonde or a redhead, and both mental spaces will be used in arriving at the interpretation. The ‘river’ example does much more than that. It adds a third space – the viewer’s mental representation, which arises out of gradually blending the representations described. Narrative viewpoint here is coaligned with this viewpoint space, even in the initial paragraphs where there are no signals of a specific narrator or conceptualizer. Here, then, is another instance of narrative viewpoint first defined in terms of the narrative space being set up or elaborated, and only then attributed to a narrator. Focusing on a representation blend (with its two inputs, even if only one is described) is different from narrating a representation of some narrative reality in a character’s mind. In these cases it is possible for the narrative to render the mental image only, even if it seems incongruent, and let the reader guess what the reality is. In (5), from Paul Theroux’s novel The Picture Palace, the on-stage Ego-narrator, a photographer, describes her experience with her first camera: (5) It was a summer afternoon in 1917. My father hung upside down in the little lozenge of glass; my mother’s chair was stuck in a canopy of flowers where my beautiful brother Orlando’s toes were planted, . . . (The Picture Palace, p. 45)
4.1 Viewpoint and representation
95
In early cameras the view available through the lens was always upside down, which is what (5) describes. The frame of ‘picture-taking’ is evoked, and the reader knows that when the shutter clicks the represented family scene will be reproduced in a photograph, which will not be upside down. Thus what is described in the fragment is the image recorded on the photographer’s retina, but the image makes sense only in relation to the representation blend prompted. The image is aligned with the narrative viewpoint recording the sense of disconnect between the narrator and her family, which is elaborated later in the story. The effect can also be achieved when the narrative acknowledges the possibility of a common optical illusion. When one is sitting on a moving train, one can perceive the world outside the window as moving, and the train as stationary. In (6), the narrator is sailing along the shore, and represents the view from the boat in a similar way: (6) Yet I could see the land creeping past the wheelhouse window. I shut my left eye and squinted, lining up a coppice of dead elm trees against the steel rigging of the mizzen shrouds. The trees were making definite but slow progress while the boat stayed still. (Coasting, p. 92)
As in the cases above, the writer ‘constructs’ his field of vision and allows the illusion to take over, while knowing that the reality matches the undistorted representation whereby the boat, not the land, is in motion. There is also a category of narrative constructions building on representation blends prompted by mirrors. Mirrors are a source of a variety of blends, where the reflection takes on a life of its own and is in some way decoupled from the object reflected. The examples below are just a tip of the ‘mirror’ iceberg in literature and art, but what seems to be common across the examples is that the reflection is treated as meaningful in its own right, and not merely in its relation to the original. In (7), a woman is dressing up as a man to be able to participate in a men-only event, while in (8) the narrator is trying on a new suit, to fit better into the gentlemanly style of the town of Guntersville where he temporarily resides: (7) She gave me a man’s jacket, a pair of striped trousers, a derby hat. I put them on and looked in the mirror. I was a man. (The Picture Palace, p. 97) (8) “Hey, John, you look real good!” sang out a passing salesclerk as I stood in front of a tarnished mirror, inspecting Rayburn in his new Guntersville uniform. (Hunting Mister Heartbreak, p. 196)
When the woman in (7) says I was a man, she is not deluded into thinking she has changed sex, but establishes an independent point of view whereby anyone looking at her in disguise (not at her reflection) would have to believe her to be a man. The power of the disguise allows her to decompress her sense of self from her outer appearance and view the latter – as she perceives it in the
96
Viewpoint: representation and compression
reflection – from the perspective of someone who has no access to her selfhood (any stranger at the event that night). This also means that the woman in her disguise IS a man from the viewpoint of anyone but herself. Example (8) is somewhat similar. The man in front of the mirror is the author, the travel writer Jonathan Raban, but his name has been charmingly revised by the inhabitants of Guntersville into “John Rayburn,” apparently fitting better into their expectations. The reflection in the mirror does not match the writer’s sense of self, but he can now see himself as Rayburn – someone meeting the expectations of the people in the town. The next two examples are again striking in how similar they are, despite coming from different texts (Raban’s novel Foreign Land and Carol Shields’s Larry’s Party): (9)
Putting the phone down, he noticed his face reflected in the dark uncurtained window . . . What was upsetting was that, at first glance, it wasn’t his own face. The hair and the beard were his, but not the plummeting cheekbones, the sunken eye sockets . . . They were his father’s. (Foreign Land, p. 215)
(10) His father’s solid, ruddy presence . . . He met it each morning in the shaving mirror of the various modest hotels where they stayed . . . He’d turn his eyes slowly toward the mirror, creeping up on his face, and there the old man would be. (Larry’s Party, p. 22)
In both cases, the characters looking into the mirror are men whose relationships with their fathers are strained. However, the crucial part is not that they see themselves as so different as to be unable to understand their fathers, but that they see themselves as worryingly similar. Their own reflections in the mirror are no longer simply representations of their own faces, suggesting a unique identity; rather, their facial features are twisted to resemble those of their fathers. While inheriting facial features is considered normal, these examples propose an uncanny loss of identity: a son is looking into a mirror, and sees someone else – his father. The viewpoint thus represented is not visual at all, but relies on a common assumption which sees physical appearance as a token of inner character. Here, however, it is reversed – the inner self’s similarity to that of another person (the father) is narrated through the perceived (though possibly nonexistent) similarity in physical appearance. For this reversal to work, the narrative has to establish a representation blend of the father and his reflection in the mirror and cross-map it with the analogous blend of the son and his reflection. With respect to viewpoint management, this narrative solution is another case of relying on representation to manipulate viewpoint. 4.2
Viewpoint compression
Further discussion of the astounding variety and narrative power of representation blends is beyond the scope of this analysis, but the viewpoint effects of the
4.2 Viewpoint compression
97
examples discussed above call for an explanation. In earlier work (Dancygier 2004a, 2004b, 2005a, 2008a, 2008b) I described the blending pattern involved under the term viewpoint compression. As I argued, the multiplicity of viewpoints in narrative discourse is conceptually manageable because of a series of compressions bringing micro-level viewpoint up to the macro level of narrative spaces. Thus various partial and very local viewpoints are interpreted as contributing to or blending with the viewpoint of the narrative space currently being elaborated. Many of the examples above provide an illustration of how such compressions work. For example, the fragment quoted in (5) presents the narrator’s family members as hanging upside down. The description represents the image in the narrator’s immediate field of vision, not the state of affairs in the narrative space (MN-space), but it is narrated as if it were a part of the MN-space, rather than the narrator’s raw perception, not mediated by conceptualization. Typically, there would have also been a mention of the camera, and there is a mention of it in the MN-space indeed, but not directly in connection with the upside-down image (except the “lozenge of glass” – the lens). The lower-level viewpoint of direct perception was thus compressed with the MN-space viewpoint of the scene describing the narrator receiving her first camera as a present. What is more, the compression gives a salient sense of the Ego-narrator’s first glimpse of the intricacies of photography, which could not have been given in such a poignant manner were the illusion to be dispelled. The mechanism of viewpoint compression thus also explains how a strikingly ‘inaccurate,’ or even false description achieves meaningfulness in the narrative. Most of the examples throughout the chapter can naturally be explained through compression, including the STR examples in (1), where highly localized perceptions of witnesses and the policeman are compressed up to the MNspace where the competing viewpoints of the ‘suicide’ and the ‘accident’ are elaborated in the course of a narrative of over 600 pages. In what follows I will illustrate the phenomenon through another visual example. In (11), a fragment of Raban’s novel Foreign Land, the main character, George, is on his boat, reflecting on his past life: (11) He got the charcoal stove going, poured himself half a tumblerful of Chivas Regal and laid out food in tins. Vera watched him from her photograph on the bulkhead. She said: “Oh, George – you eating chickenshit again?” (Foreign Land, p. 298)
The novel is structured as the most typical past-tense/third-person omniscient narrative. There is thus the stable SV-space, while the MN-space covers George’s return to England after many years abroad and his subsequent decision to sail away on his boat. The narrative space in which George prepares his meal is thus within the MN-space. Vera is George’s former friend, who belongs to the story’s past, described through numerous flashbacks, in which she often
98
Viewpoint: representation and compression SV-space Narrator: 0 MN-space NS1 George on the boat present
Photo-blend Representation (Vera) Represented past (Vera)
DS1 Speaker: Vera Hearer: George
NS2 George and Vera past
Figure 4.2 Representation and compression 1 (example 11)
admonishes George for his unhealthy dietary habits. In the fragment quoted in (11), Vera is talked about as if she were in the same space and time in which George is – in the MN-space. But she is not, since he is on the boat alone, and the only actual sign of her presence is a photograph. There are thus several levels of mental-space structure here, all of which are compressed to ostensibly participate in the general MN-space viewpoint. First, there is the SV-space (past tense, third-person narrative) and narrative space NS1 (George alone on the boat) and NS2 (George’s past relationship with Vera). Vera is introduced via a representation blend, consisting of Vera’s photograph on the bulkhead and the represented reality of the past relationship between George and Vera, evoked from the NS2 space of the past relationship. This initial configuration of spaces is shown in Figure 4.2. The representation blend is a part of NS1, which is the space currently being elaborated, while NS2 also has a discourse space in it, where Vera comments on George’s diet. Two connectors make the ensuing compression possible: the connector linking the represented past in the representation blend to the past in NS2, and the crossspace identity connectors profiling Vera’s presence in all the spaces. Crucially to the emerging blend, Vera is presented as ‘watching’ George, so that she has him in her field of vision, instead of him looking at her photograph, which naturally falls within his field of vision. How is this possible? The reader probably solves the puzzle by assuming that the words spoken by Vera are only in George’s mind or are a memory of a past situation when indeed she said something like this. But it is still puzzling to consider why the narrative does not use explicit expressions directing the reader, something like Vera would say so if she could see him or He remembered Vera saying X. The fact that there is a
4.2 Viewpoint compression
99
SV-space Narrator: 0 MN-space NS1 George on the boat present photo on bulkhead
NS3 George’s thought
Vera DS1 Speaker: Vera Hearer: George NS2 George and Vera
Figure 4.3 Representation and compression 2 (example 11)
solution available suggests that there is a standard interpretive procedure which makes finding such solutions possible. I argue that the procedure relies on viewpoint compression. In the specific example, several things are happening. First, the representation blend of the photograph evokes the reality of George’s past relationship with Vera. The evocation, though, emerges in George’s mind upon looking at the photo, not in the narrator’s space. The entire past space (NS2) is now the ‘present’ in George’s thoughts. The part of the space structure where George is looking at the photo is compressed to leave only the content of his perception (Vera, as represented in the photo) in the wording. The perception further prompts the decompression of the blend matching the photo and the reality it represents, still in George’s mind. He mentally decouples the physicality of the photo from the memories it stands for, and the representation blend is compressed in the way which brings the space of the past relationship (and not, say, the quality of the photo) in focus, along with its attendant discourse space. The narrative space of the relationship (elaborated earlier in the novel) is thus brought into the meal-on-the-boat scene within the scope of the MN-space via George’s thought-space. Consequently, the photo (i.e. the representation input) level is compressed out of the construal. Now that the relationship is in this sense ‘brought back,’ Vera can play her part of watching George and commenting on his habits. The entire temporal and narrative distance between the boat scene and the past relationship has been compressed. The effects of the compression, such as Vera’s vocal presence in NS1, are represented in Figure 4.3.
100
Viewpoint: representation and compression
The series of compressions has thus established a number of interpretive possibilities of a fragment like this one. At the MN-level, George is getting his food ready, while glancing at Vera’s photo and remembering her criticism of his dietary ineptitude. At the flashback-to-George’s-past level, evoked as the represented reality of the photo, there is some characterization of Vera, and perhaps a memory of a specific occasion or occasions when she criticized George. At the very basic level, there is the objective reality of the boat, with food and drink on the table, and a photo on the bulkhead. All of these levels of the story structure are available at the same time and open the various ways of interpreting the fragment (as a flashback, as an expression of a sense of loneliness, as the character’s awareness of his own faults, or as an ironic self-comment, et cetera). These interpretations emerge out of the availability of all the narrative spaces outlined here, but are present not as exclusive options, but rather as various viewpoints available from different positions in the narrative space network. 4.3
Decompression for viewpoint
One more aspect of viewpoint construction has been illustrated by some of the examples above, though it was not explicitly discussed – decompression for viewpoint. For example, in examples (7) and (8) the identity of the person looking into the mirror undergoes a crucial change. While any person normally conceptualizes herself or himself as one entity, whose physique, mental ability, style of clothing, et cetera are blended into one unique whole, there are situations when we see various aspects of our identity as independent. The dividing lines may run across different dimensions, as was shown in the discussion of the socalled split-self metaphors, but the most common expressions of split self assume a difference or a conflict between two major aspects of our personality called the ‘subject’ and the ‘self.’ As Lakoff argues (1996), expressions such as I’m not myself today, I have to reward myself, I’m at war with myself represent various ways in which our identity may be compartmentalized. In earlier work (Dancygier 2005a, 2005b), I discussed a number of such examples as instances of decompressions of identity. One’s sense of uniqueness is a result of a highly compressed blend, but it is natural to decompress that whole when need arises, if only to be able to recognize the changes that inevitably occur. Decompression is thus the flip side of compression in that our need to achieve a holistic understanding of complex phenomena has sometimes to give in to the need to appreciate their inner complexity. Examples (7) and (8), however, do more than decompress the concept of identity. While in both cases the person’s sense of identity remains unchanged (the photographer is a woman, the travel writer does not become a Southern gentleman), the reflection in the mirror highlights those aspects of identity which have been changed to adjust to other people’s expectations. The
4.3 Decompression for viewpoint
101
reflections in both cases are thus representative of how others in the narrative view the character. In both cases, then, the decompression is double – on the one hand, the reflection becomes a token of an identity different from the person standing in front of the mirror (decompression of representation blend), and on the other hand, the person’s identity is decompressed into two personas who will play different roles in the narrative. Example (12) is a more complex example of decompression for viewpoint. In the scene, the narrator meets his father after years of feeling distant and resentful towards him. His perception is based on childhood memories of a serious and conservative man, and that is his concept of the father’s identity. Now they meet again and the father has changed into someone much more likeable: (12) There were two men in my father’s chair. One was my contemporary; a cheerful, plainclothes, bearded, radical debunker. I could only see the other, a far older man, if I squinted hard . . . While my father talked I tried and failed to get the two men to coalesce into one person, but they wouldn’t go. (Coasting, p. 173)
The “two men in [his] father’s chair” are two decompressed parts of his identity – the one from his younger years, when he actually seemed older because of his manner and appearance, and the one at the time of the encounter, markedly different and in some way younger. These two decompressed ‘identities’ are there as anchors to two different framings which the writer is trying (unsuccessfully) to reconcile: the perception of his father as a rather unlikeable person and the pleasant man in front of him. The decompression constructed in the narrative serves the need of allocation of different viewpoints to these two irreconcilable frames of the father’s identity. Furthermore, the story here is told exclusively from the writer’s viewpoint, so that the two perceptions are described as physically present in the scene. Here, as in (11), the distance between the construal in the Ego-narrator’s mind and the events in the MNspace has been compressed, so that two conceptualizations are talked about as two story participants, even though there is just one. Similar decompressions are possible in the case of places. In (13), the narrator comes to Cairo from Yemen, and, by comparison, it seems to him to belong to the Western rather than Arabian civilization, while the tourists arriving from England see it as exotic. Cairo (as a concept constructed by a visitor) is thus decompressed into two differently framed locales, while each frame participates in the construction of a different viewpoint (Cairo feeling-like-home/Cairo as exotic). (13) My fellow-diners and I had come at Cairo from different angles, and we’d arrived at different places. They’d flown from Gatwick to the land of the Pharaohs, while I had made a homecoming of sorts from Sana’a. (Arabia, Jonathan Raban, p. 267)
All the examples above decompress concepts into two independent entities based on different frames applied. Furthermore, all these decompressions allow
102
Viewpoint: representation and compression
first-person narrators to represent events from more than one viewpoint. Consequently, these single-viewpoint Ego-narratives are enriched with a multiplicity of perspectives on the events narrated. While these viewpoints could have been represented through reporting and discussion, the lower-level decompressions of identity allow other compressions to take place across narrative space levels – so that, for example, the past and present images of a man can both coexist in the MN-space, even if only in the eye of the narrator. The term ‘decompression for viewpoint’ can thus be further elaborated into ‘decompression for viewpoint compression,’ since the decompressions at the lower narrative level are then subject to viewpoint compressions which allow the Ego-narrator to naturally represent, rather than simply report, different perspectives on the events narrated, at a higher level. Thus when Raban comments on Cairo, he is also narrating from the Ego-perspective, as the comparison of frames he makes could not have been made by the tourists whose viewpoint he contrasts with his own. ‘Decompression for viewpoint compression’ is thus a narrative technique allowing a single Ego-narrator to represent other viewpoints, constructed via decompression, by compressing them with his/her own. 4.4
Fictive vision, causation, and change
It is worth noting that in most of the examples discussed above there is a distinct ‘visual’ flavor to how the viewpoint is constructed. While this may seem a simple consequence of the fact that many of the examples are representation blends, it is not a sufficient explanation. The initial paragraphs of the ‘River’ fragment are a description of a conceptualization primarily through visual means. The image does not include the well-known non-visual framings – the dangerous power of the river, its energy, or its flow. Even though constructed out of other representation blends, this image is so intensely visual as possibly to fool the reader into thinking that someone standing on the shore is describing what he/she sees. The examples with mirrors ([7], [8], [9], and [10]) behave similarly – in all four cases the person looking into the mirror ‘sees’ someone else – not ‘imagines,’ not ‘thinks about,’ but simply sees. Also, when the traveler in (6) writes Yet I could see the land creeping past the wheelhouse window, he is simply describing what he sees and not what seems to be the case. In all of these cases what is reported as visual perception is in fact conceptualization, similarly to the situation in (11), where what is presented as the words said by Vera belongs to one of the many aspects of the narrated state of George’s mind. There seems to be a consistent strategy at play here whereby mental representations and conceptualizations are narrated as perceptions. It is probably most salient in (12), where an actual act of looking at one’s father is narrated as seeing two of them, each one of different age and appearance. Crucially, the ‘viewer’ tries to have
4.4 Fictive vision, causation, and change
103
the two fathers merge into one, but cannot make it happen. In this case, the entire complex conceptualization is narrated as literally seeing two men instead of one. Conceptualizing here is represented as seeing. In fact, the persistence of this narrative construction suggests a usage pattern which I will refer to as fictive vision. The term intends to mimic Talmy’s (1996) idea of fictive motion,5 represented by examples such as The curtain runs all the way to the corner, where mental scanning of the length of the curtain is talked about as if actual motion was taking place (Matlock 2004 talks about it as an instance of cognitive simulation). Fictive vision is similar, in that visual simulation or a visual construal is used to support conceptualization. While further discussion of the role of fictive vision in language conceptualization is beyond the scope of this project, the examples discussed throughout this chapter suggest that the usage is pervasive. There seems to be much in linguistic usage to support this narrative strategy. We often talk about metaphorically ‘seeing’ as ‘understanding’ or ‘knowing’ (as in I see what you mean, The lecture was clear) or even as ‘conceptualizing’ (as in Look at it from my perspective, I don’t see it that way). Given that visual perception is our primary cognitive source of conceptualizing the world around us, relying on expressions of visual perception in talking about mental processes seems natural, beyond the primary metaphors linking seeing to knowing or understanding. These correlations give at least a partial explanation of why narrating thoughts can be done through narrating perception, even though the perception is not literal. The other part of the explanation is probably the central question of contemporary narratives – how are thoughts to be narrated? However, I want to argue further for a distinction between ‘thought,’ as discussed in the context of STR, and experiential conceptualization and/or perception. In most discussion of STR, ‘thought’ appears to apply to various inner processes, expressing anything from propositional thought (He thought she was pretty) to narration of an emotional state (He liked her), and while some classifications are very subtle and well supported with corpus evidence (Leech and Short 1981, Semino and Short 2004, Short 2007, Vandelanotte 2005, 2009), it is still not clear how they are to be treated as complete. At the other end of the spectrum, where the entire varied collection of narrative choices is talked about as falling under the rubric of the ‘stream-of-consciousness’ techniques (Humphrey 1954), what seems to be assumed is that narratives such as the monologue of Molly Bloom give the reader direct access to characters’ thoughts or consciousness. But there is a difference between thoughts and experiential conceptualizations which in the context of the narrative is not trivial at all, and in the cases discussed here conceptualization is at stake, not thought. The point of using mock-perception as a means of narrating conceptualization is not to give the reader access to a character’s thought processes, or to verbalized statements which count as thought-content, but to allow the reader to experience
104
Viewpoint: representation and compression
the narrated reality through the eyes of the narrating Ego. In other words, using visual construals as a means of conceptualization may give the reader the kind of insight which results not necessarily from access to thought processes, but from immediate access to experience. A sentence such as There were two men in my father’s chair cannot naturally be treated as representing thought, but it is an experiential rendering of a sense of disconnect between one’s memory of a man and the reality of the person sitting in a chair opposite. The fictive vision here is a simulation of experiential on-line conceptualization (as opposed to stable categorization). Experiential aspects of textual choices, in combination with visual ones, seem to play a central role in the narrative function of examples such as (14) and (15): (14) We were shrinking . . . With nothing to measure itself against now except the open Atlantic, the ship, so enormous in Liverpool, so lordly on the Irish sea, was dwindling into a dot. (Hunting Mister Heartbreak, p. 32) (15) The cops are huge. They make the room tiny, filling it with black. (AHWOSG, p. 266)
Both fragments are examples of fictive vision, or perception-asconceptualization. In (14), the perception of the ship getting smaller is the result of the change of the background against which it is viewed, such that it seems smaller than it was at the quayside. There are two additional aspects of the construal here – the choice of the pronoun we, and the progressive form. The pronoun typically aligns the speaker with a group which she is now perceived to be a member of. In this case, it is not clear that the speaker includes other people in the we choice, but it seems more likely that we refers to the ship and all her passengers, with the ship being the core of the group. The choice of the progressive also signals the change in progress – the change in perception, in this case. Consequently, this example is not only about perception as conceptualization, but also about the change of perception as change of conceptualization. Example (15) is similar. Whether the policemen are in fact more than average height or weight is not the issue, but the change in the perception of the room (which now seems tiny) is what is being described. Additionally, the change in perception is also conceptualized causally (they make the room tiny), as a result of the appearance of the cops. The perfect combination of visual viewpoint with change and causality can be found in (16). The narrator is driving down south, noticing the gradual change of climate: (16) The trees that had been skeletal and grey the day before were coming into leaf this morning . . . The harder I stepped on the gas, the faster I could make things grow. I made the first magnolia burst suddenly into flower, woke the first snake
4.4 Fictive vision, causation, and change
105
from hibernation . . . At the rate I was going, it would be fall by Tuesday morning. (Hunting Mister Heartbreak, p. 112)
Rather than saying something to the effect of ‘spring has already come to the southern parts of the US,’ the Ego-narrator attempts to render the gradual perceived change. As in (14), the choice of the progressive (were coming, I was going) supports the conceptualization of gradual change. There is also the sense of accelerated passage of time – if the time elapsed since the day before brought trees into leaf, then spring will change into fall in a matter of days. In fact, change is perceptually correlated with the passage of time (it normally takes weeks or months for trees to go into leaf, and here the change occurred overnight, so it must have been due to some fantastic compression of time – weeks into hours, months into days). Finally, there are two constructions adding causality into the conceptualization of this trip. The X-er, the Y-er construction (as in The more, the better, cf. Fillmore et al. 1988) sets up two causally linked scales, where the increase along one scale (more people at the party) causes an increase of value along the other scale (the degree of fun). In (16), making the car go faster also makes nature go through the seasons faster. The resulting conceptualization is really complex. The driver is going south, seeing more and more signals of spring as he moves through space. The faster he goes, the more space he covers, and the more spring-like nature becomes. But from the perspective of the car the movement through space is less in focus, while the changing view along the road is highlighted. The perception of change is thus narrated as if the narrator’s actions causing motion through space were the cause of the change outside. While visual viewpoint is still at the core, the conceptualization adds causality to the picture and re-construes the narrator from an observer to an agent – a more interesting role from the perspective of the narrative. In any case, this mock-perception of causation related to speed of motion does a good job of highlighting the intensity of perceived changes and compressing the narrative time. The combination of visual compression and causation seems to be more likely than the originality of (16) suggests. The sentence which follows the description in (5) is I had stood them on their heads, but nothing dropped out of their pockets. The narrator here also compresses perception with causation, so that the awkward position seen through the lens (a little lozenge of glass) is in fact a consequence of what the narrator does. To conclude, visual structuring of viewpoint may be what strikes us as the most salient aspect of the examples discussed here, but it is more appropriate to consider these cases as instances of experiential viewpoint – which uses visual perception to render experience (of change, agency, lack of coherence between the mental and the real, et cetera). These cases are thus instances of representing conceptualization through the lens of its experiential aspects. As a result, the MN-space events are rendered through compression of lower-level mental and
106
Viewpoint: representation and compression
experiential construals upwards to the MN level. In other words, the ‘event’ of driving down south is narrated through the experience it offers, and not through objectively measurable aspects (distance covered, cities or states passed, time elapsed, et cetera). Viewpoint compression thus seems to be the mechanism which makes such experiential narration possible. Furthermore, such narrative choices seem natural in the context of the discussion of experiential grounding of conceptualization and emotion (see section 1.6). It is not at all surprising that narrative choices would capitalize on the links between experience and higher cognition by appealing to the reader’s experiential abilities, rather than rely on the ‘telling’ technique. It is thus possible to redefine and expand the idea of ‘showing’ by applying it to narrative choices which prompt experiential alignment. 4.5
The micro level, the macro level, and viewpoint compression
In most of the examples considered above decompression and compression serve the needs of aspects of the narrative being elaborated in the MN-space at the time (photographer’s preparation to appear as a man, George’s meal, et cetera). Even though in (11) the narration of the relationship with Vera (scattered throughout the text) is being evoked, it serves the needs of the George-on-the-boat-space, rather than George’s-previous-life-space. There are, however, interesting instances of viewpoint compression which also provide links across different narrative spaces. I will discuss two such examples here. Characteristically, these fragments appear at the beginning of their chapters. The first, example (17), comes from a chapter of Foreign Land which focuses on Diana, George’s new friend, who lives in Harmony Cottage. Earlier in the story (though not in the immediately preceding chapter), George took Diana for a ride on his boat: (17) Harmony Cottage was out at sea. Diana felt the floor roll away under her feet and steadied herself by leaning on a joist as her kitchen tilted and yawed. (Foreign Land, p. 186)
The initial sentence, describing the cottage as being at sea, has to strike the reader as strange and contrary to common experience. Only in the second sentence the verb felt makes it clear that it is Diana’s experiential perception that is being narrated, not the actual facts. What the fragment describes is a common sensation, especially among inexperienced sailors, whereby being on the boat feels stable, while the land feels like it is swaying after one gets off. The fragment thus narrates what Diana feels as she gets off the boat and comes back home, but does so through evoking the sequentially (not textually) preceding narrative and using the space of Diana’s experience as the current narrative viewpoint space. As a result of all the compressions, the continuity of the
4.5 Levels and viewpoint compression
107
narrative is not disrupted, and Diana’s perspective is firmly established as the one on which the chapter will focus. Let me consider one more example to illustrate the point. Example (18), coming from Alex Garland’s novel The Tesseract, opens a new chapter. It is important to note that at this point in the narrative the fact that Doming had died, in his house, is already known to readers, so that the first sentence of (18) has to be understood with respect to a narrative space other than the MN-space. (18) Doming died at the gates of the graveyard. Everywhere else – the house, the road to the church, the church itself – he had been alive. In the same way that nobody is about to leave until they reach the bus depot and see the bus: alive. And characteristically quiet. But at the gates of the graveyard he died, suddenly, and Rosa was overwhelmed by the understanding. (The Tesseract, p. 169)
Only in the last sentence of (18) is the reader told that the fragment talks about Rosa’s (Doming’s daughter) mental state. She ‘knows’ that her father died, but the realization only hits her when the funeral procession reaches the gates of the graveyard – as if he were symbolically crossing from the world of the living into the world of the dead. Consequently, one can read the first sentence as representing Rosa’s state of mind, and the rest of the fragment as the narrator’s elaboration of it. But even the first sentence here is phrased as a narration of a momentary event, a sudden and complete change of state. This is appropriate, since both ‘death’ and ‘realization’ are events of that kind, though one describes a change in bodily state, and the other a change in mental state. In fact, the connection is even more interesting since the change in Rosa’s mental state in a sense completes the change in Doming’s bodily state – not only is he now dead, but he is also dead in the minds of the people who loved him. The fragment poses a question similar to the one discussed in connection with (17) – how is the access to Rosa’s mind rendered in the narrative? It may be a rather unusual example of a free indirect report, though FID would typically require enough distance to see the mental event in its entirety, and some narrative structure wherein Rosa’s thoughts belong. It would not be appropriate to rephrase the sentence as Rosa thought (or even realized) that Doming died at the gates of the graveyard or some such. The point is that the sentence renders the emotional event ‘as it happens,’ without any distance, in progress. The narrative construction here gives the reader unmediated access to Rosa’s emotional viewpoint, without referring to her specific thoughts or words. The question these examples pose is parallel to the fictive vision examples above: the micro-level viewpoint is rendered through immediate access to a character’s bodily or emotional state. The compression goes all the way from the top space to what could have been rendered as she felt as if the house were swaying or she was shocked by the impact of her father’s death, that is, a character’s bodily or emotional experience. The expressions actually used
108
Viewpoint: representation and compression
skip over all those levels of narrative structure, but the reader is expected to rely on all the mediating narrative spaces even if some of them have not yet been set up (such as Rosa’s feelings and Diana’s dizziness). These experiential compressions seem to be in line with the examples above (especially [14] and [15]), but they render third-person conceptualizations, not first-person Ego-narrator’s conceptualizations. The compression is thus more complex, but direct experience remains in focus. Throughout the chapter, I have been discussing instances of viewpoint which rely on visual, experiential, emotional, and other types of conceptualization. While most of these instances could probably be categorized as examples of, broadly speaking, focalization, or, more specifically, represented speech and thought, represented perception, or some variety of narration, any such description would miss important aspects of meaning construction. The examples seem to be more coherently described as instances of narrative viewpoint leading to representation of conceptualizations which, in turn, participate in the construction of the events in the narrative space being elaborated. While the concept of viewpoint may often correlate with visual descriptions, that in itself does not explain their role, since they may participate in representation blends, in the construction of experiential viewpoint, and in the construal of narrative events. Whether we see the source of a conceptualization in visual perception, or attribution of causality, or framing, or representation of speech and thought, the fact remains that the manipulation of viewpoint is primarily the manipulation of narrative spaces, not of the filtering through one fictional consciousness or another. Processes such as decompression for viewpoint or compression to higher viewpoint lead primarily to the construal of the narrative space. Furthermore, viewpoint construction may require that many narrative spaces are at play – decompressed, blended, embedded, or evoked from other parts of the narrative. As a result, the complexity of viewpoint configurations cannot be consistently described either in terms of focalization with respect to a given character or in terms of STR. The review of examples quoted throughout this chapter suggests that rather than attempt to refine classifications of instances in which fictional minds are used as sources of communicated conceptualizations, we might consider mechanisms which allow various conceptualizations to emerge and contribute to the overall richness and coherence of the narrative. Viewpoint decompression-pluscompression seems to be one such mechanism. 4.6
Speech, thought, and multiple levels of representation
As the examples throughout the chapter suggest, viewpoint phenomena range over much more than focalization, and are also more specific, in that they include various types of conceptualizations. At the same time, they also involve
4.6 Speech, thought, and levels of representation
109
interlocking levels of representation. In recent work by Zunshine (2006), metarepresentation was considered a crucial aspect of fictional prose, one which tests our evolved ability to represent other minds to its limits. In fact, Zunshine also argues that narrative discourse can sustain more complex configurations of metarepresentation than spontaneous discourse does, as it is natural in fiction for a narrator to represent the story told by someone, who also represents the words and thoughts of characters, who in turn represent the words and thoughts of other characters. Zunshine’s use of the concept ‘meta-representation’ was criticized, among others, by Boyd (2006). Boyd observes that Zunshine’s framework is a much-impoverished version of either ‘theory of mind’ or evolutionary psychology, and also that the use of the concept of meta-representation in the analysis of texts is in some cases inaccurate and misleading. From the linguistic perspective, Zunshine’s claims seem striking in that they postulate a qualitative difference between the number of levels humans can process in speech and in writing. While it is difficult to prove that there is indeed an imbalance between speech and writing in this respect, the fact remains that the multilayered nature of fiction calls for an explanation. In what follows, I will argue that narrative levels of meta-representation do not cause much difficulty, as they do not need to be maintained as separate layers of representation, but are compressed in a way which makes the lowest space participate in the higher one(s). The example of The Manuscript Found in Saragossa discussed in Chapter 3 is a perfect test case – there are so many levels of meta-representation that the characters who are listeners sometimes try to recapitulate where they are in the web of stories and who is telling what to whom. And yet, maintaining the exact metarepresentation relations is not the primary task in narrative comprehension, and establishing the current viewpoint trumps other processing tasks. When several stories in the Manuscript, told by different people to different addressees, suggest the idea that beautiful women could turn out to be dangerous demons, the reader quickly forgets who told which story to whom, but empathizes with the main character, Alphonse, who keeps on meeting the same women (or devils, both options appear equally possible) and watches him deal with the puzzle. As Herman argues (2006), having additional tellers could also be explained in terms of the explicit evocation of distributed cognition, and the suggestion seems plausible in that compressed viewpoint is more useful to a group than the details of who said what to whom. Oral tradition seems to support this further, since in that context the person of the actual teller is not directly relevant to how the story is received. It thus seems plausible to claim that narrative discourse manipulates numerous viewpoints by definition, but may bring one kind of viewpoint into focus in any given fragment, whether one sentence long or ranging over longer stretches of the text. At a minimum, such a fragment is embedded in the scope of the SV-space, and within the viewpoint of the MN-space, while also reflecting the
110
Viewpoint: representation and compression
viewpoint of the specific narrative space being elaborated. In the course of reading, the reader has to come up with construals of all these viewpoints, and can then add the viewpoint of a specific character to it. Out of context, though, it may not be clear which viewpoint spaces are activated at any given time. In example (19), from Sara Paretsky’s Hard Time, the first sentence is naturally construed as narration, the second one could be the Ego-narrator’s reflection on the injuries sustained, until in the last sentence the comments are explicitly attributed to her doctor, Lotty: (19) The gun had dug a deep bruise into my side when I tumbled from the boxcar. I’d be sore for four or five days, but if I was careful I’d be okay . . . Lotty dispensed that verdict at her clinic Sunday afternoon. (Hard Time, p. 233)
With respect to STR, the passage distinguishes levels of meta-representation – the narrator, V. I., represents the words uttered by Lotty. But under the interpretation including viewpoint compression, the prediction about being sore for about five days is presented from two viewpoints at the same time – what Lotty said after examining V. I. on Sunday, and what V. I. is thinking/saying later, when she is recapitulating the situation. The compression, as in all other cases examined, brings the space which marks the lowest and most local level of viewpoint (Lotty on Sunday) up to the narrative space where the relevant events are being elaborated (V. I. planning her actions in the space where she is conducting her investigation – the MN-space). But because V. I. is also the narrator, the compression goes higher, into the SV-space, where V. I.’s health is a factor in the entire story being told. As a result of all the compressions, the diagnosis communicated at the lowest viewpoint level by one person participates in the thought processes of another person, who is also the narrator at the higher level. This example shows clearly how the concept of viewpoint compression possibly obliterates the need to distinguish clearly between V. I.’s represented thought and Lotty’s represented speech. A somewhat more elaborate example, from Anne Tyler’s The Accidental Tourist, is presented here as (20). The main character, Macon, who writes guidebooks for people who don’t enjoy travel, is preparing to leave home: (20) For his trip to England, he dressed in his most comfortable suit. One suit is plenty, he counseled in his guidebooks, if you take along some travel-size packets of spot remover. (Macon knew every item that came in travel-size packets, from deodorant to shoe polish.) The suit should be a medium gray. When he’d finished packing, he sat on the couch to rest. Or, not to rest, exactly, but to collect himself – like a man taking several deep breaths before diving into a river. The furniture was all straight lines and soothing curves. Dust motes hung in the slant of sunshine. What a peaceful life he led here! If this were any other day he’d be making some instant coffee. (The Accidental Tourist, p. 25)
4.6 Speech, thought, and levels of representation
111
There are three primary spaces at play here: the MN-space of Macon’s life after his son’s death, the narrative space of his difficult period after separation, and the space of the guidebooks he writes (in italics). Interestingly, there is also the characterization of Macon outside of the narrated events, in the SV-space where the narrator-viewpoint resides (in parentheses). The last paragraph of the fragment in (20) is naturally described as a ‘free indirect thought’ account of Macon’s perception and reflection. Even though it is not mediated by any expression suggesting access to Macon’s mind, it is naturally compressed with the ‘leaving home’ part of the narrative as another aspect of it. What is more important, though, is that all the spaces are not felt to be separate and they smoothly blend into one. It is easy to see the first sentence as pure narration, but in fact it may also be an account of Macon’s decision to wear a specific suit. The quote from the guidebooks may be the narrator’s ironic comment, or Macon’s self-reflection on how he follows his own advice, but it has to be evoked as relevant to some aspect of the narrated events. There are thus several lower-level spaces, but there is also one higher-level space, compressed into one narrative viewpoint of Macon’s experience. As a result, there is no natural interpretation here which would clarify the levels of metarepresentation, other than the narrator’s access to all the spaces at play. It appears, then, that STR phenomena such as FID constitute but a part of the viewpoint or meta-representation phenomena displayed in the fragment. The questions do not seem to be restricted to ‘dual voice’ or even dual deixis (proximal deictics here and this in the distal past context),6 because the fragment, especially in its last paragraph, compresses the character’s experiential viewpoint with the narrator’s focus on the impact of all the details of the day. In this view, the sentences describing the furniture and the light represent Macon’s perceptions, as relayed through the narrator’s privileged position, and not simply omniscient narration. I hesitate to talk about such instances as representing a ‘dual voice’; instead, I understand them as cases of viewpoint compression wherein the narrator ‘speaks in lieu of’ the character. The character’s viewpoint is clearly signaled, but compressed upwards, to let the narrator go on telling the story. It should be clear from the discussion above that the concept of viewpoint compression makes somewhat more realistic assumptions about the degree to which a narrative can be viewed as transforming the characters’ words into various forms of discourse (direct, indirect, or free indirect). It is not necessary to assume that Lotty said You’ll be sore for four or five days, but if you’re careful you’ll be okay (rendered as I’d be sore for four or five days, but if I was careful I’d be okay), or, even more important, that Macon thought What a peaceful life I lead here! The expressions in examples (19) and (20) are renderings of a character’s possible thoughts or words, but narrated from the viewpoint of the MN-space or the SV-space (depending on the type of narrator), to best represent the information from the space’s point of view.
112
Viewpoint: representation and compression
The question of meta-representation also comes up in examples like (1) above, where there are indeed different speakers. The witnesses told the policeman what they saw and he is summarizing their testimonies to Iris, who is now narrating them to the reader. But the question of a verbatim repetition or metarepresentation does not really come up, since all represented speakers have different goals in saying what they do. What really matters is only the highest level – Iris’s intended narrative account of what happened to Laura and the highest-level viewpoint which calls for suggesting an accident and a suicide at the same time. Thus the compression again goes all the way up, and the levels of meta-representation do not need to be maintained. To conclude, viewpoint compression is a blending mechanism which attempts to account for the fact that zillions of low-level facts, observations, or thoughts are compressed into more manageable viewpoint spaces and used in the processing of the narrative as a whole. There is no need to mentally keep track of each subtle change of mood or voice throughout the reading, because these specific bits of narrative structure are quickly compressed upwards to participate in the construction of major narrative spaces. 4.7
Narrative thought and intersubjectivity
The discussion of viewpoint proposed in chapters 3 and 4 raises a number of questions with respect to the actual paths of conceptualization required or used in such contexts. At this point in the discussion it might be useful, then, to go back to some ideas formulated in section 1.6, and relocate the concept of viewpoint developed above in the context of recent proposals in cognitive science and philosophy. Section 4.6 also evokes the approach suggested by Zunshine, whereby enjoyment of fiction as well as our meta-representation abilities depend on the module known as the ‘theory of mind.’ I suggested above that the need to maintain specific meta-representation hierarchies seems to be an impediment rather than an aid to narrative comprehension. Quite independently of metarepresentation, though, the application of theory of mind to narrative was intended as an explanation of our ability to engage with narrative characters in ways somewhat similar to our engagement with other people. However, recent research into the abilities summarized under the ‘theory of mind’ caption suggests a picture relying more on embodiment, activation of neuronal patterns, and, first of all, intersubjective engagement. In a recent edited volume on perspectives on intersubjectivity (Zlatev et al. 2008) a strong argument is made against ‘theory of mind’ in the form we know it. The arguments raise a number of issues, ranging from evolutionary implausibility to rejection of the model of the mind which makes such strong
4.7 Narrative thought and intersubjectivity
113
assumptions about inferential processes leading to specific formulation of other people’s beliefs and desires, or, simply speaking, their reasons for acting the way they do. Jointly, the papers in the volume present a variety of perspectives (from philosophy, psychology, cognitive science, and language) leading towards the concept of a ‘Shared Mind’ – the mind relying on embodied and experiential contact with others. Reviewing the entire argument is beyond the scope of this book, but some of the suggestions made support the general theoretical claims advocated here. First, it is proposed (primarily in Zlatev’s contribution, 2008), that the ability to adopt a view of another person emerged even before the modern linguistic ability, so that at the evolutionary level which Zlatev calls ‘triadic mimesis’ a person (a hominid?) could deal with viewpoint complexities represented by an example such as “I expect that you know that I know that X.” This stage is argued to be crucial to the development of intersubjectivity in hominid evolution. While initially such constructs must have depended on joint attention, based primarily on seeing and shared context, they evolved into a ‘shared mind.’ Interestingly for my purposes, the development had to start with what is still the default natural understanding of viewpoint (seeing things from a certain perspective), and then progress from there towards more abstract concepts – just as the embodiment hypothesis predicts. The idea of an intersubjective mind also highlights the embodied aspects of cognition and the immediacy of our responses to others. The aspect of our neural set-up which is now often discussed in the context of empathy or emotional response to the situation of another person is the existence of mirror neurons. While they have been claimed to have a crucial role in language (cf. the Neural Theory of Language described in Gallese and Lakoff 2005), they are clearly highly pertinent to the concept of intersubjectivity. Crucially, as Gallagher and Hutto (2008) point out, mirror neuron activation does not involve the distinction between first-person or third-person experience, does not require a setting up of models or representations, and is entirely natural, so that it cannot be fabricated to ‘pretend’ anything or gain access to another mind. In this light, the model of ToM, as it has been used, seems significantly over-engineered. As the intersubjective model claims further, contextualized action is more important to the ‘shared mind’ concept than the inner mental processes. The theory essentially sees ‘others’ as agents whose actions are framed pragmatically and, consequently, sees us as sentient participants prepared to act in situations including others, situations which necessarily rely on the recognition of (sometimes conflicting) viewpoints. The ‘shared mind’ approach seems to explain the viewpoint phenomena discussed above much better than other approaches available. Although the narratives I have looked at represent viewpoint mediated through linguistic expressions, the mechanism of immediately appreciating the consequences of
114
Viewpoint: representation and compression
what is being described seems essentially similar. If language evolved out of embodied action, and if it still carries the embodied core in its abstract concepts, there is no reason why language should not be a sufficient prompt to the type of processing the ‘shared minds’ assume. Moreover, as Tomasello (2008) points out, language develops along with the cultural artifacts and socially distributed cognitive needs and it increases its formal complexity in response to the increasing complexity of distributed cognitive networks. A more specific linguistic discussion of the ‘constructions of intersubjectivity’ can be found in recent work by Verhagen (2005, 2008). For example, Verhagen discusses forms of represented speech and thought (He said . . ., He thought . . .) as examples of such constructions, since they explicitly evoke conceptualizations by participants other than the speaker and the hearer. Certainly, Verhagen’s idea could naturally be taken further to explain free indirect discourse, but there is perhaps a different way to view the issue of intersubjectivity in narrative discourse. It seems plausible to claim that fictional narratives develop our intersubjective skills to their limits, since they rely on the extended, though not unlikely, construal of the very communicative context from which the narratives emerge. Narrative can be treated as a communicative act of a specific kind (and it has been, among others in Pratt 1977 and Lanser 1981), but it assumes a lot more than the text as a communicative artifact. Its very basic forms (as described in Chapter 3) assume a concept of ‘someone communicating something about someone else from some point of view’ – which is the basic intersubjective setup, but its specific micro-level choices manage all the viewpoints potentially available in the storyworld. The ‘shared mind’ concept assumes that actions, their motivations, and the emotional response to those actions are parts of the same cognitive package, situated in a pragmatically rich context. If that is true, then narratives fit the description. They do provide the context in which actions occur, they situate actions in ways which make reasons for those actions possible to explain, and they do coordinate a multiplicity of viewpoints, of all kinds. And they evoke emotional responses without necessarily attributing beliefs or desires to the characters. What the ‘shared mind’ approach postulates is that ‘understanding reasons’ is the central ability afforded by the intersubjective set-up. As Gallagher and Hutto argue (2008:27): “understanding reasons for action demands more than simply knowing which beliefs and desires have moved a person to act. To understand intentional actions requires contextualizing these, both in terms of cultural norms and the peculiarities of a particular person’s history or values.” The idea was first proposed in Hutto (2007b), as the ‘Narrative Practice Hypothesis,’ which argues that the intersubjective (or ToM) skills develop in the process of understanding actions in context, and not in attempts to get into people’s heads. The hypothesis is further put in the context of research findings (see Decety and Chaminade 2003) which show that understanding of oral
4.7 Narrative thought and intersubjectivity
115
stories is impeded when facial expressions and body language are in emotional discord with what is said. Although Hutto’s work is not intended to explain fictional narratives, the textual equivalents of bodily behavior play an analogous role. Expressions describing a character’s gaze, tone of voice, expressions of represented thought, strategic use of reporting verbs like moan or shout are all signals of reasons why characters act the way they do and why they respond the way they do. Even though no such theory is evoked there, Palmer’s work (2004) describes the many ways in which fictional minds are at the center of narrative acts and in fact proposes an understanding of narrative discourse as constructing access to other minds and viewpoints. The role of imitative and emotional responses in narrative comprehension is further discussed in Currie (2007), who proposes that reading requires adopting a ‘framework’ through which we can identify with the characters’ stance. Such a framework or stance is independent of the comprehension of the plot, and would seem to be a natural extension of the Narrative Practice Hypothesis. Together, these approaches explain how sentences such as Doming died at the gates of the graveyard or Free Indirect Discourse fragments can at all be processed by a reader. They rely on everything else the text makes available – different characters and their stance, the characters’ actions and reasons for those actions, the entire context, with its conflicting viewpoints, but they also rely on our ability to use all these prompts to construct an intersubjective configuration wherein we can perceive characters’ acts and experiences as indicative of the way in which they, in turn, perceive and understand others. There are many textually implied minds at play, from the narrator to the most lowly character, and together they create an intersubjective constellation we can grasp without analyzing each one separately, based on their actions and interactions. One might claim that postulating a line of development from triadic mimesis to narrative fiction is a bit of a stretch, but it is less far-fetched than it seems if one considers the degree to which narratives rely on multiple viewpoints. There have been numerous attempts in the literature to solve the question of viewpoint by multiplying categories or expanding them to cover race, gender, honesty, self-consciousness, et cetera (as in Lanser 1981). But these enhancements only show that what may be properly understood as viewpoint is all that human perception, interaction, culture, and social situatedness can suggest. I have here proposed a different approach, whereby narrative viewpoint is a function of textually and culturally available narrative space topology, including different perceptions of that topology. I will also claim further in the chapters to come that narrative form has continued to develop in the way which makes human minds more and more central to it and has also kept on revising its ways of making such fictional minds accessible. Oral tradition narratives mostly focus on action, and
116
Viewpoint: representation and compression
downplay the teller or the emotional response, while early novels rely heavily on what characters do or say. However, narrative discourse gradually develops forms like FID, stream of consciousness, unreliable narrators, and all the varieties of viewpoint compression, all of which enhance narrative access to minds in action, thus increasing our ‘shared mind’ abilities. Crucially, narrative discourse has developed to be a means of accessing other minds in ways much more sophisticated than spoken discourse. The narrative constructions reviewed in chapters 2 and 3 look like textbook examples of what Tomasello (2008) would predict – more and more complex linguistic forms providing access to other minds in an intersubjective context constructed in the text. Throughout this chapter, I focused on viewpoint-structuring lexical choices; in Chapter 5, I will consider a variety of referring expressions (pronouns, proper names, role-descriptors, common nouns) and, specifically, their contribution to the establishment of viewpoint and identity in the context of narrative discourse.
5
Referential expressions and narrative spaces
Do you want to grow up to be Miss New Jersey just like your mommy? (American Pastoral, Philip Roth)
This chapter presents an overview of selected patterns in which referential expressions are used in the narrative. The analyses presented here cannot exhaust the issues, but they are intended to show that the mental space complexity of narratives creates a number of cross-space links which can then be manipulated or differently construed through the use of any referential form – a pronoun, a name, or a role-descriptor. Reference in narratives is thus a matter of emerging complex referential networks. There are many specific problems in this area of language use. First of all, pronouns in the narrative need a specific treatment with respect to the different nature of deixis in fictional texts.1 In Chapter 3, I have signaled some characteristics of the first-person versus third-person choices which structure novels in their entirety, and in Chapter 7 I will return to the issue of the narrative use of pronouns in speech and thought representation blends. In this chapter, I will consider two other classes of referential expressions: proper names and roledescriptors. Section 5.4 will also propose a brief discussion of how the use of common nouns in constructions such as ‘fictive motion’ (Talmy 1996; Matlock 2004) can profile participants in the narrative space. The chapter ends with an excursion into contemporary poetry, to add to the understanding of deictic pronouns in literary contexts. The crucial way in which my view of reference differs from most accounts is the focus away from anaphoric and cataphoric relations assuming a certain flow of discourse, and towards personal pronouns as signals of connectors between mental/narrative spaces.2 This is similar to Rubba’s (1996) discussion of alternative deictic grounds, where she shows how the choice of a deictic pronoun may align the speaker with different locales, cultures, or viewpoints. But my approach also attempts to reveal the specificity of narrative discourse, which relies heavily on stable and textually maintained connections across different spaces. The question is directly connected to the broader issue of the cognitive status of identity of characters, 117
118
Referential expressions and narrative spaces
whether in fiction or nonfiction, and the relationship between identity and viewpoint. 5.1
Compression, decompression, and cross-space mappings
The concept of a narrative space assumes that such a space is structured with a certain topology, including primarily time, spatial location, and participants. These participants may, but do not have to be profiled also in other narrative spaces (a secondary character may appear for the duration of one scene or chapter only), or they may participate in all the major spaces. If the latter is the case, the continued identity in all the relevant spaces has to be supported through linguistic means. As I argued in Chapter 3, the primary difference lies in the status of the narrator, but identity is also dependent on how a text treats the very conceptualization of a character as constituting an independent and coherent whole. In earlier work I have discussed many instances of the narrative treatment of reference which requires that a fictional identity be decompressed across different narrative spaces, mostly for the purposes of marking differing viewpoints in the spaces.3 For example, when travel writer Jonathan Raban visits the Alabama town of Guntersville, the people he meets view him very differently from how he views himself. They possibly mispronounce his name a bit, so during his stay in Guntersville he assumes a new identity, that of John Rayburn, which is construed from the viewpoint of the townspeople. (1) I liked being John Rayburn. I had fallen into a routine that felt like a good life . . . But I knew that if I stayed much longer, I would make that one mistake. Scratch John Rayburn, and he’d confess my own thoughts on politics, books, religion – thoughts that wouldn’t wash in Guntersville. Late one night . . . I packed Rayburn’s life into my case. (Hunting Mister Heartbreak, p. 218)
The example suggests a specific kind of decompression of a frame. Raban definitely preserves his inner persona, and thus his new identity relies on the way the inhabitants of Guntersville choose to see him. The stay in the town is a narrative space in which the writer and first-person narrator is construed from a different viewpoint, while remaining firmly connected to the narrating persona in the SV-space. Decompression for viewpoint thus does preserve the crossspace links which allow the reader to follow the story and see the idea of ‘John Rayburn.’ At the same time, examples like these (and they are very common) suggest that referential links may be established not only backwards and forwards in the flow of discourse, but also across spaces. Example (1) also uses proper names in a way going beyond the standard definition of ‘unique reference.’ ‘Rayburn’ is much less of a unique identity, and much more of a signal of an application of a new frame to the person. I argued
5.2 Proper names, frame metonymy, and character status
119
elsewhere (Dancygier 2009) that proper names are in fact not ‘unique’ by definition, but that they make unique reference possible by calling up a very rich frame. In fact, it is also common to decompress the frame from the actual person or location and use the proper name to refer to the frame only, as in the case Everest (in the sense of ‘ultimate achievement’, as in Getting that award was his Everest), Vietnam (‘a costly and unsuccessful war,’ as in Is Iraq becoming Bush’s Vietnam?), or Einstein (‘a genius,’ as in So you think you are an Einstein?). In all of these instances the proper name is used with modification, but does not simply become a common noun, as it retains the uniqueness of the framing. At the same time, we can argue that ‘Rayburn’ in (1) is also a role that the real man assumes for the duration of his stay. Role-value mappings (such that the role of President can be filled by ‘values’ (such as Clinton, Bush, Obama) also rely heavily on frames (e.g. the role of President emerges out of the frame of a system of government). In the case of (1) it is not likely that anyone else would fill the role, but the same man then assumes another identity/role (that of ‘Rainbird’) when he arrives in Seattle. In the next two sections, I will consider more examples of the use of proper names and role-descriptors in narrative discourse. 5.2
Proper names, frame metonymy, and the status of a character
Proper names in narrative discourse may have various functions, going beyond the designation of a character as a participant in the story.4 Examples of such narrative functions are numerous and varied, and range from meaningful names such as Squire Allworthy, which do not maintain the pretence of authenticity, to names associated not only with family history, but also with shared features of physique and character, as in the case of the Aurelianos and José Arcadios in Márquez’s novel One Hundred Years of Solitude. Some authors further rely on proper names in explicitly connecting narrative spaces, as in the case of Coetzee’s Slow Man, where two women characters (Mariana and Marianna) are connected not only to each other, but also to the male character’s childhood memory of the French symbol of revolution, Marianne. The recurring name, popular across different languages, gives one main female character, Mariana, a more meaningful role in the story. Names can also signal narrative options and play with the idea of character identity, as in Nabokov’s Lolita, where proper names seem to play a special function. First of all, the narrator, Humbert Humbert, openly admits that his name has been invented, and discusses other options he has considered, as well as reasons why he rejected them. Other characters are named to evoke associations and puns of all kinds (Quilty/guilty, Cue/Q); also, all characters are referred to in multiple ways. To mention only the various names of the title character of Lolita, Dolores, Dolly, Lo, Lola, or Lolita
120
Referential expressions and narrative spaces
are all varieties of the main character’s name aligned with different viewpoints (school, family, sex, marriage, et cetera). Clearly, writers do not always treat proper names as no more than convenient labels to identify characters. Proper names can also be used without any specific reference, and then they can signal associated social and ethnic frames. In Philip Roth’s American Pastoral (to be discussed extensively in the next section) the main character is a Jew, Seymour Levov, also known by a nickname the Swede, whose ambition takes him beyond his Jewish roots. The narrator comments on the name choices Jews have made: (2) The contradiction of Jews who want to fit in and want to stand out . . . resolved itself in the triumphant spectacle of this Swede who was actually only another of our neighborhood Seymours whose forebears had been Solomons and Sauls and who would themselves beget Stephens who would in turn beget Shawns. (American Pastoral, p. 20)
The names here, all used in the plural so as to indicate the reference to generations rather than individuals, all stand for degrees of alignment with the religious Jewish past: from traditional Jewish names (Solomons and Sauls), to the somewhat Americanized generation which follows (Seymours), and the next era, where Jewishness is almost erased (Stephens), all the way to becoming indistinguishable from any other ethnicity (Shawns). The reliance on frames makes it possible to outline many years of social and religious change in a just a few sentences. Example (1) shows how decompression for viewpoint may be marked by different proper name choices. In (3), a similar instance of decompression, the main character in Anne Tyler’s novel Ladder of Years, Delia Grinstead, begins to see herself as two different people. In this case, the decompression is also aligned with location: Delia in Baltimore, where she lives with her family, is not the same person as Miss Grinstead in the town of Bay Borough, where she lives alone, having left her old identity behind. In this case, however, the decompression is not a jocular reference to the quirky preferences of the locals, but a genuine sense of disconnect which Delia recognizes and observes with some astonishment: (3) Miss Grinstead was Delia – the new Delia. (Ladder of Years, p. 94) She had noticed that Miss Grinstead was not a very friendly person. The people involved in her daily routine remained two-dimensional to her . . . . She hadn’t developed the easy, bantering relationships Delia was accustomed to. (Ladder of Years, p. 101)
Crucially, Delia remains the name the character and the narrator rely on in referring to the ‘real’ Delia – to her inner sense of self. Miss Grinstead, on the other hand, is a person who surprises Delia with her behavior and taste.
5.3 Role-value mappings as cross-space connectors
121
My final example of proper names functioning in a text comes from a movie, Murder at 1600: (4) Will Carla Town be Jack Neil’s Chappaquiddick?
Jack Neil is a fictional US president, and Carla Town is the name of a murdered White House employee. The investigation reveals some sexual connection to the president’s family and thus a scandal is possible. The name of ‘Chappaquiddick,’ a little town in Massachussetts, is used here with reference to a tragic accident in which a young woman drowned and the man who had accompanied her had left the site of the accident without notifying anyone. The man was Senator Ted Kennedy, whose political career suffered as a result. Example (4) also represents a construction, discussed in Dancygier (2009), which operates primarily on frames, but also uses the genitive form to prompt a blend in which an experience of an event is used as a frame to then be attributed to the experience of another person. In the blend in (4), all proper names are cross-mapped with other, unmentioned names, and their associated frames. The Jack Neil scenario here is blended with the familiar Ted Kennedy scenario, and the deaths of two women are also connected in the blend. The point of the construction is to use familiar frames to construe new situations, but, crucially to the point being made here, it relies on frames as prompted by proper names. In (4), the names call up frames, emerging from the story (Carla Town=her tragic death; Jack Neil=a politician whose career may suffer), or imported via another input (Chappaquiddick). The blend can then be used in constructing the movie’s story along the lines of the scenario evoked (which turns out to be wrong). Proper names are thus used in the narrative not only to refer, but also to participate in constructions which structure the emergent story itself. In the next section, I will consider another referential dimension – that of rolevalue mappings. 5.3
Role-value mappings as cross-space connectors
Role-value mappings, as briefly mentioned above, connect two entities: a slot in a socially determined frame and an individual filling that slot. However, as narrative data suggests, roles may not have to be defined outside of the text, but may be established by a pattern recurring across different texts (such as ‘rake’, ‘fortune hunter,’ et cetera) or even by a text itself. In what follows, I will discuss the case of a novel which is structured by frames and associated roles – Philip Roth’s American Pastoral. The novel tells the story of Seymour Levov, a New Jersey Jew, and his family. Seymour is introduced as a high-school friend of Nathan Zuckerman, Roth’s literary alter ego, and referred to as the Swede. This high-school nickname is prompted by Seymour’s unusual physique – he is blond and tall, and very
122
Referential expressions and narrative spaces
athletic. The Swede wants to be just an American, not an American Jew, and starts constructing his life around his central goal – having his own piece of the American Dream. And indeed the dreams all come true – the Swede makes a lot of money running his father’s glove factory, moves to an old house in the beautiful village of Old Rimrock, marries an Irish Catholic girl, Dawn Dwyer, and has a beloved daughter, Merry. The dream frame, structured by roles such as ‘money-making husband,’ ‘beautiful wife,’ ‘perfect child,’ and ‘respectable house,’ is complete. In fact, each of the roles in the frame is amply elaborated in the first part of the novel, entitled ‘Paradise Regained.’ However, other frames are at work too. To Seymour’s family his wife is just a shiksa – someone from outside of their own frame of a traditional Jewish family. Dawn is also a former Miss New Jersey, so she is definitely a beauty, but in the Swede’s eyes her primary role is the ‘wife’ and the ‘mother.’ When a local paper asks little Merry ‘Do you want to grow up to be Miss New Jersey just like your mommy?’ the Swede’s dignified answer is ‘My wife isn’t Miss New Jersey – my wife is her mother.’ He rejects the framing whereby Dawn could be seen as ‘just a pretty face’ and places her squarely within the system of family roles which lie at the core of his American Dream. Besides, Dawn is also a working mother, as she runs a cattle-breeding company known as Arcady Breeders. This completes the frame with hard, farmlike work to counterbalance the Miss New Jersey image, as well as the ‘wife of a rich man’ frame. The American Dream has to be hardearned, and even their beloved girl participates: Merry shoveled cowshit from the time she was six. Merry rode tractors. Nobody is allowed to think that the family’s success and their later tragic undoing emerge from a life of thoughtless luxury. The frame would not be complete without the ‘house.’ Driving through a village Seymour spots a house which fits his frame perfectly: (5) He saw a large stone house with black shutters set on a rise back of some trees. A little girl was on a swing suspended from a low branch of one of those big trees. (American Pastoral, p. 189)
When the local paper interviews Dawn (of Arcady Breeders), the article puts all the crucial elements of the frame together: (6) Mrs. Levov, the former Miss New Jersey of 1949, loves living in a 170-year-old home, an environment which she says reflects the values of her family. (American Pastoral, p. 199)
There is a storybook quality to the image Seymour constructs. This is only confirmed through the clearly described image of what he means by ‘a happy American’ – someone like the legendary Johnny Appleseed: (7) It was one of those kid things you keep in your mind no matter how old you get, but whom he felt like out in Old Rimrock was Johnny Appleseed . . . Wasn’t a Jew, wasn’t an Irish Catholic, wasn’t a Protestant Christian – nope, Johnny Appleseed
5.3 Role-value mappings as cross-space connectors
123
was just a happy American. ‘Who is his wife?’ ‘Dawn. Dawn Appleseed.’ ‘Does he have a child?’ ‘Sure he has a child. And you know what her name is?’ ‘What?’ ‘Merry Appleseed!’ (American Pastoral, p. 316)
The real Johnny Appleseed (John Chapman) was a man who walked the frontier selling apple seeds and starting nurseries, but the name naturally calls up a frame of a children’s story familiar to most American parents. The blended proper names given as answers to Merry’s questions make the whole family participate in the maintenance of the elaborate ‘happy American family’ frame Seymour has started. Merry is the part of the dream which is less than perfect from early on and in the end ruins the whole of it. As a child, she is a stutterer. As a teenager, she befriends people with extreme leftist views and is involved in a terrorist attack on the post office in her village of Old Rimrock. The attack kills four people, including a local doctor, and Merry disappears. Merry’s crime and disappearance reveal the cracks in the flawless image of Seymour’s American dream, and so Seymour’s brother, Jerry, talks about the little murderer herself, the monster daughter and describes the Swede ironically as Just a liberal sweetheart of a father. The philosopher-king of ordinary life. The roles of all the family members have shifted. Dawn, for example, realizes that however much she might prefer to imagine herself as another prematurely aging reader of Vogue, she is no more than the mother of the Rimrock Bomber. Their entire life focuses on the shame and guilt they feel, and they now fill roleslots in a family frame they weren’t even aware of: parents of a child who committed a crime against their American dream, parents of the Rimrock Bomber. The narrative form changes gradually to reflect the growing incongruity of the Levovs’ dream and the reality which they still cannot face. Substantial pieces of the text consist in the elaboration of contrasting frames which can never be reconciled into anything coherent. As Merry’s crime destroys the sense of the Levovs’ life, it also forces the story to start reflecting the Swede’s tormented thoughts: (8) Five years pass. In vain the Rimrock Bomber’s father waits . . . A bomb goes off in the Pentagon . . . The bomber leaves a note. “Today we attacked the Pentagon . . . while US mines and warships are used to block the harbours of the Democratic Republic of Vietnam”; The Democratic Republic of Vietnam – if I hear that from her once again, Seymour, I swear, I’ll go out of my mind. It’s their daughter! Merry has bombed the Pentagon. (American Pastoral, p. 147)
The narrative becomes more and more fragmented in this manner, going from a narration of Seymour’s psychological state, through the report of the news, and the prompted memory of Dawn’s words, to the way in which these pieces begin to fit together in Seymour’s mind. The Rimrock Bomber is no longer a role for which Merry, and only Merry, is the value, it is now the name of every terrorist
124
Referential expressions and narrative spaces
in the emerging frame which is later referred to as indigenous American berserk. Seymour is now the father of every youngster out there who has ever used the proper name Democratic Republic of Vietnam – another rich frame, giving legitimacy and democratic status to the communist part of the country of Vietnam which is in the guerrilla war with the US. The cross-frame links now extend from their own household and the memories of Merry’s adolescence to every family affected in this way and to every young terrorist who is missing. Seymour uses all means possible to find Merry. He meets with Rita, a woman who claims to know Merry’s whereabouts, but the encounter opens a new set of roles for them. She mockingly describes Merry as the daughter of the beauty queen and the captain of a football team, sneers at her mother as Lady Dawn of the Manor, while calling Seymour Mr. Law-abiding New Jersey Fucking Citizen and the great Swede Levov, all-American capitalist criminal. These insults, intended to paint Seymour’s family as morally empty and self-righteously locked in their world of privilege, do not have the desired effect. But even though the Swede cannot accept Rita’s framing of his life, he begins to see how it offered social acceptance and respect to his daughter, the stutterer, who became a soldier of freedom, a pioneer in the great struggle against repression. These words frame him as a capitalist criminal, but give Merry the role of which the Rimrock Bomber is only a peripheral fraction. The same attack which is a senseless crime to some is an act of courage and justice to others – how could one be surprised that Merry does not wish to come home to being the Rimrock Bomber? He does find her at last. She has changed her name to Mary Stolz, and is thus not seen as his daughter, but she is no longer the Rimrock Bomber or a soldier of freedom. She has become a Jain – a member of one of the ascetic religions of India. She has come to reject all that drove her away from her family, but she does not quite fit the frame of a peaceful monk devoted to meditation. While devotion to violence has destroyed her family, the devotion to nonviolence has destroyed Merry herself: What he saw sitting before him was not a daughter, a woman, or a girl; what he saw, in a scarecrow’s clothes, stick-skinny as a scarecrow, was the scantiest farmyard emblem of life, a travestied mock-up of a human being. Not even the most generic framings apply to the creature before him, and Seymour has to accept having lost his daughter irrevocably – and losing his American dream by the same token. He sums up his family’s story against the background of their intended all-American, religion-blind ideal – They raised a child who was neither Catholic, nor Jew, who was instead first a stutterer, then a killer, then a Jain. – and thus acknowledges that the American dream frame he has relied on was just an idealization. The consistent use of referring expressions to evoke and juxtapose frames can be seen as the organizing narrative strategy in American Pastoral, in agreement with the novel’s poignant critique of the frames which structure America’s
5.3 Role-value mappings as cross-space connectors
125
social and family life. Crucially, the text’s rich array of role-slots and roledescriptors, not only supports the facts of the story, but also manipulates a number of viewpoints. Each of the characters is described from different angles (the Swede, his brother, the press, Merry’s terrorist friends, et cetera), and as a result an uncomplicated sequential story becomes a multidimensional narrative of changing and conflicting roles each of the characters plays. The multiplicity of viewpoint spaces here is sustained through the unusually broad range of referring expressions, and the changing discourse patterns. The strategy extends beyond proper names and role-descriptors discussed above, and also includes relative and adverbial clauses: (9) The kid in that swing, the kid in that tree. The kid in that tree who was now on the floor of that room. (American Pastoral, p. 326) (10) Merry at the afternoon tea, the band playing, before she’d been raped. She had danced with the headwaiter, his six-year-old child, before she’d killed four people. Mademoiselle Merry. (American Pastoral, p. 404)
In (9), the phrase the kid in that swing is clearly a role-descriptor. It evokes Seymour’s first view of his dream house, with a child in the swing. The role is a part of the American dream frame, which has to include a perfect, happy child. Merry could fit the role for as long as she was small, but, as Seymour recalls the scene, she is now Merry-the-scarecrow in a repulsive room in some dilapidated building. The clause who was now on the floor of that room attempts to describe the same child that played in the swing, but there are no discernible links connecting the two. The technique works similarly in (10), where the ‘before’ image of a pampered girl dancing is juxtaposed with the ‘after’ identity of Merry-the-Rimrock-Bomber. In both examples Merry’s identity is decompressed across the temporal axis, although under normal circumstances the timeline of one’s life has some coherence and directionality to it. Here, there seems to be no path leading from the past into the present. Also the use of distal that in (9) has an interesting effect. Both frames (the happy house and the awful room) are distant from Seymour’s ‘here-andnow’, temporally and conceptually, even though now signals narrative anchoring in the part of the story when Seymour has to accept giving up on his daughter. They both belong elsewhere – in a fantasy or in the corner of one’s consciousness which had best remain closed. The distance implied by that is in stark contrast to the use of this in another scene, one which intends to render Seymour’s raw pain and despair. In (11), he is showing his glove factory to Rita, in an attempt to convince her that he is not a ‘capitalist criminal’ and to get to know her enough to reach Merry. (11) This is called a pique machine, it sews the finest stitch, requires far more skill than other stitches . . . This is called a polishing machine and this is called a stretcher and
126
Referential expressions and narrative spaces you are called honey and I am called Daddy and this is called living and the other is called dying and this is called madness and this is called mourning and this is called hell, pure hell. (American Pastoral, p. 131)
The fragment starts as a rendering of what Seymour says to Rita as they walk around the factory. It has all the formal characteristics of direct speech, including the subsequent appropriate use of you and I. Most important, all the constructions refer to the names of things (this is called X) rather than using simple predicative constructions (this is X). The smooth shift from the ‘factory machinery’ discourse to Seymour’s personal feelings suggests that while the initial clauses can be treated as direct discourse, the rest is a rendering of Seymour’s thoughts, patterned by the same construction. The combined effect is a blend of what Seymour says and what he feels. While ostensibly giving Rita a wellorganized factory tour, he is described as tormented by the memory of his daughter and the pain of her absence, which Rita could cure with one word; the passage thus represents the trauma of a desperate father at the mercy of an indifferent young woman who holds the very sense of his life in her hands. The use of proximal this throughout the fragment blends the truly demonstrative use related to the discourse ground (pointing at objects) and the sense of one’s own inner emotional state – also proximal, and also demonstrative, though in a different way. The crucial switch from one narrative space to another happens after the word honey. At some point in the conversation Seymour starts addressing Rita as honey – she seems so young and her link to Merry frames her as a ‘childhood friend.’ The choice of a form of address such as honey evokes his family role for him, so that her being ‘honey’ puts them both in the same frame and forces him to be construed as ‘Daddy,’ however incongruent it seems in the context of the conversation. The realization further prompts the memory of the horror of the bombing and the painful acknowledgement of his own trauma. The entire fragment is an interesting form of Speech and Thought Representation (STR), given that it effectively blends speech and thought and in fact uses ‘speech’ as a prompt for the inner thoughts which develop alongside it.5 At the same time, there is a clear metalinguistic element to the fragment, as the expression is called X is repeated in most of the sentences in the fragment. The inappropriateness of the ‘guided tour’ framing of the conversation, when in fact they should be talking about Merry, is what prompts Seymour’s sense of disconnect from the interaction. The demolition of Seymour’s dream does not end with the disappearance (and, finally, death) of Merry. His wife has an affair with their neighbor, William Orcutt III, a member of a very prominent legal family. Dawn thus starts her life from scratch, back to the Dawn of Hillside Road, Elizabeth, New Jersey – in an almost forgotten frame of a pretty Catholic girl from New Jersey. Obviously, Orcutt’s social position is an incentive (as the narrator comments sarcastically The mother of Merry Levov needs nothing less than William Orcutt III.), and
5.3 Role-value mappings as cross-space connectors
SV-space
127
Narrator: Zuckerman present
MN-space NS1 Jewish Community The Swede shiksa The monster, daughter NS5 Jain Mary Stolz
NS1 The Levovs Seymour Dawn Merry
NS2 American Dream Johnny Appleseed/businessman Dawn Appleseed/Miss New Jersey Merry Appleseed/the kid on the swing/Mademoiselle Merry
NS3 Media father of Rimrock Bomber mother of Rimrock Bomber the Rimrock Bomber
NS4 Rita capitalist criminal Lady Dawn of the Manor soldier of freedom
NS6 Old Money Dawn of Hillside Road; William Orcutt III (Mr. America)
Figure 5.1 Narrative spaces and roles in American Pastoral
Dawn hopes to erase the Levov episode of her life and pick up where she left off – from being Miss New Jersey. Her hopes are well summarized in the following sentence: (12) Teamed up with Orcutt she’ll be back on the track. Spring Lake, Atlantic City, now Mr. America. (American Pastoral, p. 385)
The chain of events started with the beauty pageants in Spring Lake and Atlantic City will be concluded in the relationship with the most glamorous of Americans. This final nail in the coffin of the American Dream removes any doubt as to the role of hard work and earning one’s wealth that Seymour valued so highly. Earlier, under Orcutt’s influence, Dawn insists on building a new, modern house and abandoning the historic building which Seymour loved so much. The disintegration of the frame he thought he recreated through his life is complete. Other frames have emerged, which he had never predicted, and the competing conceptualizations of his life are too disparate to be reconciled. From the linguistic perspective, American Pastoral is a story of frames and of the roles they rely on. Major narrative spaces of the story are set up with respect to the set of roles they profile, and the plot is organized alongside shifting frames and roles of the main characters. All these spaces contribute to the understanding of the story of the Levov family, which represents shifting social frames, rather than personal stories – as is shown in Figure 5.1. The lives of
128
Referential expressions and narrative spaces
its main characters are presented much less through the description of the events that affect them, and more conspicuously through the roles they take in the changing frames. Almost in the spirit of an allegory, the characters have depth and personality mainly to the degree to which the roles they take on are the true core of the story being told. Thus the way Dawn and Merry are represented is almost skeletal, and only Seymour’s character achieves some depth. The sheer number of role-descriptors, as well as their narrative salience (as in the part where Seymour is referred to almost exclusively as the father of the Rimrock Bomber), is well beyond what most narratives do, and they jointly create an allegorical story of the state of the American Dream. Seymour, however moving a character he is, is not more than the focal point of the frame he builds his life around and of its disintegration. 5.4
Common nouns
To complete this overview of referential expressions and viewpoint, I will consider examples where there are no human participants profiled in the expressions used, but they are implied through the use of common nouns and constructions. In example (13), from Tolkien’s trilogy The Lord of the Rings, several hobbits are travelling through the Shire: (13) The sun had gone down red behind the hills at their backs, and evening was coming on before they came back to the road at the end of the long level over which it had run straight for some miles. At that point it bent left and went down into the lowlands of the Yale making for Stock; but a lane branched right, winding through a wood of ancient oak-trees on its way to Woodhall. “That is the way for us,” said Frodo. (The Fellowship of the Ring, p. 86)
The fragment uses several fictive motion constructions (it had run straight, it . . . went down, et cetera), as well as descriptions of change. The presence of the observers is signaled only through the assumption that the sunset and the road are being taken in by the hobbits. But there are also subtle signals of the travelers’ decision-making process, as in but a lane branched right, which seems to report an observation of someone looking for the right road. The final sentence suggests that the person was Frodo. Such cases would fall under Langacker’s (1990b) idea of a subjective construal, since there is a clear indication of someone observing the situation, but in the narrative context the description becomes in fact a narration of where the travelers went and how what they saw affected their choices. Even when motion is not involved, the construal of the observer may be central to what is being narrated. In J. G. Ballard’s novel Empire of the Sun, a boy separated from his parents in wartime Shanghai roams the streets in search of food and help, taking in the horrors of the situation:
5.5 Personal pronouns, viewpoint, and the narrator
129
(14) Outside the tram station in the Avenue Haig, the hundreds of passengers were briefly silent as they watched a public beheading. The bodies of a man and woman in quilted peasant clothes, perhaps pickpockets or Kuomintang spies, lay by the boarding platform. The Chinese NCOs wiped their boots as the blood ran into the metal grooves of the steel rails. (Empire of the Sun, p. 40)
The description of the scene is structured as narration, but it assumes the visual and experiential point of view of the boy, Jim. At the same time, the choice of descriptive terms can only be attributed to the narrator, and thus the boy’s viewpoint is compressed with the narrator’s SV-space viewpoint, with the latter responsible for the choice of expression. Only two paragraphs later, Jim’s visual viewpoint is acknowledged in the text (Jim watched), but the narrating voice takes over again in the next sentence: (15) Jim watched the coolies and peasant women staring at the headless bodies. Already the press of tram passengers was pushing them aside, submerging this small death. (Empire of the Sun, p. 40)
These examples suggest that viewpoint of specific story participants may be embedded in the richness of the images described. In (14) and (15), the brutality of the scene described is in stark contrast with what one expects a boy to take in without shock, and the calm mood as well as the rich detail of the description create a blended viewpoint which is informative and disturbing at the same time. Without the implied viewpoint of the boy, the description of the scene would seem less brutal and cold. The examples in this section suggest that viewpoint of characters is present in the narration without the intervention of specific constructions representing characters’ thoughts. The effect is often achieved through direct rendering of sensory perception (mostly vision), but the maintenance of third-person narration also supports the blend in which the character and the narrator viewpoints are both represented. 5.5
Personal pronouns, viewpoint, and the narrator
In Chapter 3, I have postulated an approach whereby the folk theory of a ‘narrator’ reflects our reliance on the basic spoken deictic set-up. In this context, the claim of the Deictic Shift Theory that the ‘ground’ of the narrative is textually constructed is true only to a degree.6 I argue that the text does indeed construct all its narrative spaces, but it also requires a higher-level storyviewpoint (SV) space, which structures the text-wide viewpoint of the story. The space may not include information typically expected of the ‘ground,’ such as temporal or spatial location, but it is positioned as the discourse ground from which the main narrative space (MN-space) is viewed, which is partly determined by the choice of tense. In any event, the SV-space profiles a fictional
130
Referential expressions and narrative spaces
narrating subjectivity which may be fleshed out as a character in the MN-space, or remain deictically bound to the SV-space without taking on an embodied shape. These choices determine most of the ways in which the narrative unfolds, but are also crucially tied to pronoun choices.7 For example, the exchange in (16) seems unremarkable until one realizes that the I of I said in the first line is a man, travel writer Jonathan Raban. He is visiting a lifestyle specialist, and uses a fictional persona as a test case. Obviously, Ms Lee can see that the speaker does not fit the description provided, but the pronoun use structures both the situation in the MN-space of Raban’s travel, and the narrative viewpoint. (16) “I have a problem,” I said. “You’ve come to the right place,” said Ms. Lee. “I’ve just won the New York Lottery.” “That’s no problem.” “But I’m forty, I’m a woman, I work in a deli in the Bronx . . . “So you want a total makeover, right?” (Hunting Mister Heartbreak, p. 91)
The I of I said is anchored to the discourse space of the entire narrative – its MN-space (via Raban-traveler) and its SV-space (via Raban-narrator). It is, in as much as narrative discourse allows it, addressed to the idealized reader. The I of I have a problem, however, is anchored to the discourse space of the conversation in Ms. Lee’s office, where it functions as a blend of Raban-traveler’s physique and voice and Raban-client’s assumed persona, including gender. While Ms. Lee is aware that the two are impossible to reconcile, the fact that she responds to the ‘client’ role, rather than the ‘value-for-client’ person, is perfectly acceptable in the context of a consulting office (and would not be acceptable later if Ms. Lee and Raban decided to go for a cup of coffee together). To the reader, both instances of I are cases of Raban-narrator addressing them, either to report what he does in Ms. Lee’s office, or to report what he says to Ms. Lee. This effect is possible because the lowest-level speaker’s viewpoint (that of Raban-client) is compressed all the way up to the highest level of Raban-narrator viewpoint. Thus the first-person pronouns anchored to different deictic grounds in different narrative spaces can coexist side by side and jointly contribute to the same narrative. Examples like (16) may seem unusual owing to the added complication of Raban play-acting as a woman, but any first-person narrative has to distinguish between the viewpoint of the narrator and the viewpoint of the participant, so that any such text relies on viewpoint compression across the SV-space and the MN-space. Also, first-person narratives have to rely heavily on third-person pronouns, so that other participants in the events are talked about too. First-person narratives occasionally switch to the plural form we, which I have briefly discussed in earlier work (Dancygier 2004a; see also Margolin
5.5 Personal pronouns, viewpoint, and the narrator
131
1996). I have argued that the pronoun we does not necessarily include other subjectivities (whether in the inclusive or exclusive sense), but that instead it signals access to a broader frame, anchored to the first-person speaker/writer. This was the case, for example, in example (14) discussed in Chapter 4, We were shrinking, said/thought by an observer on a ship leaving port. While the ‘shrinking’ refers to the observer’s perceived size of the ship, the we does not include the writer in any role other than the observer. There is, as a result, no ‘we’ at all, in the sense of plurality, but there is an evocation of the ship and her passengers, and one person’s point of view on the entire event. We can also be interestingly used in a third-person narrative, to engage the narrator in an unusual way. In Haruki Murakami’s After Dark it reappears throughout the novel, as a token of establishing the visual viewpoint of the narrative space. The second sentence of the novel is: Through the eyes of a highflying night-bird, we take in the scene from mid-air. Then the perspective gradually zooms in on Tokyo’s entertainment district, until the next section starts with We are inside a Denny’s. Then, at the beginning of the next section: (17) The room is dark, but our eyes gradually adjust to the darkness. A woman lies in bed, asleep. A young, beautiful woman: Mari’s sister, Eri. Eri Asai. We know this without having been told so by anyone . . . We allow ourselves to become a single point of view, and we observe her for a time. Perhaps it should be said that we are peeping in on her. Our viewpoint takes the form of a midair camera that can move freely around the room. At the moment, the camera is situated directly above the bed and is focused on her sleeping face. (After Dark, p. 30)
The text then goes on to tell the story, returning briefly to establishing the viewpoint. It is an explicit description of the point of view which the narrator wants the reader to take, and, most important, it establishes the visual perspective only, including the response of the eye to the change of light. The pronoun we creates an illusion that the narrator and the reader are coordinating the way they are viewing the situation, or that the narrator gives a narrative guided tour to the reader. At the same time, the description has a cinematic quality to it (also prompted by the word camera), so that one can naturally imagine seeing what happens in the room. Also the idea of gradual zooming in – from the city, to the district, to the enclosed spaces of restaurants and bedrooms – resembles a technique often used in film, where it usually plays the role of locating the action of the movie in a broader spatial context. In terms of the narrative spaces, the narration seems to flesh out the SV-space viewpoint and at the same time blend it with the expected reader experience of the text. Similarly to the cases discussed in Chapter 4, the focus on the visual viewpoint here is best described as fictive vision (we are invited to ‘see’ the location of the story’s events), which highlights the experiential aspects of how the narrative is delivered. Even though most of the text is not different from a typical
132
Referential expressions and narrative spaces
omniscient narrative, these we-passages make the reader part of the epistemic viewpoint of the story. The examples above suggest that, along with other formal choices, pronouns in the narrative do not only serve the establishment of referential links, but are also viewpoint devices, managing and linking various levels of narrative spaces, and also engaging the SV-space. In general, they support the concept of viewpoint compression, because in the cases analyzed above the primary function of the pronouns is to maintain viewpoint cross-links leading up to the SV-space. 5.5.1
Mixing person and tense
In the analysis of The Blind Assassin in Chapter 2 I looked at an example of a fragmented text, where narrative spaces were marked with different tense and reference choices. However, there are also instances where a continuous and uninterrupted narrative mixes different narrative forms. Such texts, as, for example, Günter Grass’s novel The Tin Drum, do not fit neatly into the types of set-ups discussed in Chapter 3. The fragment in (18), for example, alternates between past and present narration, also switching between first person and third person (Oskar) to refer to the same character/narrator. (18) Oskar couldn’t adjust that easily, however, and looking about for something to take the place of his ants, shifted his attention to several flat, grayish brown creatures strolling along the edge of my Kalmuck’s collar. I wanted to catch one and examine it more closely . . . And to explain my conduct Oskar says . . . (The Tin Drum, p. 375)
Oskar is the main character and narrator of the story, a 22- year-old man who has never grown since he was a child. He tells the stormy story of his family and his own life. The scene in (18) is the scene where Oskar’s father dies, after the Russians seize Danzig. The fragment starts in third-person past, then switches to first-person (my Kalmuck, I wanted) mid-sentence, then moves on to a mix of first and third person (my conduct; Oskar) and present tense (says). These switches do not seem motivated by any specific pattern of meaning construction, other than the attempt throughout to tell the story and also represent the mind of the narrator/Oskar. There is a true decompression signaled through these switches, such that Oskar attempts to be both a teller of a bigger story and a character whose thoughts and feelings (past and present) are being reported, and maintains both roles throughout. The next scene is the father’s funeral, where a wave of reflection rolls through Oskar’s mind, until he begins to consider giving up on his decision not to grow. The question (should I or shouldn’t I?) is repeated a number of times throughout the scene, with intermittent reports of what Oskar is engaged with as he ponders the question:
5.5 Personal pronouns, viewpoint, and the narrator
133
(19) Should I or shouldn’t I? You’re going on twenty-one, Oskar. Should you or shouldn’t you? You’re an orphan. (The Tin Drum, p. 384) (20) Then I found a slender cast-iron cross with crumbling ornaments and crusted letters that spelled Mathilde Kunkel – or Runkel. Then – should I or shouldn’t I – in the sand between thistles and wild oats – should I – I found – or shouldn’t I – three or four rusty, flaking metal wreaths. (The Tin Drum, p. 384)
In the course of this narration-plus-internal-monologue blend, Oskar also addresses himself as you (You’re an orphan). The second person here is part of the internal monologue, wherein Oskar is trying to convince himself that becoming an adult is now inevitable. There is thus another decompression, such that Oskar’s dilemma is the ongoing state of his mind, while he describes other things he sees and does, and also attempts to turn arguments around in his mind. The switches are thus in fact attempts to maintain several narrative viewpoints at the same time, with the viewpoints being structured by various strands of Oskar’s thoughts. The technique is a remarkable attempt to build the multiplicity of mental viewpoints a single character may take. These viewpoints are further blended with various narrative roles in the text (character/narrator), but they depend on the setting up of various thought-spaces. Once again, multiplicity of viewpoint is a function of the narrative spaces set up, rather than of the number of story participants, and formal choices locate these viewpoints in the overall narrative structure. To conclude, personal pronoun choices are further signals of the narrative space set-up in any given text. Such choices often depend on the macro-level narrative space configurations, as described in Chapter 3, but may also reflect the micro-level maintenance of various viewpoints (as in the case of The Tin Drum). The examples discussed above show that it is not possible to determine the type of narration on the basis of the formal choices alone, but that, instead, a careful analysis of narrative spaces might offer a better understanding of how these choices are arrived at.
5.5.2
Pronouns and narratorship
Pronoun choices in the narrative may cause one to reconsider the concept of narratorship, while also questioning the position of the narrator with respect to the fact/fiction contrast. In Richard Powers’s novel Generosity, the first-person narrator is construed as a full subjectivity, while not being cross-mapped with any character. Additionally, the narrating voice is framed as that of a writer, and so the story is told as a third-person narrative, but occasionally the narrator’s voice comments on the writing process and on the decisions regarding the characters’ fate. At the same time, the ‘story’ is given some independence, so
134
Referential expressions and narrative spaces
that the writer may feel restricted in his story-production choices by the current state of the MN-space and the characters’ dispositions. The novel introduces its main character, Russell Stone, in the third person: (21) A man rides backwards in a packed subway car . . . I watch him balance a yellow legal pad on his toppling book sack. (Generosity, p. 3)
The narrator/writer is construed as observing characters rather than inventing them, while remaining invisible himself in their shared time and space. He is not planning his story, but rather lets it be prompted by what is ‘seen’: (22) Look again: the whole point of heading out anywhere tonight. The blank page is patient and meaning can wait. (Generosity, p. 3)
He also makes his character a writer and a teacher of creative writing, and describes him in terms strikingly reminiscent of Powers’s own physique. Furthermore, he also has access to Stone’s thoughts on writing: (23) Just beyond South Bend, Stone has an epiphany. He knows why he could never in his life or anytime thereafter write fiction: he’s crushed under the unbearable burden of a plot . . . Plot is preposterous . . . Story is antilife. (Generosity, p. 273)
It thus seems possible right from the start that there are cross-links connecting the writer, Powers, the narrator of Generosity, and the character/writer, Stone. While tenuous at first, these links are gradually reinforced. In the story, Stone teaches a writing class, which includes a student, Thassa. He becomes close to the girl, as her own story, central to the plot, develops. In the final pages of the novel Thassa attempts to commit suicide, while Stone is with her, trying to help: (24) He sits down on the floor, shaking, clouded and adrift. And in that instant of annihilation, art at last overtakes him, and he writes. He cannot rescind this. (Generosity, pp. 288–9)
Stone thus becomes both a participant in the events of the story, and also the writer who is responsible for these events: (25) He can do nothing for her but revise. And he has time to rework entire world anthologies. In the scene he keeps returning to, all the principals assemble in her hospital room. (Generosity, p. 290)
Through these lines, the text allows the novel’s blend to gel at last. Stone is the writer, and the text he writes creates an irreversible set of facts, ones which constitute the story we are reading. Not only is the character inside the story its ‘outside’ originator at the same time, the story told is also the story he and other characters live. Life and story become one. Thassa goes back to her home country, Algeria, where another character, a filmmaker, visits her. The narrator reports their meeting in a café. Among other things, Thassa takes out her old writing textbook from her bag:
5.5 Personal pronouns, viewpoint, and the narrator
135
(26) “it’s not mine,” she says. “Give it to Russell. He will need this.” I will need much more. Endless, what I’ll need. But I’ll take what is given, and go from there. (Generosity, p. 295)
In this fragment, the narrator’s presence in the midst of the scene is fully acknowledged through the use of the first-person pronoun (I will need much more). At the same time, it becomes clear that the I here is also linked to Russell Stone, who was a ‘he’ until that moment. The cross-space connections become relations of identity rather than just analogy. And because of the initial link to Powers himself, the author is now also a participant in the scene, and, by the same token, in the story. In the final words of the novel, the writer and the character share a moment in silence, and then the ‘story’ disappears, back into the blank page of reality. (27) She looks across at me . . . She answers in all kinds of generous ways. And for a little while, before this small shared joy, too, disappears back into fact, we sit and watch the Atlas go dark. (Generosity, p. 296)
Through the pronominal choices, among others, Powers questions the boundary between fact and fiction, by gradually redefining the relationships between implied author and narrator, as well as a character. He shows how fiction incorporates but also disturbs reality, and he presents the boundaries between fact and fiction as flexible. He also gives characters an ontological status, such that, within the space of the book, they can interact with the author/narrator/ character. At the same time, he gives the author only limited control over characters’ lives, and puts the author at a parallel level of participation. The blend constructed in Generosity seems to rely not only on the presentation of the story, but also on an underlying construction of frames which define books and fiction. The novel seems to treat storytelling as consisting of a blend of two frames: the ‘book’ frame (with its author, its characters, and its readers) and the ordinary discourse frame (with the speaker, the addressee, and participants). In the ‘narrative’ blend, the speaker is integrated with the narrator, addressee with reader, and participant with character, but the author remains backgrounded. In Generosity, the emerging blend is more complex: it blends the author, the narrator, and the character into one, and thus makes the boundaries between fact and fiction permeable. Crucially to the point being made here, pronouns and other referential expressions play a large role in the emergence of the blend constructed in Generosity. The first-person pronoun is used in the text, but initially it represents both the narrator and the implied author. The structure of the SV-space thus profiles two subjectivities, though blended into one. Throughout the text, the I-narrator is not cross-mapped with any character, but the ‘intrusive’ nature of this narration does not guide the reader through the process of story construction. Instead, it seems to follow the author through the process of story-writing.
136
Referential expressions and narrative spaces
The MN-space, while still in the viewpoint scope of the SV-space, is also partly independent of it, in ways which resemble reality, rather than fiction. The final blend of I and he is also prompting the blend of reality and fiction, defined through cross-space participation and epistemic transparency, not identity. Throughout the section, I have considered uses of personal pronouns which represent narrative-space configurations and text-specific construals of the relationship between the SV-space and other narrative spaces. These pronominal uses require that we understand textual deixis as more than the function of the text’s ground, but as a direct representation of the set-up of narrative spaces and identity. Pronouns here serve as crucial signals of narrative space crossmapping, embedding, and compression. They participate in referential blends, rather than simply in reference relations. The fact that deixis may reflect complex relationships between language and the surrounding reality is not new (see Rubba 1996; Hanks 1990). But the narrative treatment of that ‘reality’ poses further questions both about the nature of the ground and the nature of referential relationships. In the final section of this chapter I will return to the issue of deixis once more.
5.6
Deictic I and the construal of subjectivity
Throughout this book, I have been considering the idea of a textually construed ground which profiles a disembodied construct called the narrator. I have also shown how the setting up of a mental/narrative space which includes such a construct affects the cross-space mappings throughout the narrative and the use of pronouns (and tense) in the text. However, I have not looked in detail at the interpretive implications of such a set-up. In order to show what a ‘subjectivity’ like that implies, I will look briefly at two examples from poetry. Even though there are no narrative elements involved, the observations are relevant to my overall argument. Poems often use I (profiled or implied) as a reference to the so-called ‘poem’s persona’ – a subjectivity whose viewpoint structures the content of the poem, very similarly to the way in which the concept of the narrator structures the viewpoint of a novel. It is thus assumed that there is a ‘communicator’ behind the scenes. However, it is possible for a poem to use I to refer to an object, and the resulting interpretation reveals much about the nature of literary discourse. Below, I will consider two poems by Wisława Szymborska (both from the collection Nothing Twice [1997]). In one of them, “Advertisement,” the I stands for a tranquilizer drug, in the other, “Archeology,” for the discipline mentioned in the title. The drug presents itself as capable of handling many problems.
5.6 Deictic I and the construal of subjectivity (28)
137
I can take exams or the witness stand. I mend broken cups with care. All you have to do is take me . . . I know how to handle misfortune, how to take bad news. I can minimize injustice, lighten up God’s absence or pick the widow’s veil that suits your face.
The drug promises to alleviate all kinds of emotional pain and distress. We know that the power of such drugs is to enable us to deal with such situations, but we also do not feel that taking drugs causes us to relinquish our agency. And yet this is exactly the effect that the choice of I brings about: the enablement we expect from drugs is presented as causation and control: the drug, not the widow, would pick the veil. This slight shift within the broad range of causal relations creates an effect whereby the drug taker is no longer fully agentive. What this shift suggests is that the profiling of a speaker (I) also profiles an agent, capable of controlling emotions and performing various acts. The drug should be an instrument (an ‘enabler’), but because it can speak in the first person, it also acquires agentivity and can now cause things to happen. The further result is the construal wherein the addressee (drug taker) is only a patient or an experiencer, but has no control. The point of the poem seems clear, but it is achieved primarily through giving the drug the ability to speak and, consequently, to cause. In the second poem, “Archeology,” the set-up involves epistemic abilities as well: (29)
Well, my poor man, seems we’ve made some progress in my field . . . I no longer require your stone gods, your ruins with legible inscriptions. Show me your whatever and I’ll tell you who you were . . . All I need for my ends is your layer of dirt and the long gone smell of burning.
The use of the first-person pronoun here concerns our epistemic agency. We create academic disciplines in order better to understand the world around us, but we assume that we have control over what kind of knowledge will emerge. Shifting us into the role of the addressee and giving control of the discourse to the discipline construes us as mere recipients of the knowledge we think we are creating. Again, the use of I comes with additional abilities and control, this time in the epistemic domain. The reader, profiled in the position of the addressee is thus given the experience of being dominated by constructs normally considered under his/her control.
138
Referential expressions and narrative spaces
There are several aspects of this construal which call for an explanation. First, the use of I without the support of an available deictic ground, which would be the norm in conversation, in fact sets up such a ground and attributes the construct identified as I as a carrier of all the abilities normally associated with speakerhood. The nature of the construction is essentially the same in spoken and written contexts, but in literary texts requires more of an elaboration of the space functioning as the ground. Primarily, what needs to be profiled is the addressee, typically aligned with the reader’s participation in the exchange. But there are also abilities such as agency and epistemic stance which are associated with the role of the speaker. The relationship between the reader and the literary I is typically based on having an equal status (as in the intrusive narrator case); also, the narrator may yield all the ground to Ego-viewpoints and remain off-stage. But when the first-personhood is attributed to an object, it becomes clear to what degree that actual literary ground is dependent on the specific construal created by the text. Thus it is not that I can be used because some speaker is available in the ground – on the contrary, the ground becomes available, and is then structured accordingly, because I has been used or implied. There is also the problem of making non-human participants speak. The possibility emerges thanks to the frames evoked by these constructs and the subsequent blends in which semantic roles such as ‘agent’ or ‘experiencer’ could be attributed to unusual participants. But the frame of, say, drug-taking already evokes the enabling function, the agent (drug taker) and the instrument (the drug); at the same time, it profiles the drug taker as a patient or experiencer, so that the role is available right away. What the blend does is use the frame evoked by the object (drug), put the object in a context where it obtains the status of speakerhood, and assign roles differently to participants. But the meaning emerges primarily out of the availability of the frame. The kinds of blends described here are not the only ones in which frames evoked by objects participate in new construals involving a speaker role in a newly set-up discourse ground. In the next chapter I will look at blends in which objects are construed as addressees – with a very different effect. I will start with a final look at the concept of literary deixis and then consider dramatic discourse as yet another example of a discourse set-up specific to literary texts.
6
Fictional minds and embodiment in drama and fiction
. . . O happy dagger! This is thy sheath; there rust and let me die. (Romeo and Juliet, Shakespeare)
In earlier chapters I have focused on the contribution the language of narratives makes to various aspects of story construction. I have relied on cognitive theories of language to reveal the general mechanisms involved, and I tried to link the discussion to broader issues of embodiment, framing, and intersubjectivity. In this chapter I will consider further aspects of embodiment, as represented in the narratives intended for the stage. Specifically, I will look at the linguistic means employed in Early Modern theatre to represent the characters’ minds, and trace some of the subsequent changes in literary style. Overall, I will argue that literature develops its formal means in order to find more accurate ways of representing thoughts and emotions. 6.1
Deictic ground in literary discourse
Cognitive grammar and mental spaces theory analysts repeatedly stress the need to include a more refined discussion of the ‘ground’ and the deictic set-up within it.1 As the discussion in preceding chapters suggests, literary texts pose further challenges to what we understand deictic ground to be. Generally, the ground is claimed to include the presence of the speaker and the hearer, as well as the location of the exchange, its history and goal, and the specific communicative goals of the participants. Recent work (Sanders et al. 2009) also expands the standard set-up into the ‘Basic Communicative Space Network,’ which assumes not only the presence of the ground, but also epistemic, speech-act, and metalinguistic spaces, ready to be filled with contextually available information on the speaker’s and the hearer’s attitudes and communicative goals. Generally, the concept of the ground is intended to be more specific than what is typically referred to by ‘the context,’ since it attempts to reveal the specific ways in which the structure of the ground affects interpretation. Furthermore, 139
140
Fictional minds and embodiment in drama and fiction
Rubba (1996) used mental spaces theory to expand the idea of deictic ground and show the effects of speakers switching across different alternative grounds and adjusting their pronominal choices to the relation between the base space and the alternative space profiled in the construal. Crucially, the ground is used by Verhagen (2005) to clarify the ways in which grammatical constructions (which he terms ‘constructions of intersubjectivity’) serve the goal of structuring not just the content of the interaction, but also the way in which it addresses the specific communicative relationships between the speaker and the hearer. For one, what the speaker says may involve other, off-ground participants (as in He thinks you’re right), or comment in some way on what the hearer has said earlier in the exchange. But the primary conclusion from the whole range of mental spaces work on deixis and the ground makes it clear that the standard default set-up (speaker and hearer, here, now) may yield a number of construals depending on shifts of perspective or considering multiple mental spaces at the same time. Deixis seems to refer to written narratives only partially. Deixis in narrative is typically talked about as a text-internal phenomenon: for example, Deictic Shift Theory (Duchan et al. 1995; see also Rauh 1981) explains changes of person and tense through text-internal concepts. Otherwise, there seem to be good reasons for not applying deixis in the context of narration. There is no identifiable ground which the reader shares with the narrator, also the time of reading does not coincide with the time of writing (the writer may in fact be long dead before the reader accesses the text), et cetera. Nevertheless, the lack of an actual shared ground does not prevent us from conceptualizing reading as a communicative act in which the text plays a role. Depending on the inclination of the analyst, the reader may be claimed to be responding to an author or a text, while the reader’s own participation in the act of reading may be seen as central or tangential. However, nobody is likely to deny the existence of some kind of ground, such that the reader is prompted into some conceptual or emotional response by the text, as produced by the author. The communication is one-way, but communication it is. In Chapter 3, I have reviewed some of the most common narrative options, focusing on the construction of the story-viewpoint space. One of the major points made there was that the SV-space sets up the topology which is projected over all the narrative spaces of the text and structures the overall narrative viewpoint established by the text. The set-up of the SV-space may be understood to represent the text’s ground – with the narrator as a kind of speaker (regardless of whether the narrator is on- or off-stage) and the reader as a kind of hearer. The more intrusive forms of narration, where the reader (or narratee) is directly addressed by the narrator, seem to suggest that such a conceptualization may in fact underlie more experimental contemporary set-ups. Furthermore, just as the hearer may become engaged with participants not present in the
6.2 Mental spaces in dramatic narratives
141
ground, who are brought into the discourse through intersubjective construals such as reported speech and thought, the reader is also given access to story participants, via the narrative choices made by the narrator. In the context of such a default set-up we can draw some parallels between spoken discourse and narrative discourse. The fragments typically described as narration are thus equivalent to the narrator’s ‘direct speech,’ while represented speech and thought, in its amazing variety, plays a role similar to intersubjective constructions in speech. It is not within the scope of this book to speculate whether colloquial discourse has fewer such constructions than narrative discourse, or, on the contrary, more of them. But the pervasiveness of intersubjective construals in speech, as described by Verhagen, may suggest that various ways of bringing other participants into the ground are common both in speech and in writing. The question which remains open is the exact role of so-called direct speech (or dialogue) in narrative fiction. It is natural to assume that giving unmediated voice to characters constitutes some kind of stepping back on the part of the narrator, a temporary yielding of the story viewpoint to a character. But there are reasons not to accept such an interpretation, since it would mean that characters’ discourse is beyond the scope of the SV-space, and, more important, that narrative fiction changes viewpoint levels all the time, without any mechanism ensuring the coherence of the whole. We need an explanation which allows us to see how individual instances of direct speech are incorporated into higher narrative levels, including the MN-space. Such an explanation in general terms seems to be offered by the concept of viewpoint compression, as elaborated throughout this book. But in Chapter 7 I will also look at a number of cases of direct speech, to suggest that in the narrative it is in fact representative of a range of various viewpoint phenomena. In this chapter, though, I want to look at some selected aspects of a narrative form which, at face value, consists entirely of direct discourse – drama. I will discuss the ways in which material aspects of theatre are coaligned with its language,2 and the drama-specific ways of representing fictional minds. 6.2
Mental spaces, physical spaces, and dramatic narratives
Even though one of the goals of a play, like other fictional narratives, is to tell a story, it is also clear that the means through which the story emerges are different. Theatre is thus a good experimental ground for testing the assumptions proposed in earlier chapters. I will be interested in one major question: how does a play tell a story? Given the obvious materiality of the storytelling environment, one can expect the verbal layer of the medium to play a role which complements the physical materiality of the stage and the embodied presence of the characters. It is also to be expected that the representation of
142
Fictional minds and embodiment in drama and fiction
the characters’ minds will take a different form. Finally, the engagement of the audience is different from the response of the reader, given the appeal to the senses alongside the exposure to language. A mental-spaces analysis of a theatre event requires many distinctions. The primary dividing line falls between the in-theatre reality and the out-of-theatre reality. The significance of the latter manifests itself not only in the contribution of the author and her/his historical context, but also in that of the director, and the availability of material objects which are used on the stage to represent the frames associated with them. And, naturally, the bodies of the actors appearing on the stage participate in the outside reality as well. All these observations may seem obvious, but they contribute to the material aspects of theatre and the storytelling mode thus constructed. The in-theatre reality is more structured and more complex as well. We have to distinguish two physical spaces – the stage space and the audience space, divided by an invisible barrier.3 The categories are somewhat analogous to the distinction between the story space of a novel and the situation of reading. However, the audience space is limited spatially to the enclosure of the playhouse, while being physically separate from the out-of-theatre reality, and only conventionally separated from the stage, which contributes to the holistic nature of the theatrical experience. At the same time, the audience space can be thought of in terms of its visual and aural fields – both in its access to on-stage events and its separation from the out-of-theatre reality. The stage space functions as the physical embodiment of the story space, and thus houses all the story dimensions: space, time, characters, story events, et cetera; even the events which take place off-stage (another space!) are somehow made available to the participants on-stage, via the stage set, a report by a character, the very events of the play, et cetera. The stage space is further structured with stage-framed objects such as props, costumes, the sets, the lighting. It is, however, relatively independent of the text space, which helps construct many of the stage-space elements, but also adds the whole discourse dimension of the theatre. Crucially, the text space of the play is structured differently from any other storytelling text, as it consists in the turns taken by different characters, some of which represent discourse types different from the ones available in the out-of-theatre reality, such as a soliloquy, or the chorus. It also divides the story into chapterlike chunks (acts, scenes) and provides stage directions as a kind of discourse which mediates between the stage and the text. Stage directions are usually not numerous and guide basic movements of the characters, such as coming on or going off the stage. But they do suggest the need for further blending of the stage and the text in the emergent story. Each of the spaces is further structured by more conceptually important distinctions. The stage space, for one, poses interesting questions with respect to its spatial organization. In spite of its openness, it has gradually come to be
6.2 Mental spaces in dramatic narratives
143
understood in terms of a bounded space, so that the inner region is allocated to the representation of the story space. Contemporary theatre often returns to the traditional set-up and breaks ‘the fourth wall,’ but that does not make the wall disappear – rather, the concept is being played with. Apart from the in/out orientation, the stage is also structured by the front/back schema (downstage versus upstage) – where ‘front,’ or downstage, is closer to the viewers, and thus also closer to the center of the entire in-theatre space. Finally the up/down axis also plays a role, as I will argue in section 6.3. In some cases, the in/out and front/back orientation is used to distinguish the region off-stage as contributing to the story-blend. The region beyond the stage is not to be seen, but can signal events happening off-stage. Examples are numerous. In Chekhov’s play The Cherry Orchard, the final destruction of the trees, crucial to the story, is signaled only by the sound of wood being chopped, reaching the characters and the audience from off-stage. This case is simple, though, since the whole play leads to such a turn of events. In Stoppard’s Arcadia, sounds reaching us from off-stage are far more important. Arcadia is fairly unusual in its use of the stage space, since it alternates between the past (1809) and the contemporary present, never changing or removing the set or even the props accumulating on the stage. The past and present events take place in the same schoolroom in Sidley Park, giving a glimpse of the surrounding park through french windows. The park, however, is the center of attention, as it is being remodeled in the nineteenth-century subplot and then investigated in the twentieth-century subplot. Some crucial events talked about throughout the play are located in the park, including amorous encounters, but also hunts and duels. As a result, the sound of a popping gun can be associated either with the killing of a rabbit, or with a duel in which one of the characters can be expected to die. The sound of a gun coming from off-stage is used repeatedly to suggest the duel (which in fact never happened) and then the play goes on adding hunting for ducks and a killing of a rabbit as explanations, thus playing very cleverly with the viewers’ story-constructing processes and reminding the viewers of the duel as the central event. In this sense the sound of a popping gun is used as a means of evocation of frames existing outside the play (danger, killing) and constructed within the play (the duel, the hunt). But the sound alone cannot clarify which frame is to be accessed at any given point in the play. All the frames mentioned here belong to the past space of Arcadia. The play uses the past and the present to make the viewers solve the mystery and guess ‘what really happened,’ but it also blends them in interesting ways – for instance, the final scenes in the schoolroom feature the characters from the past and the present together on the stage, not talking to each other directly, but engaging both temporal spaces of the narrative. However, before the blend comes together on the stage and in the viewers’ minds, the sound coming from off-stage is used to begin the construction. Even though popping guns are heard
144
Fictional minds and embodiment in drama and fiction
throughout the play, they are primarily used to distract, until Scene 4, the last scene of Act I, which is taking place in the present and ends with the following stage direction: (1) The empty room. The light changes to early morning. From a long way off, there is a pistol shot. A moment later there is the cry of dozens of crows disturbed from the unseen trees.
Later, the ‘past’ sequence (Act II, Scene 6) starts very similarly: (2) The room is empty. A reprise: early morning – a distant pistol shot – the sound of crows.
The directions clearly suggest that the final moment of Scene 4 is a transition from the present into the past, and then Scene 6 picks up at that very moment. However, the transition also clarifies that the sound of the gun does not attest that there was in fact a duel, because a moment after the shot is heard in Scene 6 Septimus Hodge, one of the past characters, comes in with a rabbit he has just killed. The sound of the shot is thus not simply a profiling of events off-stage, but a narrative anchor negotiating between the real and the imagined versions of events at Sidley Park and guiding the viewer. This is just one example of how Arcadia uses the off-stage space, because the sounds of guns are heard in every act, further confusing the events narrated. Similarly, through most of the scenes in Arcadia, there is the sound of a piano next door, though it is not clear who is playing and whether the music comes from the present or the past. More narrative anchoring through sound off-stage means more story-construction effort on the part of the viewer and greater complexity of the narrative. The off-stage space can thus be seen as a narrative space which complements the events presented on-stage, while not necessarily repeating the narrative material. An interesting aspect of the structure of the stage space is the use of eye contact between an actor and the audience. It is not common, so as to preserve the boundedness of the stage space. In natural discourse, eye contact triggers the construal wherein the participants share a discourse space. In a sense, eye contact is often a correlate of an establishment of a deictic ground in which the participants can act as the speaker and the hearer. The fact that actors typically do not make eye contact with the audience is reinforced by the fact that they rarely speak to the audience. It happens, of course, but when it does, it signals a deliberate violation of the boundedness of the stage space. Furthermore, eye contact between an actor and the audience is even more conspicuous in a movie, especially since then it only goes one way – an actor-as-character speaks to the viewer, or, more accurately, to the eye of the camera. Every time this happens, as in some scenes in Woody Allen’s Annie Hall, the viewer is being pulled into the space of the story and made to feel like a participant, not just a witness. This is different from the standard construal of the theatrical or movie audience,
6.2 Mental spaces in dramatic narratives
145
wherein viewers are expected to construct the story in their minds, not to take a role in it. However, the audience is also an addressee of the entire performance, and thus the occasional breaches of etiquette which make the viewer feel like a participant are not in fact exceeding the limits of the entire theatre set-up. The text of a play is of course different from the text of any other narrative. Not only is it structured primarily as dialogue, but it also includes varieties of discourse which are typically not found either in other narratives or in colloquial speech: a character may be speaking to her/himself, to a material object (a prop), to a mental image, or to a dead body of another character, et cetera. All these forms of discourse play a role in story construction and need to be seen in the context of the narrative goals a play has to fulfill. As a result, construction of meaning and the emergent story in a theatrical performance require that we account for all the spaces involved, as represented in Figure 6.1. The basic division runs through the middle of the diagram, so that the left-hand side represents all the material aspects of the play, and the right-hand side is correlated with the contribution of the text. But neither one is a holistic and uniform concept, as each consists of more spaces, some of them overlapping to a degree. This combination of the material and the linguistic is at the core of the genre. In what follows I will use selected examples from Shakespearean plays to illustrate the functioning of all spaces and to clarify the nature of the interaction among them. I will argue that a play tells a story through textual and extra-textual prompts and that the drama story is a blend emerging out of the audience’s interaction with on-stage prompts and off-stage frames. Theatrical discourse uses all the spaces available, but it relies crucially on the theatre-specific blends of the material and the textual.
Stage space set props up/down front/back time performance Back-stage space
Text space language (form and frames) discourse and constructions
Emergent story space characters events
Reality space actors Audience space objects historical facts listening seeing constructing meaning
Discourse spaces chorus/audience character/character character/object character/body ... Narrative spaces narrating plot construction characters’ minds
Figure 6.1 The mental space schema of a theatrical performance
146
Fictional minds and embodiment in drama and fiction
6.3
Materiality of the stage and fictional minds
Throughout this section, I will be looking at the material aspects of the stage space and their interaction with the story construction processes. There is an inherent difficulty in matching the narrative construction taking place in novels with the parallel processes occurring in a context where every word is spoken by someone. It is probably clear that there is no natural way to treat dramatic discourse as an example of spoken discourse, although in contemporary texts there is more and more convergence. However, in what follows I will be considering examples from Early Modern plays, hoping to find the core concepts of storytelling in these texts; later, different stylistic means represent the concepts in the context of a novelistic tradition. Discussing the materiality of the stage requires distinguishing various instantiations of the phenomenon: the stage itself, the sets, the props, but also the actors’ bodies, as well as material and embodied actions, such as manipulation of objects, body posture, gesture, and movement. Some of these elements are present on the stage as a part of the representation of the storyworld – a character needs to be played by an actor with matching physique, a chair is usually on the stage to be sat on, and a sword is used in stage fights. These are all elements of the materiality of theatre which rely more or less accurately on representation blends, in which the represented reality (including a fictional reality) is fused with the representing reality. I have talked at length about representation blends in literature in earlier chapters, and will not devote more attention to them. I will focus, instead, on the cases where material objects and bodies on the stage are used to do more than represent and participate in the story construction and the emergence of interpretation. The difference between representation and emergence of meaning is rather basic here. The former consists in the kinds of blends we interact with every day, while the latter requires a context wherein the object is used beyond the ordinary. Out of context, an empty chair will naturally be understood as available for sitting on, and probably not as a prop in a cabaret performance. But when Liza Minelli sings in Cabaret, she uses a chair as a dancing prop, never actually sitting on it. That is where the chair acquires its new meaning. Similarly in Arcadia. A gun, or a sound of a popping gun, evokes the frame of using the gun as a weapon – fighting, hunting, or killing.4 But for it to evoke a specific duel and participate in the construction of the entire story, which depends on whether the duel happened or not, more than an ordinary evocation and representation blend is required. The frame used is story-specific and emerges as the play goes on, rather than being available instantly. The same is true about material objects in Arcadia. The table in the center of the schoolroom, which acts as the centerpiece of the set, gradually accumulates many objects, left there by participants from both the past and the present subplots: an apple, a
6.3 Materiality of the stage and fictional minds
147
sleeping turtle, a book, et cetera. The objects, such as the apple, have no specific role to play in the story, but when a character in one subplot takes a leaf off it and then another character bites it in the other subplot, the apple begins to work as an anchor, as it connects two inputs of the blend gradually emerging through the play. Since the anchoring and blending function of objects does not properly belong to the common meaning of the word ‘prop,’ I will not use the word in my discussion. By the same token, this approach does not quite mesh with the discussion of props in literature (see, for instance, Sofer’s fascinating book, 2003). Similarly, when actors’ bodies are used on the stage to support the metaphorical meaning of a scene or to foreground crucial metonymies constructed through language, this is not a specific aspect of acting in general (and so is not included in Graver’s in-depth review [1997] of options), but a construal in which various aspects of the theatre environment are used in meaning synchrony. In the cases discussed below, the body is used to represent the mind, but this requires that language and embodiment work together in a stage-specific environment. 6.3.1
Narration on the stage
The role of a narrator involves various tasks: introducing the setting and the characters, managing the flow of time, revealing or hiding information as the rules of suspense demand, and, quite often, also telling the reader ‘what happened.’ Also, one of the major roles of a novel’s narrator is to provide some access to the minds of the characters, either by giving the reader direct access to thoughts or by reporting them or mediating between a character’s thought and the narrator’s own evaluation. None of these tasks can naturally be performed in a play, and, crucially, it is clear that all the relevant information cannot be presented on the stage – either because it cannot be staged realistically, or because it would take too long, or because the information (such as a character’s mood) cannot easily be expressed in plain words. However, the need for some rudimentary narrative mode has been present in theatre since its earliest forms. A classical tragedy chorus performs some of the narration tasks and comments on the events in the story in a mode best described as ‘metanarrative’ – commenting on what has happened rather telling the viewers what happened. But a chorus is rare and can only frame the narrative in the most general terms, as the choruslike characters in Romeo and Juliet, who describe the basic context at the beginning and comment on the events at the end. What dramatic discourse does, then, is also use characters as tellers – a blend which is very natural in the context of a play. In Elizabethan drama, characters report events which have taken place off-stage, often as eyewitnesses, messengers, et cetera, so that typically it is realistic to assume that the character offering a report had reasons both to witness the events and to come and give an account
148
Fictional minds and embodiment in drama and fiction
of them. However, on occasion such limits of plausibility are crossed. When Ophelia dies in Hamlet, by drowning in a stream, her death is reported to the other characters by Gertrude, in beautiful and moving detail. However, Gertrude never makes it clear how she came to know what happened, and there is also a nagging question her report prompts – if someone did witness the death of a young, innocent girl, why wasn’t she saved?5 The question, though perfectly reasonable, also presupposes its own answer – because what happens on the stage is not meant to be realistic, because Ophelia couldn’t have drowned on the stage, and because someone had to tell the story as a result, so that the viewers know it. Thus what a character says to a character on the stage is not necessarily a token of actual discourse between two people, but it may in fact be addressed to the audience. It is of little importance that Gertrude was not the person who witnessed Ophelia’s death, but in view of the fragment’s narrative role it is more important that Gertrude’s own viewpoint as a character is not involved – she is hardly more than a mouthpiece. There are also events which cannot be plausibly represented by characters. In such cases the narrating can also be done by ghosts (as in Hamlet) or other supernatural beings. Thus the witches who tell Macbeth about his fate not only foreshadow the events of the play, but also seem to reveal Macbeth’s most secret desires. The witches are undoubtedly adding to the visual appeal of the story, but they have a role to play, beyond being mysterious and scary, as the viewer needs to understand the forces driving Macbeth to action. It is not surprising, then, that in Kurosawa’s film Throne of Blood, an adaptation of the story of Macbeth, the witches are replaced by an apparition more appropriate to Japanese aesthetics – a single pale spirit, turning a spinning wheel (a symbol of fate in numerous cultures). At the same time, the words of the spirit and of other characters in the movie clearly identify the role of the prophecy as a prompt or a catalyst, as they attribute secret lust for power to Washizu, Macbeth’s counterpart in Kurosawa’s adaptation. In a sense, the movie relies on the aspect of Macbeth’s story which I find crucial – its focus on the character’s mind, not on the events. All in all, supernatural interventions in both texts have a role to play in both narrating events and revealing the characters’ dispositions. 6.3.2
The vertical dimension of the stage and representation of mental states
The stage is conceptualized primarily as a bounded region, structured from the point of view of the audience as having a front and a back. It is less common, though, for the vertical direction to start playing a conspicuous role, so it is all the more surprising to see the up/down schema structuring the primary themes of a play – in Shakespeare’s Richard II. It is made possible in the context of an Elizabethan stage, which was divided into a lower and an upper level, so that the
6.3 Materiality of the stage and fictional minds
149
spatial division along the up/down schema fits naturally there. Crucially for the approach taken here, the material feature of the stage is exploited throughout the play and turned into the main motif, linking various aspects of the story together, but also being reinforced through all available modalities. The vertical schema emerges in the language of the play, but is also matched by the use of the upper and lower stage, as well as by the gestures of the characters and the metonymies relying on the verticality of the actors’ bodies. The synchrony across all these modes of expression is surprisingly salient, and is then used metaphorically to structure the complexities of power struggle and personal degradation which are central to the meaning of the play. The up/down schema is the basis of many orientational metaphors. First described in Lakoff and Johnson (1980), these metaphors use the vertical orientation to structure many polarized concepts, such as ‘status is up/lack of status is down (He is climbing up the social ladder), ‘well-being is up/lack of well-being is down’ (We’ve been sinking for a while), ‘happy is up/unhappy is down’ (Good to see your spirits are up!), and many others. Many of them are relevant to the play, in which King Richard loses his crown, is humiliated by his rival, and suffers discouragement, rejection, and pain. Recurrent vertical images in the play foreshadow events and highlight their meaning in the context of the story, while also reflecting the king’s feelings of loss and disappointment. Early on, in Act I, the king’s rival, Bolingbroke, approaches his sovereign with due respect: (3)
Lord marshal, let me kiss my sovereign’s hand, And bow my knee before his majesty; (I.3)
Kneeling in front of another person and bending your head down to kiss the king’s hand are both body postures signifying submission and respect. Richard receives the gesture kindly and offers to abandon his high position in return: (4)
We will descend and fold him in our arms. (I.3)
This simple exchange of courtesies has some significance, though. Here, so early in the story, Bolingbroke still maintains the pretence of being the loyal subject, and this is signaled through body posture and its metonymic function. On the other hand, the king’s willingness to ‘descend’ not only shows friendship, but also foreshadows his imminent downfall as partly caused by his softness. But the Duke of York is not fooled by Bolingbroke’s gesture and comments later: (5)
Show me thy humble heart, and not thy knee, Whose duty is deceivable and false. (II.3)
His response not only underlines the intended meaning of the body posture, but also highlights another body-part metonymy, wherein the heart is the locus of one’s feelings.
150
Fictional minds and embodiment in drama and fiction
Later in the play, in Act III, Richard knows he is losing the battle for his crown, but reassures his men, evoking the up/down schema again: Look not to the ground, Ye favourites of a king; are we not high? (III.2). In the next scene Richard awaits Bolingbroke’s arrival on the walls of the castle. He is thus still ‘high,’ occupying the upper level of the stage and his high position, but he knows the war is lost. He engages in a long reflection on the coming events: (6)
What must the king do now? Must he submit? The king shall do it: must he be deposed? ... Or I’ll be buried in the king’s highway, Some way of common trade, where subjects’ feet May hourly trample on their sovereign’s head; For on my heart they tread now whilst I live; And buried once, why not upon my head? (III.3)
The vocabulary of this monologue reinforces the downfall Richard expects (submit, be deposed). Richard also constructs a spatial image of the reversal of his fortunes. He should be towering over his subjects, but his burial might mean that they will trample on him, and his power of control. In effect, to describe his humiliation, Richard creates more levels of up-and-down structure, by imagining himself buried under the feet of his subjects. The body-part metonymies are very meaningful here: the head, as the center of control, and the heart, as the center of feeling, have both fallen so low that the subjects feet are above. The power of the scene comes directly from the broad figurative exploitation of metonymy, and the projecting of much of it onto the bodies of the participants. Finally, Northumberland arrives as Bolingbroke’s emissary, and asks Richard to come down and meet Bolingbroke in the court. In the context of his actually being ‘up,’ both on the castle wall and on the stage, the request has a very straightforward spatial sense. But Richard’s words immediately expand the meaning of the scene into the metaphorical sense of ‘losing power’: (7)
Down, down I come; . . . In the base court? Base court, where kings grow base . . . In the base court? Come down? Down, court! Down, king! . . .
As in the preceding scenes, the wording describing descent to the lower court is exploited to prompt the metaphorical reading. Crucially, the structure of the stage and the motion of the actor’s body are also co-opted into the rich construal of the fate awaiting Richard. Later in the scene, Richard is waiting in the court as Bolingbroke arrives. The king knows the meaning of this, but Bolingbroke upholds the pretences. Because of all the metaphorical load of the previous scenes the viewer will not be deceived by Bolingbroke’s politeness as he kneels again. Richard as well
6.3 Materiality of the stage and fictional minds
151
understands the empty gesture, as he comments Fair cousin, you debase your princely knee . . . Then he lifts Bolingbroke and makes him stand up, but, contrary to his words, this is naturally understood not as courtesy, but as an admission of defeat. Then he continues: (8)
Up, cousin, up; your heart is up, I know. Thus high at least, although your knee be low. (III.3)
The contrast between Richard’s earlier Down, down I come and Up, cousin, up is in each case built into the bodies of the actors and their motion, but the metaphorical sense prevails again. Perhaps the most interesting part of the line is the expression Thus high at least – stage directions suggest that Richard touches his head as he says this, but even without them it is entirely clear that his head, the locus of the crown and the center of control, must be in focus here (especially since both the ‘heart’ and the ‘knee’ metonymies have been exploited in the same sentence). The play is again building its meaning into the actors’ bodies. Throughout the play, the same metaphors (good is up; status is up; happy is up; . . .) and metonymies (head for center of control; heart for feeling; knee for submission; . . .) are profiled through a complex blend of linguistic and material means, including the body, motion, gesture, structure of the stage, and language. Crucially, meaning is prompted through various combinations of linguistic and embodied means. In the case of Down, down I come the actor is announcing and/or performing motion downwards and speaking at the same time. Such a description of action to be taken is less likely to happen in spontaneous discourse, but it may occur when the consequences of the action are highlighted, as when a parent warns I’m coming in so that the teenager is given one last chance to start behaving. But in the play the words and embodied actions may be used to prompt for independent meanings, as when Richard touches his head while talking about his opponent’s heart. Only his gesture suggests the crown, while the language remains a more typical instantiation of an orientational metaphor of status. In Cienki’s discussion of metaphoric gestures (2008), embodied motion typically represents the source domain of the metaphor (so up/down motion could signal well-being or social status), but Richard’s gesture, being metonymic, points to the head to signal the crown, even though in the scene Richard is its bearer, in the political sense, or literally wearing it. In other words, the gesture makes sense in the context where Richard is thought of in terms of his role, the king, rather than in reference to his personal tragedy. The various embodied and linguistic means thus prompt various instances of a rich blend relying on the body posture, metonymic meanings of body parts, social and political context, et cetera. It is because of this rich blend that Richard’s deposition is given its political,
152
Fictional minds and embodiment in drama and fiction
emotional, and moral complexity. We can refer to this overall mapping as Body/ Mind blend. The Body/Mind blend prompts for up/down metaphors throughout the play,6 giving unity and poetic force to the story. Furthermore, the means of story construction here support the earlier discussion of a variety of roles theatrical discourse can play. When Richard is describing his emotions and commenting on his fate, he is speaking in the presence of his courtiers, and thus, ostensibly, to them. But the lengthy monologue does not really function as ‘direct speech,’ but provides a view of Richard’s understanding of the events, and thus addresses the audience first of all. There is perhaps no need to distinguish in every case between words addressed to characters and words addressed to the audience, since, obviously, the audience is the ultimate addressee in any case and thus the boundedness of the stage space is naturally permeable to anything that can be seen or heard. But, linguistically speaking, the discourse that profiles a genuine interaction between characters treats the audience as primarily an ‘overhearer’, and is thus different in terms of both linguistic form (such as style, length of the turn, et cetera) and content. 6.3.3
Material objects and the human mind on the stage
Shakespearean drama has often been described as creating the ground for literature to represent and shape the human mind. The source of the phenomenon is often sought in the power of poetry and the complexity of the characters and their plights. I want to argue, however, that the effectiveness of Shakespearean drama is partly due to his truly ingenious use of the materiality of the stage and the construction of complex blends which represent thoughts and emotions in creative ways. I will look in more detail at the tomb scene in Romeo and Juliet (V.3). Romeo arrives to find Juliet lying in her tomb, and assumes she is dead. He is not aware that she is only in a drug-induced deep sleep, and thus he succumbs to despair. He kills himself, drinking potent poison, and falls dead. A moment later, Juliet awakes, and seeing her husband dead, kills herself with his dagger. The events of the scene are dramatic and emotionally loaded, but the crucial aspect of what we see is the representation of the feelings of the two young lovers. How can this be achieved? One of the interesting features of the scene is that both Romeo and Juliet are, in turn, the only characters speaking. That is not unusual, and most Shakespearean soliloquies are based on a similar idea, but the discourse of this scene is still surprisingly animated and uses a number of constructions resembling spoken discourse. Both Romeo and Juliet are justifying their decision to take their own lives, but their words sound much less abstract than, say, Hamlet’s reflection on the reasons to refrain from suicide in the famous To
6.3 Materiality of the stage and fictional minds
153
be or not to be soliloquy. The crucial difference, it seems, is that Hamlet talks about death as an abstraction, while Romeo and Juliet perform acts of suicide, which requires that they both talk about death, but also perform the acts of killing themselves. Hamlet’s proverbial inaction seems to be reflected also in the discourse the character engages in. For comparison, each of the lovers in turn addresses the body of the other, they also speak to objects (poison vial, dagger) and to their own body parts. None of these ‘addresses’ can seriously be taken to be genuine vocatives, but they do serve a function which needs to be described in terms of both linguistic constructions and embodied cognition. Also, the use of framing and metaphor here is again, as in the Richard II scenes, closely intertwined with the material aspects of the scene. Overall, I will argue that the scene represents the characters’ minds by relying on material objects, language, and actors’ bodies in a very complex way. When Romeo finds Juliet dead, he sits by her and looks, surprised to see her unchanged. He delves into a complex metaphor wherein death is an adversary in a love contest, trying to take Juliet away from her Romeo. The metaphor of death as an adversary is common in colloquial language, while the only one perhaps even more common is death is departure, as in She passed away, She left us, et cetera. Thus we might expect Romeo to talk about leaving the world of the living, going away, following Juliet, but he does not. Instead, he says: (9)
For fear of that, I still will stay with thee And never from this palace of dim night Depart again: here, here will I remain.
The decision to remain ‘here’ makes perfect sense if we consider that the scene is taking place in a tomb, where only the dead dwell. But at the same time, Romeo’s decision to die is portrayed as less dramatic, almost obvious – what could be more natural than a husband remaining by his wife’s side? Moreover, this metaphorical reversal also reinforces the setting of the scene, and would not make much sense without it. Again, the text is relying on the materiality and framing of the stage and thus foregrounding its metaphorical meaning through what might seem a plain description of a spatial location. As in all the best uses of metaphor, the literal and the figurative are becoming the same. But Romeo is also aware that dying means losing consciousness, and thus not being able to delight in the presence of his love: (10)
Eyes, look your last! Arms, take your last embrace! and, lips, O you The doors of breath, seal with a righteous kiss A dateless bargain to engrossing death!
These faux addresses to the body (eyes, arms, lips) prompt a different series of blends. They evoke the body parts and the sensory and emotional frames
154
Fictional minds and embodiment in drama and fiction
associated with them, so that the viewers can appreciate Romeo looking at his wife for the last time, embracing and kissing her in a gesture of parting, even if they cannot see the lovers properly. The way these linguistic forms work is quite complex. From the formal perspective, the constructions consist of two parts – a vocative (O, eyes) and an imperative (look your last). But the entities evoked by the vocatives have no intentionality, and thus cannot be implored to act in any way, which raises questions about the meaning of the imperative here. In fact, imperative form in English is often used not to impose actions on others (as would be the case in standard instances such as Bob, open the door, please!), but to profile imagined future events (Just open the door and you’re finished) (see Dancygier and Sweetser 2005 for more discussion). The door-opening might happen in either of the two instances exemplified, but the verb form itself does not determine whether the speaker desires it or not. Similarly, in (10), Romeo’s desires are not part of the construal, even though he is describing actions his own body might take. But the goal, in the context of the stage, is not to describe what he wants or orders, but what he will imminently do. The ‘address’ and the ensuing imperative (look your last, take your last embrace), evoke the actions Romeo is taking as he is speaking, along with their emotional meaning – which is that of parting. In effect, the language is describing what Romeo is doing, and framing his actions emotionally in an unambiguous way. It is thus a performative blend similar to the ones described by Sweetser (2000), such as performing a hunt-dance before the actual expedition, as a way of increasing the chances of a successful hunt. Crucially, the information which needs to reach the viewer is reinforced through poetic evocation of the actor’s body and its actions. A moment later, Romeo ‘addresses’ the poison vial in his hand and drinks the potion: (11)
Come, bitter conduct! Come, unsavoury guide! Thou desperate pilot, now at once run on The dashing rocks thy sea-sick weary bark! Here’s to my love! – O true apothecary! Thy drugs are quick. – Thus with a kiss I die.
As in the cases above, the address and the imperative describe (rather than demand) the imminent act of drinking poison, while announcing Romeo’s performance of it. Also, the scene is typically played with Romeo’s hand holding the vial raised in front of or above his face and his eyes looking up at it – incidentally, this also allows the viewers to recognize the vial for what it is and see Romeo’s face well. But there is an additional element – the terms used to refer to the vial. The nouns are conduct, guide, and pilot – all of them metaphorically referring to the expectation of what the poison will do, rather than what it is, thus describing a change of state (death) conceptualized as transfer from one
6.3 Materiality of the stage and fictional minds
155
location to another. The initial metaphor of ‘staying here’ is substituted with the standard departure metaphor for death. At the same time, the adjectives – bitter, unsavoury, desperate – gradually move away from the expected sensation of taste to Romeo’s emotional state. The emotional attributes of the young man committing suicide are transferred onto the instrument he uses in achieving his goal. Also, in a brief evocation of the apothecary who provided the poison, Romeo announces that the drug is taking its effect quickly. And so he dies, having also talked his way through the process in very moving detail. Remarkably, this part of the scene takes significantly more time on the stage than it would be likely to take in real life, even including all the parting gestures. The story events are actually crisp and fast-moving, but the pace of the play is different. But good storytelling does not have to be realistic in matching stage time and story time. In fact, as in all narratives, narrative time is allocated based on the emotional load and overall importance, and not some realistic rendering of time flow. The play does the same, but cannot just extend the time by adding words or slowing down the movements of the actors’ bodies: the engagement of the actors through discourse and gesture helps in animating the stage time devoted to an emotionally loaded event. A moment after Romeo dies, Friar Laurence arrives, too late. Juliet wakes and sees Romeo dead, but she is immediately urged to run and seek shelter elsewhere. Her response matches the lines earlier spoken by Romeo, as she is clearly determined to stay: (12)
Go, get thee hence, for I will not away.
Again, her words match both the literal meaning of her refusal to escape and the metaphorical mapping death is a staying in a location, set up earlier by Romeo’s words. Then a pageboy warns her about a noise, and she makes her decision quickly: (13)
– then I’ll be brief – O happy dagger! This is thy sheath; there rust and let me die.
Again, her words mimic the pattern established earlier in Romeo’s words, as she speaks to the instrument of her own death. The crucial difference is that while Romeo used the adjective desperate, she describes the dagger as happy – the use of the adjective signals her more willing acceptance of death, but can also be read as framing the dagger not so much as an instrument of death, but as an embodiment of the lost lover, welcoming the intimacy. The use of language throughout the scene suggests its reliance on blends which allow the characters to reveal their thoughts and feelings without stating them, and while engaging the material objects in unusual linguistic patterns. In the blends, feelings are attributed to objects which only humans can experience, and
156
Fictional minds and embodiment in drama and fiction
Discourse frame speaker addressee
Killing frame killer victim instrument
On-stage frame speaker/killer speaker/victim addressee/instrument
Figure 6.2 Acting/Speaking blend
prompted to act in ways which match human actions. These are indeed blends, specific to the stage space, where things appear in order to participate in the construal of the story.7 The cases of the poison vial and the dagger are similar. With the exception of their being used with adjectives suggesting different emotions, they both suggest death, and they frame people both as users of instruments of death (killers) and as victims. But on the stage, these props are also framed as addressees, and characters about to use them as speakers. As a result, the roles in the action of killing are blended with the roles in the discourse on the stage. Speaking and acting become indistinguishable aspects of the performance. As Figure 6.2 shows, the blend involves two frames. One is the frame of killing, metonymically evoked by the objects (poison vial, dagger); the frame profiles the roles of ‘killer,’ ‘victim,’ and ‘instrument.’ The other input is the general discourse frame, profiling a speaker and an addressee. In the on-stage frame, a blend of the other two, the speaker is both the killer and the victim, while the instrument is an addressee. In the emergent meaning, speech and action are blended, in that the characters’ words evoke the act of committing suicide, which is taking place in the same narrative space, albeit moments after the discourse ends – so that the viewers can absorb the frames and apply them to the events on the stage. I’ll refer to this kind of blend as the Acting/Speaking blend. The address to the body parts, quoted in (10), prompts a very similar blend (see Figure 6.3). The frames evoked, in the context of the relationship between the characters, are connected with affectionate behavior: eyes evoke admiration (as they consistently do in Shakespearean sonnets), arms intimacy, and lips passion. There is no ‘acting’ in the proper sense, but there is showing one’s feelings of love and parting. In the on-stage blend the speaker is the experiencer of the feelings, the addressee is the body part associated with the feeling, and the other lover is an object of emotion. The nature of the acts profiled is
6.3 Materiality of the stage and fictional minds
Discourse frame speaker addressee
157
Emotion frame experiencer associated body part object of emotion
On-stage frame speaker/experiencer addressee/associated body part addressee/object of emotion
Figure 6.3 Feeling/Speaking blend
different, as the role prompted is not ‘agent’ but rather ‘experiencer,’ so in this variety we can talk about a Feeling/Speaking blend. Both blends work similarly: material objects in the stage space metonymically evoke frames. These frames profile roles which are not directly available on the stage, and then the speaking character is profiled in one (or more) of the roles. The use of the blends allows characters to talk their way through important acts on the stage which, if they were to be simply performed, would seem more like pantomime. The discourse of the drama is a crucial element of the functioning of the stage space, but often needs to be materially supported through objects brought onto the stage in order to evoke frames necessary in processing the events. Ultimately, these blends also make better use of stage time, as they give more temporal and discursive salience to the crucial elements of the scene. The blends connecting acts on stage and feelings of characters involve props and the actors’ bodies. None of the blends above are intended to represent what characters think, but we might expect to find such blends as well. It seems that they are represented by the cases where Romeo speaks to the dead bodies, lying there in the tomb. He addresses the body of Paris (“O, give me thy hand . . . I’ll bury thee in a triumphant grave”), of Tybalt, whom he has killed (“Tybalt, liest thou there in thy bloody sheet?”), and, of course, of Juliet (“Ah, dear Juliet, why art thou yet so fair?”). Romeo finds himself in a tomb, surrounded by the bodies of people who have died in the recent events, so it is natural that he would reflect on their fate. As in the other parts of the scene, he is choosing addressees who cannot answer, but their role is that of objects of reflection. The fact that the vocatives may play such different roles in the constructions does not make the construction any less consistent in its performative sense – when a single character on the stage addresses any material aspect of the surroundings, be it
158
Fictional minds and embodiment in drama and fiction
a material object, a dead companion, or his own body, the on-stage meaning is bringing the object to the attention of the audience and describing the character’s thoughts, feelings, or actions focused on that object. Thus Romeo buries Paris as he talks about it (Acting/Speaking blend), while both Tybalt and Juliet are addressed as objects of his reflection, in a variety of the blends established above – Thinking/Speaking blend, wherein ‘speaking to’ stands for ‘thinking about.’ It seems useful to distinguish the three kinds of blends, in spite of their similarity, because they are achieved through different discourse patterns. They all profile aspects of the characters’ actions, emotional responses, and mental processes, and they are all marked by faux forms of address. However, the cases differ not only with respect to the choice of the faux addressee, but also in the aspects of the mental and emotional functioning to be represented by various aspects of the body. As in the case of Richard II, discourse is structured by the Body/Mind blend. The specific blends profiling actions, emotions, and thoughts are all instantiations of this larger blend. The kinds of constructions we find in these cases, as well as the kinds of blends, distinguish the discourse of drama from other forms of literary discourse, since it relies heavily on the availability of material objects and on the presence of actors’ bodies. It also agrees with the story-constructing needs of the viewers, who have to rely on all the cognitive modalities the theatrical performance appeals to: hearing, vision, processing language, understanding gesture, and evocation of frames. The Body/Mind blend, in its various theatrical instantiations, is not as unique to the theatre as the specific forms of discourse are. The ‘Mind Is a Body’ metaphors are very common, and sometimes utilize very complex domains. For example, the domain of ‘food’ is naturally co-opted into the construal of thought, as in thinking is chewing, swallowing is accepting an idea, digesting is understanding, et cetera. However, there is an important difference between such uses and the blends just described. The onstage blends do not rely on the knowledge of the body and its functioning – on the contrary, they rely on the direct evocation of embodied experience, as relayed through the body of the actor. The fact that Romeo talks about the ‘last embrace’ is only a part of the impact – the actor playing Romeo embraces the actress whose body represents Juliet. That is why the language surrounding such acts can be unclear or unusual, or rely heavily on metaphor (recall the conduct, guide, and pilot terms, all describing poison), since the blend relies in the same degree on the performance of the acts in question. The props and the actors’ bodies thus function as ‘material anchors’ in the sense introduced by Hutchins and as ‘narrative anchors’ (see Chapter 2). The anchoring effect is clear in Romeo and Juliet, where props play such an important role, but there are objects used in other plays which are central to the entire narrative construction – enough to mention Desdemona’s handkerchief in Othello (cf. Yachnin 2002). The handkerchief is a token of the bond
6.3 Materiality of the stage and fictional minds
159
between Othello and Desdemona, and thus its appearance in the hands of another man has to be understood as betrayal. In fact, it does not seem to matter much what kind of object plays the role of such a token – it could have been a ring, for one. What matters is that the object can be framed as a gift, as something one can carry around at all times, and which is at the same time easy to take away (as the actual events in the play demand) or to lose (as Iago’s plot construes it). The choice of the handkerchief is possibly the best narrative choice, as it fits all aspects of the description perfectly and is easily seen from a distance. The actors’ bodies are also used as anchors in all plays that use cross-dressing (Two Gentlemen of Verona, Twelfth Night, et cetera), as they prompt identity blends which inevitably play with the femininity and masculinity of characters. In the context of Elizabethan theatre, the blends are further complicated by the fact that boy actors may thus represent women who pretend to be men – a fascinating issue for blending theory, though beyond the scope of this chapter. The blends shown are also interesting in that they can be seen as precursors of the numerous constructions of represented speech and thought, used abundantly since the nineteenth century. Some of the forms, such as free indirect discourse, can be claimed to be specific to literary discourse, and thus the above discussion of the Body/Mind blend does not make these blends so exceptional. The representation of the characters’ minds may be simple – they can just tell the audience how they feel, via a confession to other characters or in a soliloquy. But in the scene described, feeling is only a part of what needs to be appreciated, as there is action as well – and action so crucial to the story that there can be no misunderstanding (what if the viewers missed the fact that what Romeo drinks is poison?). Moreover, it is infinitely more true to the genre to create constructions which also use the very bread-and-butter of theatre – the materiality of the stage, of props, and of the actors’ bodies. Shakespeare has clearly mastered the skill. The linguistic constructions he uses are clearly a part of the Early Modern poetic repertoire, as can be seen in the final scene of Marlowe’s Dr Faustus. The main character is on the stage alone, being dragged to hell by frightening creatures. The monologue is rather long, and in the course of it Faustus is using a substantial number of constructions strikingly resembling the vocativecum-imperative pattern of the Romeo and Juliet scene. He speaks to himself (Ah Faustus, Now hast thou but one bare hour to live), to God, to Christ, and to Lucifer, to mountains, to stars, to Earth (Earth, gape), to his body and soul (Now, body, turn to air . . . O soul, be changed into little water drops . . .), to serpents and adders, and finally pleads Ugly hell gape not, come not Lucifer. The scene allows Faustus to display a range of emotions (desire for a reprieve, fear, guilt, regret), but it also allows the viewer to follow what happens to Faustus and appreciate the bodily pain of being dragged to hell by fierce serpents. The Body/Mind blend constructions are the primary means in the scene of allowing the audience to grasp what Faustus’s body and mind go
160
Fictional minds and embodiment in drama and fiction
through as he dies. The fact that all the vocatives call up frames of imagined situations or otherworldly occurrences explains the lack of material support on the stage, but the constructions remain the same – they represent Faustus’s thoughts and feelings. 6.3.4
Ghosts and other supernatural occurrences on the stage
The preceding section mentions the witches in Macbeth as a good example of how the stage uses embodiment to represent mental constructs. The witches take a material form, but they seem to be more of an unacknowledged voice in Macbeth’s head than the sources of knowledge of the future. They may at the same time be modeled after a tragedy convention when an oracle or a sphinx predicts future events which seem very unlikely, so that the play can then show the workings of fate. But considering the rest of the play, Macbeth is not simply manipulated by fate, and the role of the witches may be closer to revealing to him what he does not wish to acknowledge. For comparison, Oedipus does all he can to avoid his prophesied destiny, but fails because in the classical tragedy world fate rules a person’s life entirely. Macbeth could not avoid becoming thane of Cawdor, but the fulfillment of the rest of the prophecy is his own doing. The witches may be telling him what will happen, promising ultimate power, but they may also be helping him to realize what his desires are. Another example of a meaningful supernatural intervention is the Birnam wood. In this case, Macbeth reads the conditional prophecy as a promise of impunity, assuming nature would not allow the forest to move. So when the ‘forest’ arrives, Macbeth is supposed to receive this as a realization that his crimes violated the natural order in a way that cannot be forgiven. Macbeth is a play full of supernatural interventions and mysterious apparitions, but all these instances are carefully constructed to contribute to our understanding of what is going on in the characters’ minds, rather than in the story alone. From this perspective, it seems possible to talk about Macbeth as a play about the mind. Not all of the subjective construals in Macbeth involve material occurrences on the stage. In the famous scene in Act II where Macbeth speaks to the dagger ‘in his mind’s eye’ (“Is this a dagger which I see before me, The handle toward my hand?”) the viewers hear about the dagger but cannot see it – indeed, how could they, if Macbeth himself seems unsure about the nature of this vision? It is possible, of course, that the difficulty in giving a real dagger its ghostly appearance has prevented the writing of the visible dagger into the play, especially since Macbeth also draws out a real dagger, an instrument of the murder he has planned. But the entire monologue, preparing the viewer for Duncan’s murder, is another example of the faux address, as described above, and thus uses a verbally constructed vision of the material object as a catalyst through which Macbeth’s unspoken thoughts could be represented. The
6.3 Materiality of the stage and fictional minds
161
underlying materiality of the immaterial dream dagger plays the same role it did in Juliet’s suicide scene. In fact, the monologue constructs two strands of thought, not one, which was clearly shown in the movie directed by Polański. Macbeth and the viewers both see the dagger, eerily floating in the air, to the accompaniment of ghostly sounds. However, the dagger is visible only when Macbeth is directly addressing it in his thoughts, and in these fragments the text is heard as a voiceover: (14)
I see thee yet, in form as palpable As this which now I draw. Thou marshall’st me the way that I was going; And such an instrument I was to use. (II.2)
But when Macbeth reflects more soberly upon the strange appearance in his mind, the dagger is no longer visible, and the actor turns his head away from the apparition and begins to speak: (15)
Mine eyes are made the fools o’ the other senses, Or else worth all the rest.
The visual choices made by the director help underscore the claims made above: the discourse of the play needs to represent the hidden inner thoughts of characters, and often relies on material objects to serve as addressees in linguistic constructions specializing in thought representation. It may also represent various levels of inner reflection, through choices involving all the material aspects of the stage/movie environment – the actor’s body, the actor’s voice, and material objects. Returning now to the question of what is or is not visible, there is a significant difference between things only a character can see (and thus has to describe to the viewers), those that the audience can see as well, and those that other characters can see too. The witches, for one, appear to Macbeth and to Banquo, but apparently not just because the fate of both of them is revealed to each one and to the audience. The way I am reading the scene, Banquo is there so that Macbeth’s pride in the high position he is promised can be hurt by the prediction that Banquo’s descendants will inherit the throne after him. As all the stories of this kind convince us, the ambition to be king is nothing compared to wanting your children to continue being kings – and the witches put Banquo in the way of that desire, creating a conflict between the two characters. Interestingly, Banquo himself is not tempted to take action to help fulfill the prophecy – otherwise, what would stop him from killing Macbeth? But the logic of the play is focused on Macbeth and the conflict between his hidden desires and his sense of morality. This might support my view of Banquo’s role in the play. The whole scene with the witches is crucial for the viewers to appreciate the conflicting desires. But they don’t have to see the dagger, especially if it is to be
162
Fictional minds and embodiment in drama and fiction
clear that it is in Macbeth’s imagination. Once it is mentioned, it metonymically evokes thoughts of killing, and Macbeth’s words clarify the rest. Another such instance is the blood spot on Lady Macbeth’s hand, which stands metonymically for the murders she committed and thus also for her guilt, because, like the aftermath of the crime, it does not go away. What is more, the hand is commonly used metonymically to represent willful participation (as in have a hand in something, give someone a hand), and thus the spot also points to Lady Macbeth as the instigator of the killings – a memory she wants to erase. She tries to wash off the blood, addressing it in the now familiar construction (“Out, damned spot! out, I say!”), and thus we know she is haunted by remorse and horror of her crimes, but also afraid of being discovered – in her hallucinatory state, she is not aware that others cannot see the spot. Both the dagger and the spot are products of the characters’ imaginations, and thus are visible only to them. Other characters cannot see them and the audience does not have to, because the discourse patterns profile their presence with sufficient salience. As in the other cases, speaking to an object, even an imagined one, is used to prompt for blends which profile the speakers as experiencers of emotions and thoughts (wish to kill, fear, guilt, et cetera). Ghosts on the stage are interesting blends. The ghost is assumed to be a disembodied spirit of a person who is now dead, and yet it has to appear on the stage in some embodied form, so that it can be appreciated as a ghost and yet seen by the audience, and possibly the characters. Turner (2004) discusses the use of death and ghosts in Henry the Sixth, Part One, and in Hamlet, relying to some degree on the concept of ‘ghost physics,’ as proposed by Talmy (2000). Wetmore points out (2008) that ghosts appear quite often in Early Modern revenge tragedies and compares their function with the role ghosts play in Japanese nō plays. As he argues, Early Modern ghosts are not really agents in the play: they come to call for revenge, reveal crimes and their perpetrators, and sometimes also witness the acts of revenge to ascertain that their bidding has been done. Japanese nō ghosts, by comparison, come to tell their whole story and often act as avengers themselves – they can be narrators, but also have full agency in the play. This fits the description of the ghost of Old Hamlet. First of all, he, like all the ghosts considered by Wetmore, is visible to everyone – he is thus much more of a participant than just an instigator of revenge. He tells the part of the story which took place before the events on the stage start developing, so he does some crucial narrating. Thus this ghost connects the past space of the story with its present, and provides motivation for the plot as we watch it. There is also an interesting difference in how the bodies of the ghosts are represented. Western ghosts are eerie but not necessarily scary, while, as Wetmore suggests, Japanese ghosts come in full embodied shape, either the way they looked when they lived, or the way they looked when they died (and thus possibly disfigured). But they also can appear in more than one form. As Surma points out
6.3 Materiality of the stage and fictional minds
163
(2009), a nō ghost can truly exist in both the otherworld and the real world, and can choose the form it appears in, depending on the role the scene plays in the story. However, due to the highly stylized nature of the poetics of nō, the choices of how a ghost will appear can also guide the viewers. For example, in the play Atsumori, the ghost of the main character first appears in the shape of a simple man, a grass-cutter. The grass-cutter carries a flute, which was in Atsumori’s hand when he died, so the flute is an anchor to the slain youth. Even though ghosts have to be played by masked actors to be distinguished from mortals, the grass-cutter is unmasked. However, his appearance will still be recognized by savvy viewers of nō as being only a disguise because he would be played by a high-ranking (shite) actor. This same actor would also play the ghost of Atsumori in his true warrior form, which appears in the second half of the play. As this example shows, nō uses various anchors, including the actor’s body, the actor’s position in the world of theatre, the mask, and material objects to project some of the information about the role being played. Crucially, the mask defines a character as a certain type (say, a noble youth, as in the case of Atsumori), but does not allow emotions to be expressed through the actor’s face – all is done through words and actions. Ghosts in nō are clearly participants in both the present and the past of the play – they exist as if in both worlds, and are fully embodied, and thus visible to all. By comparison, the ghosts in Early Modern revenge tragedies have very little agency, but they also link the present of the play with its past – it would thus be accurate to claim that one of the roles of all vengeful ghosts is to construct a past narrative space which serves as the viewpoint space for the events of the play’s present. Moreover, as Surma points out (2009), ghosts in nō also serve as bridges between the real and the unreal, which coexist in some sense. They come back from the past to tell their story, but can also be agents in the story’s present. Later, in kabuki theatre, the presence of a ghost on the stage can be signaled by fireflies, but otherwise a ghost may look like an earthly human being. Alternately, it may wear make-up in colors which signal death (an equivalent of a nō mask), which enhances the material aspects of the play’s meaning. While the vengeful ghosts such as Old Hamlet may be more common in Early Modern drama, they are not the only kind. One good example of the different type of apparition is Banquo’s ghost, coming to sit at Macbeth’s table during a feast. This ghost comes from a very recent past, as Banquo has just been murdered, but he does not tell his story to anyone and does not seek revenge. He appears to Macbeth, and Macbeth alone, so that he is visible to the audience but not to other guests at the table. Macbeth shows signs of fear and distress, but nobody knows why. Banquo’s ghost is thus a material form similar to Lady Macbeth’s blood spot – an embodied token of Macbeth’s guilt. One can ask, then, why he is the first ghost to appear in the play – there have been earlier murders and the victims could have appeared to haunt the killers. If the ghosts
164
Fictional minds and embodiment in drama and fiction
were kinds of projections of feelings from the characters’ minds, there would be no reason to wait till Banquo dies. But Banquo’s ghost appears towards the end of the play, when it is time for everyone to start wondering what is bothering Macbeth. He appears at the banquet, not in Macbeth’s sleep (as the blood spot) and not in the privacy of his being alone (as the witches). I argue that Banquo needs to appear not just so the viewers can appreciate Macbeth’s distress, but so that his friends and courtiers can. Furthermore, Banquo takes the seat at the table which has been reserved for Macbeth, another visual and material means of reminding Macbeth that his royal seat will not be his for long. Literally and metaphorically, Banquo takes the royal seat away from Macbeth. As the examples suggest, ghosts may play a variety of roles in the play, but what they share is an embodied form which can occupy a specific position in the narrative, including prompting access to past narrative spaces, never represented on the stage. They are also perfect means of evoking revealing behavior of other characters, and prompting discourse patterns which allow characters to freely express their thoughts. Alongside other material objects and presences on the stage, they are catalysts for on-stage demonstrations of hidden thoughts and feelings. Overall, discourse of the play performs most of the functions later identified in novels, albeit in stage-specific forms. There is clearly a place for narration on the stage, as well as for direct and indirect speech and thought representation. In other words, when characters speak on the stage, they are not necessarily speaking only to other characters within the story space – they also speak to the audience, and sometimes speak to the audience via material objects and bodies as purported addressees. Material aspects of theatrical space are exploited to profile subjective construals beyond the characters’ words and play a central role in prompting story construction processes. The reliance on materiality has a clear advantage over relying on discourse alone, since supporting psychological components of the story with speech and objects conveys information to the audience through multiple channels. Also, stage time is allocated to events based on their emotional impact, not verisimilitude. The meaning of the play emerges as a blend prompted by textual and material means, and this close correlation between text and materiality is genre-specific. 6.4
From dramatic narratives to novelistic narratives
As section 6.1 proposes, literature can be seen a special case of a deictic set-up, where the reader is receiving communication via the text and can rely on the authority of a narrator. At the same time, it is also a complex intersubjective construction, wherein the reader is given access to the lives and minds of characters. In this sense, a literary text is also a site of joint attention (Tomasello 1999, 2008; Oakley 2009; Tobin 2006; 2010), so that the effect of reading is achieved
6.4 From dramatic narratives to novelistic narratives
165
through the text or the narrator drawing the reader’s attention to the aspects of narrated situations. Similar claims can be read from Bakhtin’s concept of dialogic imagination. All of these concepts support the idea that literature, being essentially a linguistic form, relies in its specific ways on the broad conceptualizations governing any exchanges of utterances. In the preceding section I looked at the discourse of drama. Ostensibly, it is the closest to the deictic set-up wherein there needs to be a shared ground and one has to see something happen or hear it said to receive the message. But, as I argued above, even such a presumably unambiguous set-up may be used in interesting ways to communicate mental and emotional states without them being directly verbalized. Even if the linguistic constructions suggest otherwise, however, the pretence of the ‘you have to hear it or see it’ communication is retained in the structure of discourse, assigned to different characters. In prose fiction, the pretence of the deictic ground does not disappear too quickly. One of the telling examples of the transition is La Celestina, a 1499 Spanish ‘novel in dialogue’, by Fernando de Riojas. La Celestina is considered a literary achievement (some claim it is the first Western novel) mostly because of its interesting characters and complex plot, but it is remarkable from the point of view of the topic of this section in that it is a bizarre mixture of drama and prose. The text is divided into acts, though there are sixteen of them, more in the spirit of contemporary chapters. Characters speak as they usually do in a play, but their turns can be pages long and include very good examples of narration and report. While much of what is interesting about the text is its novelistic spirit, what is said has to be said by one character to another. The first-hand access to information is also what seems to underlie another early narrative form – the epistolary novel, such as Richardson’s Clarissa. While much later than La Celestina (it is dated 1748), the novel is also infinitely more complex in its representation of the characters’ feelings and intentions. Its central narrative form is the strand of letters between innocent Clarissa and Lovelace, the rake of the story. While what she says is to be taken as a direct representation of the heroine’s thoughts and feelings, his text is meant to deceive Clarissa into accepting his advances. Zunshine (2006) offers a detailed analysis of how these exchanges can be construed from the point of view of the ‘theory of mind,’ but it is equally important for the understanding of the novel to note that the very format of the epistolary work assumes that the reader also gets first-hand access to the character’s words, and processes them as originating in the mind of the character. The epistolary formula backgrounds narratorial choices and downplays the role of the SV-space in the novel, offering instead an illusion of direct communication of thoughts and feelings, a variety of Egoviewpoint. Of course, the letters also have to contain all the narrative elements (so that the reader finds out what the plot was), but a text like Clarissa seems to represent two needs, very salient at the time: the need to tell the story, but also
166
Fictional minds and embodiment in drama and fiction
the need to represent the characters’ thoughts and emotions. And what could be more reliable than the voices of the characters themselves? As it turns out, Lovelace deceives Clarissa, though not the reader, but the narrative has now clearly established its important goal – representation of characters’ minds. The need to see the mind at work is further elaborated in the so-called itnarratives – stories told from the point of view of a material object or a pet (a lapdog, a guinea coin, a coach, et cetera).8 As Lupton suggests (2006), this ephemeral literary form, very popular in the eighteenth century and then gone, reflects the emerging interest in consciousness as the topic of literary production. From the point of view of the model explored here, it-narratives constitute a very interesting viewpoint set-up – with an object or animal taking the role of the character, so that the frequent changes in its ownership and location can be used to profile more narrative spaces. However, what it-narratives add to the discussion in this chapter is the observation that material objects can naturally be used as anchors to narrative spaces and can have human perceptions and emotions projected onto them. The connection between Juliet’s dagger and the story of a coin seems tenuous, but it relies on a cognitively salient pattern of projection of mental phenomena onto non-human participants on the narrative ‘stage.’ The eighteenth century is also when the intrusive mode of storytelling is common. Again, it seems that the presence of the on-stage narrator derives from the expectation that a novel is a kind of witness report, even though the illusion of the narrator as actually being able to know all the information reported is not being seriously upheld. However, it seems possible to argue that drama starts a narrating model wherein all the relevant information has a deictically identifiable source. That source is a direct, on-stage participant, who can be narrating, commenting, or genuinely involved in a conversation with another participant, but the information is provided first-hand. The access to participants’ minds is not postulated beyond what the participants themselves talk about, but attempts are made to use material aspects of the narrative space (whether dramatic or novelistic) to relay the information on mental states which are not being communicated directly. Clearly, narratives are naturally interested in the minds of characters, but the stylistic means of making mental states accessible to the reader are initially more reliant on material anchors. However, we need to acknowledge that the techniques in all cases involve the emergence of specialized linguistic constructions. These constructions develop, especially with the emergence of modernist techniques, but the path to interior monologues or free indirect discourse starts with the vocative-cum-imperative forms and intrusive narration. Even after the novel is well developed, authors may resort to theatre-like forms, as in the case of Melville’s classic Moby-Dick, published in 1851. The narration of Moby-Dick is truly multidimensional, with frequent excursions into
6.4 From dramatic narratives to novelistic narratives
167
nonfiction, longish treatiselike fragments, embedded shorter stories, and even a part which looks like it has been taken out of a play. The style is so varied and original that it would require a longer study to do it justice, but I want to mention just two fragments in which representation of the characters’ minds takes a truly innovative form. One of the turning points in the novel is the scene in which Captain Ahab announces that the Pequod’s expedition has a special goal – that of killing the white whale (rather than just finding any whales to hunt). He communicates his passion to the crew and receives their pledge of support. The goal is still controversial and dangerous, though, and the reader can expect that the men on board may have different emotional reactions to what has happened. At this point the narrator, Ishmael, ostensibly just one of the crew but also the novel’s primary voice, steps back to let others speak. It is narratologically not an easy task, since the novel is essentially the first-person narration of an eyewitness. Melville thus resorts to a theatrical setting for the duration of four short chapters, named Sunset, Dusk, First Night-Watch and Midnight, Forecastle. These chapters are thus overtly covering one evening, right after Ahab’s passionate announcement. In Sunset, the first-person voice is given to Ahab, in Dusk it is given to first mate Starbuck, and in First Night-Watch to second mate Stubb. Finally, Midnight, Forecastle represents all the mixed voices of the crew, in the form of a play. All four chapters contain ‘stage directions’, including rather traditional ones, such as (Aside). Essentially, the first three of these chapters are internal monologues of the main officers of the Pequod, reflecting on the situation. The chapters represent reflective thought only, and each of the characters represents a different emotional attitude. The viewpoint of the narrative here is almost modernist in its complete shift to Ego-viewpoint in each case and in the way in which the narrator’s role is reduced to the ‘stage directions.’ However, these stylistic choices are clearly not just narratorial experiments – they have a very specific role to play in the development of the plot. Elsewhere in the novel, mental states or participants other than the narrator are not central to the telling of the story, but the dramatic scene in which the crew come to share Ahab’s obsessive goal requires some stepping back and taking in of the consequences of the emotional shift. When the ‘faux stage directions’ return at the end of the novel, it is again to announce that a character (Ahab, or Pip) is either left alone on the ‘stage’ or speaks to himself (and the reader, but not other characters) – again, the formula is a kind of monologue. The theatricality of this narrative solution is somewhat artificial, but it effectively takes the floor away from an on-stage narrator and gives full voice to a character for an extended turn, which could not have been handled with dialogue alone. Crucially to my argument, the need to represent thoughts and feelings is what prompts this narrative innovation, and the most natural solution involves giving unmediated, first-hand voice to the character. The ‘seams’ are showing, but the goal has been achieved. The role of the hearer in
168
Fictional minds and embodiment in drama and fiction
the theatre has been blended with that of the reader of a novel, and the goal has not changed. In this section, I have focused on the role of supernaturally embodied narrative constructs in the framing of the minds of characters. The next section briefly considers the recent emergence of texts which rely on the body in more natural terms, while also focusing on the mind. 6.5
Fictional minds, bodies, and brains
While earlier literary narratives use bodies and materiality as inputs into mindconstructing blends, some recent texts display different tendencies. On the one hand, bodies are directly used metaphorically as representations of the characters’ minds; on the other, there are numerous texts featuring characters with various disorders related to unusual or deficient brain activity. The latter tendency is particularly interesting, as it suggests that literary production remains sensitive to the issue of the representation of minds and uses recent knowledge in neuroscience as inspiration. Rushdie’s Midnight’s Children is an excellent example of how bodies are used in contemporary texts. Some of the main characters are described as having particular and much exaggerated bodily features, which the text aligns with their psychological states. The main character, Saleem, has an unusually large nose, always congested and runny. But this very feature of his anatomy is what makes it possible for him to communicate telepathically with other children born on the cusp of India’s independence; once his nasal passages are clear, he loses his gift. His mother, Amina, hides her continued love for her first husband, but when she starts seeing him again in secret, she also starts suffering from awful corns on her feet; naturally, the corns go away when she breaks the relationship. Also, Saleem’s counterpart and enemy, Shiva, is distinguished by his unnaturally large and protruding knees, which are a mark of his brutal strength. Other characters are also marked by their physique – Naseem’s warts, Padma’s body hair, Jamila’s voice, et cetera. While all these narrative choices can easily be subsumed under the novel’s magic realist style, these bodily features become central to the narrative as metonymic for the characters’ special roles and gifts. Examples of what I refer to as ‘embodied characterization’ are common in other contemporary texts. The main character in Coetzee’s Slow Man has his leg amputated after an accident, but his limited ability to move is also aligned with his mental state, which prevents him from moving beyond the tragedy that befell him and trying a different life. Also, the main character of Günter Grass’s novel The Tin Drum, Oskar Matzerath, refuses to go on growing at the age of three, and remains physically, though not mentally, a child for many years. This embodied refusal to participate in the society in which he finds himself qualifies
6.5 Fictional minds, bodies, and brains
169
the novel as a form of magic realism, but also foregrounds its primary themes throughout the text. The heavy-bodied character of E. Annie Proulx’s The Shipping News is also not too agile mentally; the two clueless lovers in The Blind Assassin are cross-linked to another pair of lovers – she mute, he blind, et cetera. We can also note a recent emergence of numerous novels using characters with various mental and neurological disorders. To mention just a few, there were main characters suffering from Tourette’s syndrome, Asperger’s syndrome, Capgrass syndrome, Huntingdon disease, and de Clerambaut syndrome. All these texts, which I have dubbed ‘neuro-literature,’ inspect the human soul (and build good stories) through the functioning of the brain, rather than via psychoanalytically defined inner conflicts or postcolonially confused identities. They try to reach into the very embodied core of the human mind and human emotions. I consider this narrative tendency to be the new form of the literature of embodiment, informed by current advances in neuroscience and the flood of information on the brain pouring off the pages of books for the general public and popular magazines. Literary production, I argue, has strived to give readers access to the minds of characters since its inception. The means have relied on the body and the material world all along, but have gradually adapted both linguistically and psychologically. At the linguistic end of the spectrum, the old faux imperatives were substituted with more and more intricate constructions of speech and thought representation, including achievements such as the stream-of-consciousness techniques or the form of free indirect discourse. At the same time, the body regained its space in our attention, as a crucial means of shaping our mental life and giving it an expression. Also, the general knowledge of what our mental and emotional life depends on has led to literary solutions exploring the consequences of what neuroscience has taught us. To sum up, many of the changing literary conventions I have looked at occur as improvements in the representation of fictional minds. These improvements rely on our folk theories of how other minds can be accessed, such that are available at any given time. Early on, human thoughts are primarily accessed through embodied actions and self-reflexive linguistic expressions, but, as time goes by, providing ‘access’ to other minds becomes an important function of literature, one which can be identical with plot construction. The fictionality of the minds portrayed creates the kind of focus and time compression which real interactions do not provide, and thus literature gradually develops linguistic and narrative constructions which allow story-time to be allocated in ways which help represent the minds, rather than events. From the narrative perspective, direct textual access to the minds of characters has become a norm; intrusive narration, with its overt assumption of shared deictic ground, is now mostly used in literature for children and humorous texts. It is assumed that some linguistically negotiated access to the mind is possible,
170
Fictional minds and embodiment in drama and fiction
even a direct access, as modeled after modernist texts, especially stream-ofconsciousness ones. Interestingly, these narrative forms, which one could refer to as ‘direct thought,’ are often linguistically indistinguishable from ‘direct speech.’ Literary style seems to be redefining the role of constructions such as direct speech into tools of close representation of thought. In Chapter 7, I will consider the emerging constructional peculiarities of the speech/thought contrast.
7
Speech and thought in the narrative
Can Rita and I blow up the post office? (American Pastoral, Philip Roth)
In this chapter I will consider a range of literary language forms in which the formal distinction between speech and thought is used in different constructions specific to narrative discourse. Narrative discourse is typically treated as a discourse genre independent of colloquial usage, although, as I argued in Chapter 6, there are conceptual similarities, even if accompanied by formal differences. Further distinctions within the broad category typically address the difference between generic narrative purposes (pure narration) and the representation of discourse attributed to specific characters (Speech and Thought Representation, or STR); the latter are in turn talked about as direct discourse (as in He said, ‘I like it’), indirect discourse (He said that he liked it) and free indirect discourse (He liked it). The literature on these constructions is extremely rich, with much of the work recently brought together in Vandelanotte (2009), but since the issue is beyond the scope of my present concerns, I will only discuss selected instances of uses pertinent to the issue of viewpoint and story-construction. Above, I only used the term ‘discourse,’ without distinguishing speech from thought.1 In fact, I will be arguing below that not only is the distinction not salient enough, it is also misleading. As I have shown in Chapter 6, linguistic forms specific to speech have been co-opted into the representation of thoughts and feelings quite early on. In fact, these forms themselves suggest that the nature of all literary forms is such that, being based in the language resources, they need to rely at least partially on available constructions to express meanings which are not the core of the constructions’ semantics. In drama, indeed, speech, and more specifically speech profiling a potential addressee, is a common choice in representation of thought and feeling. Languages rarely (if at all) use constructions which formally distinguish speech from thought, since what matters most is the attribution of content to a specific character and thought is often communicated through speech. The crucial point is not whether any given content was communicated (speech) or not (thought), but how to distinguish the literary representation of 171
172
Speech and thought in the narrative
communicated or communicable content from the state of a character’s consciousness. For example, an actual perception of warmth and sunlight, registered by someone’s consciousness, may never need to be verbalized, but in a novel its representation may take any number of different forms, including a description of the weather (It was sunny and warm), a description of a character’s bodily experience (She felt the warmth of the sun on her face), and an STR construction (How nice and warm, she thought/said, turning her face towards the sun). Any of these choices (and many others) would perform the function of letting the reader know what the character experienced, but the issue of STR only becomes relevant when the perception is communicated (speech) or consciously formulated in the way in which it could be communicated (what counts as ‘thought’ in STR).2 In Chapter 4, I discussed narrative forms which use the pretence of visual perception to represent conceptualizations. In this chapter, I will look at a related phenomenon, wherein speechlike constructions in novels clearly play the role of the representation of the characters’ minds, without being mediated through the viewpoint of a narrator. In the final section, I will compare them with selected examples of more typical uses of STR. 7.1
Types of discourse spaces in the narrative
Relying on the concept of mental spaces and their special case, narrative spaces, offers some possibilities for a treatment of narrative discourse which does not rely entirely on the opposition between narration and STR. It is based on the assumption elaborated earlier in this book that narrative spaces constitute a multileveled network, with each space marked by a viewpoint (such as an SVspace, the MN-space, or an Ego-Viewpoint space). When a narrative space is marked with a viewpoint of a character, all the structure of the space is accessed that way, regardless of the linguistic realization of that space (whether it is structured more like speech, like thought, or like narration). As we have seen in Chapter 6, a narratorial experiment such as Melville’s theatrical chapters may blur these categories and blend them into one. Each of the three Ego-Viewpoint chapters in Moby-Dick is at the same time an instance of speech (since it is framed in theatrical terms), of thought (because it represents a character’s reflection on the events), and, to some degree, of narration (because the viewpoint is compressed up to the SV-space, and participates in the construction of the story of the hunt for Moby-Dick). Such instances are common in novels. It matters much less whether a given space is structured as a specific discourse type, but it does matter which viewpoint is represented and how the space participates in the overall storyconstructing blend. In (1), Anne Tyler’s character Delia is preparing herself mentally for a job interview:
7.1 Types of discourse spaces in the narrative
173
(1) Good afternoon, she would say. I’m here to ask if . . . No, not ask. Ask was too tentative. (Ladder of Years, p. 94)
The italicized fragments in the first line participate in several narrative spaces. As speech, they are lines in an imagined future conversation Delia is planning. As thought, they belong to the space of Delia’s plan for the interview the next day. And technically speaking, along with the rest of the fragment, they are also parts in the narratorial rendering of Delia’s thoughts about the interview. However, the narrator’s intrusion is limited to she would say, which also reports the content of Delia’s plan as she rehearses it to herself, and is thus itself an example of free indirect thought, along with the second line of (1). What is more, ask in the second line of (1) belongs to a metalinguistic space in Delia’s thought (as reported by the narrator), and is a quote from the imaginary conversation in Delia’s mind. It seems that the crucial distinction is between Delia’s reflective space and the imagined conversation space, and not in making the (difficult) decision as to the allocation of Good afternoon to speech or thought category – in fact, it is both, depending on the space whose viewpoint we are considering. Also, ask represents two different ‘thought’ spaces – Delia’s plan and Delia’s metalinguistic evaluation of the word choice. Finally, temporal and referential compression is what obliterates the categories (as well as allowing for a consistent choice of tense and reference), so that in fact the distinctions are interpretive only. The viewpoint really matters here, but the speech/thought distinction not as much. Examples of speech in imagined conversations are not uncommon, and participate in very interesting construals. The fragment in (2), from Eggers’s novel, is a part of a complex narrative space network. The main character, Dave, is driving to meet a girl while also worrying about the fact that he has left his little brother, Toph, with a babysitter. He imagines a scenario wherein the babysitter is a murderer, and kills the boy, whereupon Dave is accused of parental negligence: (2) There will be a trial, a show trial – How did you meet this man, this babysitter? We found a posting, on a bulletin board. And how long did your interview of him take? Ten, twenty minutes. (AHWOSG, p. 126)
The conversation (between the prosecutor and Dave?) is imagined as a deposition in the trial in which Dave (not the babysitter!) is the defendant. The entire court scenario is a narrative space embedded in the train of thought starting with the imagined murder. It is again both direct speech and free indirect thought, depending on which viewpoint is adopted. But primarily, what is being elaborated is the ‘emotional’ space of Dave’s paralyzing though unreasonable guilt, woven into every bit of the narrative he constructs.
174
Speech and thought in the narrative
The approach which distinguishes narration from various forms of STR has much to contribute to our understanding of viewpoint phenomena. Later in this chapter I will consider some examples to show how the framework being developed here treats the important grammatical effects of STR (tense choices, pronominal choices, et cetera). First, however, STR phenomena must be seen in the larger context of a fictional narrative, including the way reported content is embedded in higher spaces and the complexity of such embedding/embedded relations in the context of the entire narrative structure. 7.2
Speaking for thinking
It is generally true that narratives use similar forms to represent speech and thought, especially in the ‘reported’ or ‘embedded’ cases. What is less talked about, however, is the forms which are appropriate to the representation of ‘direct thought,’ as opposed to ‘direct speech.’ The shared understanding seems to be that the term ‘thought’ refers to the ideas in a character’s mind which are not structured in a dialogic form of exchange of utterances, and are thus not addressed to anyone in particular. As I will show below, however, direct thought can rely on the dialogic structure of conversation, profiling various viewpoints as discourse participants. I will refer to this narrative technique as speaking for thinking. The question of a potential addressee is crucial to this discussion. In the examples analyzed in Chapter 6, thought can take dialogic form because the text profiles a material or imaginary addressee – hence the constructions in which material objects, bodies, or mental images are being spoken to so that the speaker’s thoughts and feelings can be expressed and actions identified. But in the context of a novel, dialogic structure typically constitutes only a portion of the text, and then the speaker and the addressee are characters, situated in the context of the same ground, as emerging from the story being constructed. Moreover, the reader may be an off-stage addressee of all this, but becomes an on-stage addressee only in the case of intrusive narration. However, it is possible to treat much of what would qualify as ‘direct thought’ as a one-way dialogue similar to narration, though with an important difference. When the narrator is on-stage, the viewpoint of the SV-space is directly accessible to the reader. But when a character’s thoughts are communicated, they are relayed through an off-stage narrator. In an example like (3), from Roth’s American Pastoral (discussed in some detail in Chapter 5), discourse structured similarly to conversational discourse is presented through the filter of the SV-space viewpoint and construed as though the Ego-Viewpoint here were a part of a conversational structure in which the person asking the question also answers it – an effect which is possible also in colloquial discourse.
7.2 Speaking for thinking
175
(3) Had there ever been a happier child? A less destructive child? A little signorina any more loved by her mother and father? No. (American Pastoral, p. 229)
The questions are rhetorical, as is in fact the answer, since the point of the fragment is to express the amazement at the discord between the pampered child of the past and the teenage terrorist she became. The text in (3) reflects a shocked and pained reasoning of the father, as reported by a narrator. The narrative space structure of the novel as a whole includes a number of levels of embedding – from the on-stage narrator, Nathan Zuckerman, who runs into his high-school friend, through Zuckerman’s attempts to find out more about the Swede’s life, down to his complete identification with the character whose story he is telling, where (3) belongs. The example thus represents plausible thoughts of the Swede, as they are imagined and reported by Nathan. Structured as speech, they are clearly thoughts; formed as questions, they are expressions of shock and disbelief. But Zuckerman’s narration, which dominates the entire novel, is indistinguishable from an off-stage narration with a zero-narrator, that is, a narration where the viewpoint of the character is the only one represented. Zuckerman quickly moves away from acknowledging what people tell him in conversations, and yields the ground to his imagined experience of the Swede’s life. While much of the text represents the Swede’s thoughts and experiences, mostly from the third-person perspective, constructions of represented thought are rare and fragments like (3) are used instead. They are a signal of viewpoint compression, prompted by the constructions typical of direct speech, and rendering the raw experience of the character. What seems to dominate in such cases is the ‘directness,’ miming first-hand access to the thoughts, rather than marking any narratorial presence. But the possibility of such a choice of expression is based in the mechanism of viewpoint compression, not STR. A similar degree of compression, across different spaces, can be found in (4). Having shifted the viewpoint into the space of the Swede’s mind, Zuckerman recreates the man’s memories of his daughter’s childhood. They are brought in, as in (3), to underscore the desperate state the family is in after Merry’s disappearance. The context of (4) is the Swede’s reflection on Rita, who presents herself as Merry’s terrorist friend and torments the Swede with veiled threats and false promises. Thinking of Merry’s friends, he remembers Merry and her childhood friend Patti, and the space is elaborated through questions of the kind Merry could have asked at the time. Again, there is a pretense of direct speech, while on a higher narrative level these are the Swede’s thoughts, imagined by Nathan. (4) Mom, can Patti stay for dinner? Mom, can Patti wear my boots? Mom, can you drive me and Patti to the village? (American Pastoral, p. 321)
Later in the fragment, thoughts about Rita are set against the memory of Merry’s childhood.
176
Speech and thought in the narrative
SV-space
Narrator: Zuckerman present
MN-space: The Levovs Roles: Mother, Father, Daughter, Daughter’s Friend NSI – Merry’s childhood; past Daughter – Merry / child Daughter’s Friend – Patti Childhood activities
NS1 – the Rimrock Bomber; present Daughter – Merry / terrorist Daughter’s Friend – Rita Terrorist activities
DS 1 / 2 / 3/...
Imagined DS
Speaker: Merry
Speaker: Merry
Asking permission to do things with Patti
Asking permission to do things with Rita
Figure 7.1 Discourse blend (examples 4 and 5) (5) Can Rita stay for dinner? Can Rita wear my boots? Can you drive me and Rita to the village? . . . Can Rita and I blow up the post office? (American Pastoral, p. 321)
The blend in (5) builds on the generic role of ‘Merry’s friend’ (see Figure 7.1). The ‘Patti’ input in (4) belongs to the past of the story, and explicitly evokes an addressee (Mom). The ‘Rita’ sequence of questions builds on the frame established in (4) and substitutes a young terrorist in the ‘friend’ role, which changes the nature of what the girls would now want to do. The blend belongs to the present of the story, and is crucially built on disanalogy – the innocent childhood activities are supplanted by dangerous acts of terrorism, and yet, in the Swede’s mind, Merry is still his beloved child, seeking the parent’s approval of her friend. The lack of form of address in (5) suggests that the Swede may be imagining himself as an addressee of these requests, which underlines his dramatic situation of a father who has to accept the daughter’s friend if he wants to continue accepting the daughter. The ‘speech’ pattern in (5) is thus not identifiable as speech at any level. It is, instead, a representation of a mental state, that of having to deal with irreconcilable frames. Crucially, Roth’s narrative choices completely blend actual facts with imagined experience, and minimize the role of the narrator. In American Pastoral, Zuckerman is not just a narrator, he is more like a medium who gives voice to the Swede’s story. Viewpoint compression is the primary narrative technique here. What is also interesting about such examples is that the fragments which mimic conversation may not only be imagined, but also generic – it is hard to say whether a sequence such as (4) was ever uttered and whether it is this particular occasion that is remembered. It could also be intended as frame
7.2 Speaking for thinking
177
metonymic – so that it evokes the frame of ‘childhood friend activities,’ which the discourse in (4) is representative of. Similar examples of quotations used as ‘selective depictions’ have been discussed in Clark and Gerrig (1990) to refer to instances where a particular stretch of discourse is selected to play the role of a depiction, not a verbatim quotation. What one might add to this observation is the point that such ‘depictions’ cannot play their role unless the selected quote has a frame-metonymic function, so that it evokes a frame in which the particular expression played a role, to use that frame in the construal of the situation.3 There seem to be numerous examples of direct speech expressions used metonymically, for example in Eggers’s novel. Example (6) comes from a fragment when Dave agonizes over the impression strangers have of Toph’s happiness. He is frustrated by a school photo, in which, in his view, Toph looks sad, and this makes him remember ‘the Phone Voice Problem’ – Toph’s habit of answering the phone in a weak and insecure voice. (6) And of course people wonder. What’s wrong with Toph? they ask me when he hands me the phone. And always I have to be cavalier. That’s just Toph! ha, ha. (AHWOSG, p. 322)
People can be expected to wonder about the happiness of a small boy being raised by a rather colorful character of a brother, but we are aware that the question reported in (6), as well as the answer, do not have to be quotes of speech fragments uttered on any specific occasion. Being generic, they evoke a frame wherein the ‘Phone Voice’ is indeed a ‘Problem,’ even though it may appear problematic only from the point of view of Dave’s exaggerated fears. As (7) shows, the discourse then continues into another generic pattern, of Dave’s attempts to change Toph’s habits. At the same time, the description of what the voice ‘sounds like’ is vivid and detailed and soon develops into what is clearly Dave’s childhood memory of being frightened by an overpowering parent: (7) So I’ve been imploring him to sound normal. Please sound normal, Toph, you are normal, we are normal so just sound normal please can’t you? Don’t sound like I’ve been beating you, like you’re in the bathroom hiding from me, because I’ve been there, have hidden from parents before . . . and I have hidden there, and have seen, darkening the white slit of light under the door to this closet, his shoes. (AHWOSG, p. 322/3)
It is not quite relevant to ask if this is speech or thought. It starts as a metonymic speech pattern, but ends up being Dave’s personal reflection on his own childhood, one which he probably never included in his actual conversations with Toph. There is a shared viewpoint, of Dave fearing being thought abusive, and building the fear on his own childhood experiences. That viewpoint space participates well in the overall narrative, and thus the status of specific mock-utterances remains unresolved.
178
Speech and thought in the narrative
The problem of the school photo which triggered the fragment on the Phone Voice Problem continues when Dave looks at the picture set again: (8) All these pleading Tophs, Eleven Tophs saying Look at my sad life, you people, you viewers of junior high pictures! . . . Save me from him, because every night before dinner he’s asleep on the couch and so dead to the world, and when he can’t get up he tugs on my shirt and begs, he makes me cook for us and then later . . . (AHWOSG, p. 325)
What the pictures are ‘saying’ is presumably a fair description of Dave’s routine behavior, not always in keeping with a ‘perfect father’ figure. The viewpoint compression here is quite complex: Dave’s guilty awareness of his behavior is the highest viewpoint here, but is attributed to Toph instead. Dave would not want to accuse himself of bad habits, as this would imply an attempt to improve, but he attributes the accusatory thoughts to Toph, while also worrying that other people will become aware of them too, and blame Dave. In the end, the confusion of viewpoints here means that Dave wants Toph to look happy, so that people don’t assume he is unhappy, so that in turn Dave can feel good in his parental/brotherly role. The fact that Toph is communicating unhappiness through the picture is an interesting construction – viewers may interpret the represented reality of the picture in this way, so the picture is ‘saying’ that it is so. In this representation blend, the viewpoint of the potential-viewer-cumaddressee is the one that Dave constructs, and then tries to respond to. But all these scenarios are playing out in Dave’s mind only, and are properly a part of his thoughts. As these examples suggest, networks of spaces in narratives may be structured by present and past events, memories, imagined scenarios, frame-metonymic prompts, representation blends, roles, et cetera. All these spaces are subject to viewpoint compression, so that we can see the relevance of any specific space set-up in the text to the story as a whole. But whether these spaces are linguistically structured as speech or not is not correlated with their place in the network and their viewpoint role. 7.3
Levels of embedding in thought representation
The examples above signal the fact that a network of narrative spaces may involve a number of levels of embedding, rather than a two-level structure profiling the narrator and the character. The depth of embeddings can then affect the way in which various forms of discourse will be interpreted. At the same time, thought-spaces may be set up not only from the point of view of the narrator (when they are reported as actually occurring in the character’s mind), but also from the viewpoint of other characters – so that we learn what characters think about each other’s thoughts. One such example is presented in (9), which comes from Philip Roth’s novel The Human Stain. The main
7.3 Levels of embedding in thought representation
179
characters there are Coleman, a retired college professor, and Faunia, his lover, a much younger woman with a very tragic and difficult past. Given how much she has been through, she can be expected to ruminate on her fate, and Coleman attributes just such thoughts to her: (9) He thinks she is thinking about how long it has all gone on, the mother, the stepfather, the escape from the stepfather, the places in the South, the places in the North . . . and maybe she is. Maybe she is even if, alone now on the grass while the boys are smoking and cleaning up from lunch, she thinks she is thinking about crows. She thinks about crows a lot of the time . . . and today they are cawing all over campus, and so instead of thinking of what she is thinking the way Coleman thinks she is thinking it, she is thinking about the crow that used to hang around the store in Seeley Falls. (The Human Stain, p. 164)
What the fragment makes clear is that thought does not have to verbalize clearly the actual issues considered; it may meander through different topics, while at the same time, deep down, it always revolves around the problem. How, then, are such thoughts to be reported? Roth’s narrative choices, again relayed through Zuckerman, are similar to those in American Pastoral, where the lowest-level Ego-viewpoint is compressed up to the narrative level. In (10), the train of Faunia’s thoughts is recreated: (10) That crow’s voice. She remembers it at all hours, day or night, awake, sleeping, or insomniac. Had a strange voice. Not like the voice of other crows probably because it hadn’t been raised with other crows. Right after the fire, I used to go and visit that crow at the Audubon Society, and whenever the visit was over and I would turn to leave, it would call me back with this voice. (The Human Stain, p. 165)
The narrative gradually goes deeper into Faunia’s thoughts. Starting with the present-tense third-person narration of the nature of Faunia’s thoughts (She remembers it), which follows the pattern of the rest of the narrative, through free indirect discourse past-tense report of the content of her thoughts (Had a strange voice.), to the sudden viewpoint shift when the text takes a first-person past-tense form, representing Faunia’s thoughts without the mediation of the narrator. Not only does this narrative choice open a lengthy internal monologue, but it also affects the entire structure of the novel, gradually compressing out all the levels set up earlier, and separating Faunia’s thoughts from the SV-space. It is thus in fact a shift down into the consciousness of one of the characters. Faunia’s thoughts revolve around the nature of crows, which she admires and seems to understand. She feels so close to the crow she once went to visit that she feels fully identified with it: (11) (. . .) why can’t I be a crow locked up in this body? Yeah, and where is the doctor who is going to do what they do to let me out? Where do I go to get the surgery that will let me be what I am? . . . I am a crow. I know it. I know it! (The Human Stain, p. 169)
180
Speech and thought in the narrative SV-space Narrator ‘Nathan Zuckerman’ he
Coleman thinks
Faunia thinks she (about her unhappy life)
Faunia thinks about her life by blending it with the life of a crow
she
Faunia thinks (about crows)
I I’m a crow
Figure 7.2 The ‘crow’ blend (examples 9, 10, and 11)
Overall, the fragment sets up a blend of narrative spaces which not only gives the reader access to Faunia’s thoughts but also elaborates on the mutual relationship between the main characters (see Figure 7.2). Two ‘thought’ spaces have been set up for Faunia – one in which Coleman imagines her depressed about her past, and the other wherein she is indeed thinking about it, but in a different way (instead of thinking of what she is thinking the way Coleman thinks she is thinking it, she is thinking about the crow). She rethinks her life through her bond with crows and finally imagines herself being one of them, even if locked into a human body. Without acknowledging it, the narrator is constructing a blend in which different concepts in the minds of two different characters can yield a similar meaning, even though Coleman clearly underestimates the complexity and depth of Faunia’s trauma. It is achieved through the additional level of mental space structure – he ‘thinks’ she is thinking, while she ‘is thinking.’ This narrative choice manipulates viewpoint in interesting ways, and also allows characters’ thoughts to be expressed and work jointly in the narrative. Roth’s narrative technique is worth a pause. In the two novels mentioned here, he starts with an on-stage narrator, the writer Zuckerman, whom his faithful readers know from earlier novels. As the narrator hears more from or about the characters, the narration becomes gradually off-stage, as the story being constructed comes into focus. In both cases, the texts go into lengthy internal monologues, until there is a complete shift of focus to the character’s Ego-viewpoint. But a careful reader remembers that all of the text is embedded under an SV-space in which Roth’s alter ego, Zuckerman, is in fact weaving it together from the facts structuring the SV-space (e.g. his role of a high-school
7.3 Levels of embedding in thought representation
181
friend of the main character), the stories heard, and finally those he constructs himself. It may not be Roth’s intention, but the successive descent into the character’s mind seems to be a perfect description of how fiction might emerge. The narrative role of direct speech can be somewhat complicated in the context of the network even when the expressions represent what has been said. Example (12), from Anne Tyler’s novel A Patchwork Planet, describes a scene where the main character and first-person narrator, Barnaby, comes home to find a message from his parents: (12) My answering machine was blinking . . . I leaned over and pressed the button. “Barnaby,” my mother said, “this is your mom and dad.” What a thrill. “We just wanted to say Happy New Year, sweetie . . . Call us sometime, why don’t you? Bye.” Click. I flopped back on my bed and looked up at the ceiling. (A Patchwork Planet, p. 43)
All the direct speech in this example is the message, and so represents what Barnaby hears in the MN-space, and only secondarily what his mother has said earlier, when she called. The viewpoint overall is thus that of Barnaby as the addressee of the message, which is further reinforced with his inner comments (What a thrill.) and further things he hears (Click.) Thus, even though it is a genuine example of direct speech, its embedding in the temporally later space of Barnaby’s return home contextualizes it differently; for one, if Barnaby had been home when his mother called, he would not have such a natural opportunity to comment sarcastically on his parents’ calling. In the fragment, all the viewpoints profiled in the phone call are embedded in the space of Barnaby listening. Space embedding, even when there are clearly two different sources of what is being said, does not have to take a marked constructional form. In the movie The Hunt for Red October the main character, Jack Ryan, talks to the National Security Advisor, Jeffrey Pelt, asking for his help. Pelt answers: (13) I am a politician, which means that I’m a cheat and a liar, and when I’m not kissing babies I’m stealing their lollipops. But it also means that I keep my options open.
It is clear that not all that Pelt says represents his viewpoint, and that he is evoking another discourse space wherein speakers other than himself describe politicians in negative terms. In spite of the fact that the ‘cheat and liar’ part of the utterance is naturally seen as echoic of some contextually available belief, this description would not suffice for two reasons: one, there is no indication of such an utterance being available in the context of the movie, and two, an echoic utterance is naturally described as embedded in the current discourse, even if
182
Speech and thought in the narrative
there are no constructional signals present. Thus, in spite of the fact that (13) represents an entry in a conversation, the example blends two viewpoints. Throughout the utterance, Pelt is referring to himself as I, without adding any hedges such as people think (I’m a cheat and a liar), but it is clear that other people’s thought-space is being evoked. The use of the first-person pronoun is thus a result of the compression of the ‘he must be a cheat and a liar’ viewpoint into the higher discourse space of the conversation, in which Pelt is the speaker, and thus refers to himself as I. The fact that (13) comes from a fictional dialogue does not invalidate the point. In colloquial discourse speakers often say similar things. When one reveals a secret but wants to remain anonymous, it is natural to say something like You never heard it from me, which puts the official ‘line’ into the mouth of the addressee, thus preventing him from identifying the source. Such cases also rely on blended discourse, as the full utterance would probably include an embedding phrase such as You mustn’t tell anyone (that you heard it from me). Embedding of discourse without profiling a complete STR construction is in fact more common than is generally thought. Recent work by Pascual (2002, 2006a, 2006b, 2008) describes a class of discourse blends which she refers to as ‘fictive interaction.’ The description applies to many formal constructions in which a discourse fragment that functions as a token of a discourse situation is incorporated into another utterance. Some of Pascual’s (2006a) examples are given in (14): (14)
a. You need to go in with the attitude that yes, I can do this . . . (allnurses.com forum) b. You’ll learn the winning attitude of YES, I CAN DO IT! (personal training website) c. Develop a “Yes, I can do it” attitude. (“Twice – Exceptional Students” Newsletter)
What is crucial to these constructions is that a discourse fragment begins to play a meaningful role of its own, and can contribute to another utterance by occupying various syntactic positions (a clause in 14a, an NP in 14b and an adjective in 14c). What fictive interaction shares with examples like (13) is the possibility of adding a second viewpoint to the viewpoint of the broader utterance, but what is different is the pronominal choice. In (13), the compression of the embedded space into the embedding one requires that the referential features and the ground of the higher space are preserved. In all the instances in (14), the pronoun I is preserved, while the utterance as a whole uses second person you (also implied in the imperative in [14c]). Switching to you in any of the ‘Yes, I can do it’ expressions would cause confusion or unacceptability, which suggests that the ‘token’ nature of the phrase requires that any person assuming the attitude will blend her/himself with the I, because the I is in fact not identifying
7.4 Viewpoint compression and constructional phenomena
183
any speaker in any specific discourse ground. Example (13) is crucially different in this respect, as the speaker blends his identifiable critics’ views into his own discourse ground. The examples throughout this section suggest that there are various possible embedding configurations in representations of thought, and that they may yield a variety of different constructions. While these constructions may not be well established as stable grammatical patterns, they all rely on viewpoint blends. They sometimes suppress the overt expression of embedding, but they also maintain a constant connection to the SV-space of the narrative. Viewpoint compression thus offers an explanation of a range of narrative choices intended to negotiate different viewpoints. 7.4
Viewpoint compression and constructional compositionality
The preceding section looked primarily at the expressions which represented thoughts without profiling any STR constructions. In most cases, the actual linguistic form was modeled after direct speech, even though the actual interpretation involved numerous levels of embedding. In what follows I will look at selected constructional phenomena related to STR. However, the range of constructions specific to STR is defined differently by different analysts, and thus I will not discuss such proposals here or attempt to view each from the perspective of my framework. Instead, I will use some very standard examples of STR to clarify further the mechanism of viewpoint compression, as it participates in textual representation of speech and thought. Example (15) comes from Anne Tyler’s novel Ladder of Years (see Figure 7.3). The main character, Delia, is being interviewed by Mr. Pomfret, her future employer: (15) Mr. Pomfret didn’t mention references. His sole concern was the nature of her past duties. Had she typed, had she filed, taken shorthand? . . . He said she would start tomorrow; her hours were nine to five. Sorry the pay was just minimum wage, he said . . . Also she was expected to brew the coffee; he hoped that wasn’t a problem. Of course it wasn’t, Delia said brusquely, and she rose and terminated the interview. (Ladder of Years, p. 95)
The fragment represents some salient forms of represented speech: indirect speech (He said she would start tomorrow) and free indirect speech (Had she typed, had she filed, taken shorthand?; her hours were nine to five.). It also features some less typical solutions, wherein free indirect utterances are accompanied by an additional he said (Sorry the pay was just minimum wage, he said). All these forms are marked with some standard shifts in grammatical forms, resulting from the shift away from the represented speaker’s and hearer’s deictic ground and an alignment with that of the narrator. Consequently, the pronoun referring to Mr. Pomfret as the speaker is shifted into he, while the you with
184
Speech and thought in the narrative SV-space
Narrator: 0 present MN-space Character: Delia (she) past
(he) (she)
Discourse space 1 Speaker: Mr. Pomfret (‘I’) Addressee:
Delia (you)
Figure 7.3 Third-person report (example 15)
which Mr. Pomfret addresses Delia changes into she; also tense forms are shifted from present to past. These changes (even if not always equally consistent) have been described in the relevant literature.4 Within the framework advocated here, the same facts can be accounted for naturally. The viewpoint of the SV-space in the case of (15) is the present, with an off-stage narrator, which accounts for all the characters in the narrative being referred to via third-person pronouns. Two of the characters in the past MNspace are Mr. Pomfret (he) and Delia (she); in (15) they are in conversation, and thus in a separate discourse space within the MN-space. In a discourse space, one expects its specific ground to be profiled, and thus the participants would refer to themselves and each other according to the rules: I for the speaker and you for the hearer. But there are two options available from here onwards. Either the discourse space is a direct speech space, and then does not need to be incorporated into the network other than through stable cross-mappings maintaining the identity of characters, places, et cetera, or it is to be linked with the higher space (via embedding, reporting, STR, et cetera). If the discourse space is integrated into higher narrative spaces, it needs to be coordinated with them with respect to viewpoint established by the SV-space. Thus if direct speech (i.e. independence of the discourse space) is maintained, the I/you forms remain restricted to the discourse ground of the conversation, and cannot be projected upwards. But if the discourse space becomes a part of the blend, it has to start using the basic deictics (pronouns and tense) as they are used throughout the network of the MN-space (he/she). The discourse remains a part of the MNspace in either case, but its degree of incorporation is different: from establishing cross-space mappings, to some degree of blending. The topology of the MN-space in (15) is embedded in the SV-space with no on-stage narrator. It is possible to argue that a zero-narrator (off-stage narrator),
7.4 Viewpoint compression and constructional phenomena
185
such as the one in Ladder of Years, cannot be cross-mapped to any of the characters and never requires the use of the form I. In order that the consistency of reference is preserved throughout the network, the unspoken I remains reserved for the position of the narrator, but there is no character in the lower space whose viewpoint could be compressed up to the SV-space, thus justifying the use of I in the MN-space. Discourse spaces in which direct speech is used are different, since they rely on their own ground, even if the characters from the discourse space are cross-mapped with the characters in the MN-space. Thus, even though Delia and Mr. Pomfret are profiled as characters in both the MN-space and the discourse space, and these spaces are cross-linked, they would use I/you only if their conversation were represented as direct speech, in an unincorporated discourse space. In (15), the discourse space is fully integrated into the MN-space, and thus the forms shift to he/she. Now, what about the form of STR known as indirect speech? Example (15) has one instance of it, He said she would start tomorrow. How is it different from the examples of free indirect speech in the fragment, such as Had she typed, had she filed, taken shorthand? and from the free indirect/indirect hybrid Sorry the pay was just minimum wage, he said? It seems that the he said expression in both cases participates in the MN-space, not the discourse space, since it belongs to the descriptions of actions performed rather than representing discourse of a character. The clause representing the words of a character, whether fully incorporated syntactically (she would start tomorrow) or not thus participates in the discourse space/MN-space blend. Both cases preserve much of the discourse style of the discourse space: there is the temporal deictic tomorrow, which is future with respect to the discourse space ground, not the MN-ground, and the emotive marker Sorry. Such local gaps in full viewpoint coordination signal the degree to which the blend is integrated, but do not change the nature of the process. To sum up, the MN-space has discourse spaces embedded in it (alongside or as a part of other narrative spaces). These discourse spaces are embedded in the MN-space but may be marked with varying degrees of integration. They may not be integrated at all, and then the form of choice is direct speech, with the I/you pronominal choice. They may be embedded via appropriate MN-space expressions (he said) and thus integrated to a high degree. Or there may be a more loose connection between the spaces, such that the pronouns and tense are typically integrated, but there is no MN-expression in which the reported sentences are directly embedded, and the expressive flavor of the discourse space is preserved. Theoretically, nothing prevents a given discourse space from being incorporated into the MN-space in any one of the three ways – example (15) uses only indirect and free indirect speech, but if the conversation were longer, one could imagine different parts reported differently. Contrary to what much of the
186
Speech and thought in the narrative
discussion of STR suggests, discourse spaces are embedded in the MN-space as spaces, not as individual sentences. It is not surprising that the choice of the STR construction may vary from sentence to sentence, because a constructionally consistent report (say, indirect speech only) would be stylistically unacceptable. But there needs to be minimal consistency as well – any form other than direct speech needs to be blended to some degree into the MN-space; at a minimum, the temporal topology and reference need to be coordinated. This does not prevent shifts (as in the case of [10] above), but it precludes random changes from one sentence to another. The story network functions differently when the story’s SV-space profiles an on-stage first-person narrator. Such a narrator is often cross-linked to a story participant who describes the events he or she participates in. Discourse spaces within the MN-space do not necessarily involve the narrator/character, and these behave roughly as in the off-stage narrator set-up. But when the narrator/character (N/CH) does participate in the discourse space, then in the instances where N/CH is the speaker the pronoun I is used. However, the I is then aligned not just with the discourse space in question, but also with all the narrative spaces in which N/CH is a participant. In (16), for example, Dave (in Eggers’s novel again) reports talking to his brother about his day at school and then quotes what he says to the parents of another child at school. He is consistently using I to refer to himself throughout: (16) “What happened today?” I ask. “Today Matthew told me that he hopes that you and Beth are in a plane and that the plane crashes and that you both die just like Mom and Dad.” “They didn’t die in a plane crash.” “That’s what I said.” Sometimes I call the parents of Toph’s classmates. “Yeah, that’s what he said,” I say. “It’s hard enough, you know,” I say. “No, he’s okay,” I continue, pouring it on this incompetent moron who has raised a twisted boy. “I just don’t know why Matthew would say that. I mean, why do you suppose your son wants Beth and me to die in a plane crash?” (AHWOSG, p. 89)
There are two (or more) discourse spaces here: one is the conversation between Dave and his brother, the other is the generic discourse space of conversations with parents in which he tries to make them feel guilty about their children. Whether Dave is the speaker in one space or the other, or whether he narrates (I ask, I say, I call, I continue), the pronoun I consistently links all his narrative space roles across the contexts. There is a cross-space mapping linking the narrator, the character in the MN-space, and the speaker in the discourse space, so that, via Access Principle,5 any one of these instances can use the pronoun I. Interestingly, this makes for a much more integrated narrative, since the
7.4 Viewpoint compression and constructional phenomena
187
presence of the first-person voice throughout connects the emerging blend very tightly via a shared viewpoint. Example (17) shows another instance, this time of Dave’s internal monologue, scattered as usual, but largely consisting of questions he is considering. (17) There is so much to do . . . I have to get a résumé together, and we have to find a new place to live when the sublet ends, and how will Toph get to school if I get an early job? Will Beth pull her weight, will she be too busy, will we kill each other? . . . How much should I/ can I/ will I burden Kristen? . . . Should I lighten my hair? (AHWOSG, p. 66)
What is central to the narrative space set-up here is that Dave is considering questions about his own possible future, so the pronoun I in each of the questions (such as Should I lighten my hair?) means that he is the person asking the question and also its addressee; this blended speaker/addressee viewpoint is further compressed up to his role as a narrator, making the narration very close to the core of the main character’s mind. Crucially, this narrative space set-up is different from an STR form recently identified and described by Vandelanotte (2004a, 2004b, 2009) under the name of DIST (Distancing Indirect Speech and Thought). Vandelanotte’s description of DIST compares it to FIST (Free Indirect Speech and Thought) and notes several differences. One directly relevant to my discussion here is the issue of the meaning of the pronoun I. FIST would involve uses like the one in (17), but DIST allows the first-person pronoun to identify the addressee, not the speaker in the discourse space being reported, but that addressee also has to be what Vandelanotte refers to as the ‘current speaker’ – which in the context of firstperson narration means the narrator. The DIST use of I is well illustrated in (18), which is a fragment from Henry Miller’s non-fiction account of his visit to Crete, The Colossus of Maroussi. He talks there to a Greek intellectual whom he calls Mr. Tsoutsou, since he has forgotten his real name: (18) He started off on Aragon – had I read Le Paysan de Paris? Did I remember the Passage Jouffroy in Paris? What did I think of St. Jean Perse? Or Nadja of Breton? Had I been to Knossus yet? I ought to stay a few weeks at least – he would take me over the island from one end to another. He was a very hale and hearty fellow and when he understood that I liked to eat and drink he beamed most approvingly. (The Colossus of Maroussi, pp. 115–116)
Miller, as the first-person narrator, is reporting a conversation in which Mr. Tsoutsou asks a lot of questions. All the reported questions use the pronoun I to refer to Miller as the addressee (you in the discourse space), while also marking him as the N/CH of the narrative. In this case, then, the pronoun choice is aligned entirely with the SV-space and the MN-space, but not with the discourse space – this seems to be a narrative space rendering of the nature of DIST. The same applies to the sentence I ought to stay a few weeks at least, as it
188
Speech and thought in the narrative SV-space
Narrator: ‘I’ present MN-space Narrator/character ‘I’ past Discourse space 1 Speaker (‘I’) Addressee/narrator (‘you’)
Figure 7.4 First-person report (example 18)
clearly renders the Greek’s suggestion, not Miller’s own determination. But in the final sentence, he understood that I liked to eat and drink (indirect speech), the text uses I in a more common reported sense, as referring to Miller’s own appreciation of his enjoyment of earthly pleasures, albeit acknowledged by Mr. Tsoutsou (see Figure 7.4). We can now compare the role of pronouns in the use of questions in the three examples above. Had she typed? is an example of free indirect discourse, or FIST, and the pronoun she substitutes you (as a marker of the addressee) while aligning it with the MN-space framed by an off-stage narrator. Should I lighten my hair? is also an example of FIST, and the I connects to the first-person onstage narrator-cum-character, as a speaker and as an addressee. Finally, Had I been to Knossus yet? is an example of DIST, where I links to the roles of the addressee and narrator, but not the actual speaker. Each of the pronouns is thus a prompt for a complex blend, linking the viewpoints of the relevant narrative spaces, and, perhaps primarily, situating the utterance with respect to the viewpoint of the SV-space. It seems an oversimplification to treat such grammatical choices as sentential only, as they appear to be the means narrative discourse relies on in achieving story-wide coherence and in prompting the processes of narrative space integration. To put it simply, such pronominal choices participate in overall processes of story construction. It is thus not the case that all of the aspects of pronoun interpretation in the narrative are a part of the meaning of pronouns. On the contrary, I argue that a pronoun like I is indeed no more than a pointer to the current speaker (also in the textual sense of the current narrator), seen against the available ground. The complexities of reference emerging in the examples throughout this chapter result from the narrative space network, not the pronoun itself, but there are specific regularities in the network. For example, an off-stage narrator in an
7.4 Viewpoint compression and constructional phenomena
189
SV-space cannot be referred to as I, but once the narrator comes on-stage (in intrusive narration or first-person narration) the first-person pronoun can be used. Moreover, if there is a character cross-mapped with the narrator, that character will also be an I in various narrative spaces, as long as there is a crossspace connection coming through a network of blended spaces. The coordination of grammatical (and especially deictic) forms in the narrative thus happens via three processes, but the same processes also explain the emergence of the story. First, lower-level spaces are embedded in the higher-level spaces. This applies to forms such as indirect discourse, where the embeddings are at the sentential level, and thus coordinate pronouns and tense to the required degree. But it also applies to larger spaces, such as flashbacks, spaces prompted by different temporal and spatial topology, subplots, et cetera. Grammatical coordination may be less strict then, but some is required: for example the embedded novel in The Blind Assassin is coordinated with the rest of the novel by strict avoidance of proper names and complete reliance on pronouns, but the narrative tense is present, even though we learn early on that the space elaborates a sub-story which resides in the past of the narration. Second, there is blending. Clusters of spaces are gradually integrated into larger chunks of the story, until the complete story emerges (this does not preclude lower-level, sentential blends such as FIST). This process is coordinated with respect to the SV-space, whose constant viewpoint allows temporal progression to emerge and also maintains identity cross-links (so that a character referred to as Alex in one of the spaces of The Blind Assassin is finally crossmapped with he in the novel and the ‘blind assassin’ role in the pulp fiction novel, of which he is also the author). The process of reading constantly increases the number of cross-space connectors, and thus the process of the emergence of the story (through projection and integration) progresses. Finally, there is the mechanism of viewpoint compression. I have argued throughout the book that compression of viewpoint is what allows spaces to be integrated by their shared viewpoint. Even though their viewpoints may be different (let’s say, because of a different temporal ground or different narrator), one of the spaces may be selected so that its viewpoint is the main narrative viewpoint for a part of the novel or the whole (see the relevant discussion in chapters 2 and 4). Also, as I argued elsewhere (Dancygier 2005a), additional spaces may be set up when a space is decompressed into more spaces to mark different viewpoints within it, and then the viewpoint of one of the decompressed spaces is compressed with the main narrative viewpoint, substituting it for as long as the space remains active. Similar phenomena can be observed when decompression is not involved. For example, in example (15) above, the part of the main narrative space being elaborated is the interview. The report includes primarily what Mr. Pomfret said, and Delia’s concluding words, but none of this is done through direct speech. The interview space thus contains
190
Speech and thought in the narrative
another space – the report of Mr. Pomfret’s words. The report is first introduced through the narrator’s summary of how Mr. Pomfret conducted the interview, then sets up a number of sentential-level STR spaces, detailing his words. But when a sentence starts with He said she would start tomorrow, it explicitly sets up the space of ‘Mr. Pomfret speaking,’ where the viewpoint is clearly that of the narrator (SV-space). The ensuing content of his utterance is embedded in the ‘speaking space.’ The next sentence, her hours were nine to five, is a blend of the reporting space and the reported space, but the viewpoint is that of the lower space, so the words of the character become the narrative viewpoint for the duration. Viewpoint compression thus foregrounds the words of the character while not abandoning the narrative viewpoint of the fragment. Viewpoint compression here is the mechanism which contributes to the integration of the FIST blend, and to the coherence of the story as a whole. Another issue is the distinction between two cases: embedding and blending. Mental space embedding, such as indirect discourse, involves some phenomena such as pronoun or tense shift. But this is primarily the result of a phenomenon widely talked about in mental space theory, that of projection. A higher, embedding space (such as He said) projects its tense and reference pattern into the lower, embedded spaces (she would start tomorrow, rather than you will start tomorrow), very much like in conditional constructions, where the if-space (If I were you) projects its stance, marked by past tense, and its referential pattern, into the lower, main-clause space (I would marry him). Blended spaces, for comparison, do not necessarily maintain the hierarchical organization or the syntactic dependence, while also allowing the tense and reference to be adjusted to the participating spaces. As a result, a free indirect utterance such as Had she typed? marks many of the same shifts, but does not maintain the embedding structure. However, as I suggested above, the viewpoint of the higher spaces still dictates the grammatical choices. The fact that these embedding and blending processes occur is the reason why lengthy narratives can be read as coherent. Contemporary narratives break some of these rules and experiment with much more fragmented structure, but they can be appreciated for what they are precisely because we have a tacit understanding of how viewpoint compression smoothes out narrative space boundaries. Discussions of STR often consider reasons why adverbs of time are sometimes excluded from the viewpoint adjustment, as in He said she would start tomorrow above, where tomorrow is future with respect to the time of the conversation and not the time of the report. This is in fact similar to what has been labeled the ‘was-now’ construction, recently put in the constructiongrammar perspective in Nikiforidou (2010). I argued earlier that such notfully-integrated uses of adverbs should be treated as signals of the maintenance of their spaces in the network. My interpretation is slightly different from Nikiforidou’s, in that I do not assume that this is a constructional pattern
7.4 Viewpoint compression and constructional phenomena
191
resulting from coercion. Nikiforidou discusses the resulting blend of spaces elsewhere, but my point is that this is much less a signal of a blend than a signal of incomplete integration which serves a specific purpose. Tomorrow is not only the day after the conversation between Mr. Pomfret and Delia, it is also the next day from the point of view of the story as a whole. Deictically, tomorrow is situated identically with respect to the interview space and the SV-space, which has a narrator in the present and tells the story in the past. The ‘was-now’ construction is found in the narration proper as well as in the STR constructions. Since the former case seems more interesting from the point of view of the framework being developed here, I will consider some narration examples below. Example (19) repeats example (6) from Chapter 3; it is a fragment from Fielding’s Tom Jones; example (20) is another classic – the first paragraphs of James Joyce’s “Eveline”: (19) He now lived, for the most part, retired in the country, with one sister, for whom he had a very tender affection. This lady was now somewhat past the age of thirty. (Tom Jones, p. 17)
Example (19) comes after two initial paragraphs in which Squire Allworthy is introduced, along with his story, and thus the now functions as a kind of resumptive pronoun, announcing that the excursion into the pre-story past has ended and the in-story past begins. It is thus a means to organize the readers’ understanding of where along the axis of time narrative spaces belong. The resulting construal highlights the current temporal viewpoint of the narrative, setting it against a broader temporal space. Interestingly, while being stylistically salient, the now in the fragment is not indispensable to the processing of the narrative. The function of now is more explicitly resumptive in (20). The story starts with the image of Eveline sitting at the window, looking outside. The next part describes her childhood memories, prompted by the current situation – in this part many past adverbs are used, so that the elaboration of the narrative space of Eveline’s childhood is maintained. However, the last sentence of the paragraph starts with now and then frames the imminent events which have prompted the memories – Eveline’s planned departure for America. (20) She sat at the window watching the evening invade the avenue. Her head was leaned against the window curtains and in her nostrils was the odour of dusty cretonne. She was tired . . . . . . One time there used to be a field there . . . Still they seemed to have been rather happy then. Her father was not so bad then; and besides, her mother was alive. That was a long time ago; she and her brothers and sisters were all grown up; her mother was dead. Tizzie Dunn was dead, too, and the Waters had gone back to England. Everything changes. Now she was going to go away like the others, to leave her home. (Dubliners, p. 25)
192
Speech and thought in the narrative
Being used at the beginning of the sentence, now is here also a discourse marker announcing a sudden shift of viewpoint. Both (19) and (20) show that the adverb is not simply an unexpected reference to the present in the middle of a past story, to be blended into the past background, but that it is a specialized viewpoint marker, signaling the return to the narrative space which is being currently elaborated. So if it is a temporal marker at all, it picks out a specific narrative space as the current ground of the narration. It is also worth noting that in (20), unlike (19), now is a necessary addition to the text, since without it the paragraph-ending sentence would be difficult to process. Clearly, there is a shift of viewpoint from one space to another happening here, and the sentenceinitial use of now marks that shift. Example (20) supports the claim that narratives operate through the setting up of narrative spaces and their layering in terms of their embeddedness and subsequent blending. There is much textual signaling of the different spaces, but the use of now considered here shows that viewpoint shifts may also be marked through specific means. This use of now does not mark an independent meaning, specific to the adverb’s narrative function exemplified here, because the new meaning emerges at the intersection of the deictic meaning of the adverb (a pointer to the current space) and the structure of the narrative network (so that the past space can be selected as the current space to be elaborated). Both (19) and (20) do not have narrators who are also characters. Example (21), for comparison, uses now in the first-person narration of The Blind Assassin. In the fragment, Iris, the narrator and character, describes looking at old notebooks, hidden by her sister to prevent the truth about her life from being revealed: (21) I riffed through the other notebooks. History was blank, except for the photograph Laura had glued into it – herself and Alex Thomas at the button factory picnic, both of them now colored light yellow, with my detached blue hand crawling towards them across the lawn. (The Blind Assassin, p. 627)
The use of now here is not clearly resumptive, but it also signals the temporal structure of what is being described. The adverb comes within a description of the photograph Iris sees in the notebook, in which a hand at its edge has been tinted blue. But the blue belongs to the time when Laura colored the photo and put it in her notebook, while the yellowish shade of the photo overall (we can assume it was black-and-white) has emerged over the years that have passed. So the ‘yellow’ here is the color observed years later, when Iris is looking at it. The fact that Laura glued the photo in the notebook is the pre-past of the story, while the yellow tinge belongs to the now of the scene narrated – when Iris finds the notebooks. Now thus emerges as a marker of time, but not the present time. Instead, it highlights the narrative space currently in focus, especially when the context
7.4 Viewpoint compression and constructional phenomena
193
highlights some earlier events. The now marks a return to the temporal space currently elaborated, and in some cases this use may further be resumptive. Throughout the chapter I have looked at the way in which the lowest-level discourse spaces profiling speech or thought are represented in the text. While there are clearly various constructional phenomena involved, it is not the case that traditional categories such as direct, indirect, or free indirect discourse naturally refer to classes of sentences falling under a cohesive linguistic description. Rather, the history of research into STR shows intensive negotiation of the boundaries of each category, of new categories needed, or of the actual status of adverbs and expressive elements. In this section, I have not attempted either to summarize or to revise the existing categories. Instead, I have proposed an approach wherein individual formal signals, such as a pronoun, a temporal adverb, a verb form, or a syntactic pattern, contribute their meaning to the overall construal, but allow the construal to emerge as a result of a blending process reconciling all the components of the expression. Thus a pronoun I points deictically to the speaker situated against a ground, but if there are several grounds in the network, I may prompt a chain of projection and compression in which different spaces will be involved, and the designated participant may appear not to be the speaker in the standard sense. This approach goes beyond Rubba’s discussion of alternate grounds (1996), because the matter of reference in the narrative may require involving the network of spaces, rather than different or competing spaces. The discussion of STR often yields new categories – such as DIST, proposed by Vandelanotte. There is indeed a crucial difference in its design, the shift towards the deictic centre of the current speaker (rather than represented speaker). But the difference between DIST and FIST is primarily in how deictic forms such as pronouns are aligned with the two mental spaces in the required network (current versus represented). Perhaps this is a criterion which could be further studied profitably – in fact Dancygier and Vandelanotte (2009) discuss other uses of a similar deictic alignment in a range of colloquial and literary uses, under the term distanced discourse. Clearly, alignment with the current speaker is a compositional feature of more than one construction, though it also reliably distinguishes two forms of STR. In a completely different context, the vocative-cum-imperative construction discussed in Chapter 6 builds on the same mechanism of constructional compositionality – the imperative evokes a desirable though not predictable future occurrence, and in a typical context it involves a speech act of ordering or requesting, with its appropriate felicity conditions. But this is, so to speak, the ‘strong’ form of the imperative, which is absent from uses such as Open the door and I’ll shoot! (cf. Dancygier and Sweetser 2005), where the action of opening the door is not only not ordered or desired, but outright discouraged. When Dr Faustus urges the Earth to ‘gape’ and swallow him so he can escape
194
Speech and thought in the narrative
hell, he is not ordering, but is imagining the desirable state of affairs occurring. The most salient use of a form (such as the ‘order’ for the ‘imperative’) is not necessarily the most accurate. In fact, it is usually better to have a minimalist description of a form’s meaning, and then find out how this minimal meaning gets enhanced in actual constructions. The narrative is too complex a language form to be naturally analyzable in terms of a set of fully determined constructions. Throughout this book I have attempted to focus on selected mechanisms responsible for the emergence of the forms and meanings readers pull out of texts. But, as I tried to show, all those meanings are functions of specific kinds of narrative space networks, and of phenomena such as embedding, projection, and, perhaps first of all, viewpoint compression. These mechanisms, however, do not yield specific constructions, but rather propose interpretive strategies which make sense of the amazing variety of narrative choices available.
8
Stories in the mind
Life is a trace in the ocean of stories.
The analyses and theoretical constructs proposed throughout this book apply to some of the more common questions raised by fictional narratives, but there is a host of more specific problems which can gradually be addressed. In this chapter, I will sum up some of the main directions the examples considered suggest and outline the general conclusions emerging.
8.1
The linguistics of literature
Stories are linguistic artifacts, but they require a language theory which allows the analyst to consider questions of meaning and form in some correlation, and deal with patterns of emergent interpretations. The primary interest of cognitive linguistics is the elucidation of form–meaning mappings and the uncovering of underlying cognitive concepts, and thus it seemed worthwhile to test it in a broad literary perspective. In the narrative context, where forms are less naturally in focus and meanings are of primary interest to readers and analysts alike, cognitive linguistics opens new avenues of investigation, with respect to both form and meaning. The nature of narrative meaning is a further question I tried to address. I have argued that such meaning emerges in the reader’s mind and cannot be attributed to the textual prompts alone, but relies also on the frames available, on the reader’s disposition, on the cultural context, et cetera. Thus, while the text serves as a prompt, it does not limit the ways in which the prompts will be interpreted. What matters most is the range of processes which may yield different readings, and not the readings themselves. This approach appears to question some assumptions analysts of literature may hold dear – either the expectation that linguistic tools can predict the interpretive result, or the belief that such a result is primarily a matter of influences beyond the text. However, no such claims are being made. While interpretive acts are motivated by the linguistic processes I tried to describe, they are not predictable. Also, the frames available in the reader’s mind participate in the construal alongside the text and 195
196
Stories in the mind
come into play at any level. It is the interaction of the linguistically motivated aspects of interpretation with cultural or historical frames that may yield a specific reader’s specific reading. But the nature of that interaction is at the core of the process and needs to be formulated. Throughout the book, I talked about macro-level phenomena and micro-level phenomena, and relied on two mechanisms that explain the cross-level coherence. The first mechanism is blending – a convenient term to represent processes which yield emergent meanings. The second mechanism, central to my argument, is viewpoint compression. While blending gracefully explains the fact that narrative meaning emerges through selection of conceptual material and its integration into new, higher-level structures, viewpoint compression was argued to be the main driving force behind these processes, playing a major role in the actual textual choices. I defined narrative viewpoint independently of the tradition which assumes a viewing Ego, and linked it to the concept of a narrative space, as a partial construct structured by a certain topology. A narrative space affords various viewpoint dimensions – temporal, spatial, emotional, et cetera – but it also relies on the presence of participants and their dispositions. Viewpoint can thus be aligned with one of those participants, but may shift to time or space, or to an unprofiled, imagined participant, or to another narrative space. Those viewpoint shifts, crucial to the complexity of fictional narratives, are managed across narrative levels through compression – a process which may incorporate a viewpoint of a lower space into the higher space, and gradually reach the highest narrative level, wherein the story as a whole is construed as being narrated. The approach tested some assumptions I had made – both about the role of the text and about the role of frames in the processing of texts. It was also a test of the applicability of blending in the analysis of longer texts. I will briefly review the conclusions below. 8.1.1
The sentence versus the text
The main linguistic problem I tried to address is the ways in which the lowestlevel grammatical and lexical choices participate in the meaning of the text in its entirety. Naturally, not all such sentence-level forms contribute significantly to how a narrative will be read, but there are at least two instances discussed here where the choices matter throughout the text as a whole. One such example is the choice of the structure of the SV-space and its relation to MN-space. Grammatical options such as pronouns or tense reduce or expand the distance between the locus of the narrating and the actual storyworld. The degree of distance, in turn, affects the epistemic impact of the text and, consequently, the ways in which what is narrated is processed. First-person narratives usually mark less distance, but also limit the range of viewpoints available and leave
8.1 The linguistics of literature
197
significantly more unsaid, or require that it be said differently. For example, the first-person narrator of Atwood’s The Blind Assassin yields a very personal tone of the narration, but the text also needs to construct other major narrative spaces to profile epistemic viewpoints which are indispensable in the grasping of the story. Similarly, in Raban’s travel writing, frequent decompressions of identities and concepts, marked by sometimes unusual textual devices, serve the purposes of viewpoint management at the level of the story. At the same time, Woolf’s multi-Ego narrative plays with many perspectives all the time, but creates no pretense of a consistent epistemic viewpoint and requires that all the work of piecing the story together is done in the reader’s mind. In all specific cases, lower-level narrative choices are responsible for the kind of interpretive work the reader does. Also, there is the question of lexical choices or minute reactivations of images and scenes, which I discussed under the rubric of narrative anchors. While my most complex example is that of anchors in The Blind Assassin, there is a fair degree of anchoring and reactivation patterns in all narratives. Quite often, the anchors are material anchors written into the story (such as the photograph in The Blind Assassin), so there is a clear correlation between evocation patterns identified in cultural artifacts and those developed in narratives. One such example, not mentioned earlier in the text, is Melville’s Moby-Dick, where unusually numerous objects contribute to the understanding of what motivates the characters and open more general avenues of interpretation (let me mention only the doubloon, Queequeg’s coffin, the Pequod’s tiller, or Ahab’s leg). In each such case the object built into the story is endowed with story-specific meaning, which is reactivated every time the object is mentioned and contributes to story construction. As I argued throughout the book, these, and other, lowest-level choices play a role in the understanding of the text in its entirety and achieve their narrative power through the processes of viewpoint compression. The compressions reduce and remove the distance across various narrative spaces, thus contributing to the compact and well-integrated structure which finally emerges as the story. Viewpoint compression, as I argue, is the central mechanism leading to story coherence, as the emergent construct gradually reaches its completeness. The process of reading is thus not simply linear and does not rely primarily on the accumulation of information. It is a multidimensional process, reaching across narrative spaces in different directions. In the responses to my earlier work I often heard colleagues sounding uneasy about the idea that a sentence may have a narrative status comparable to a fragment, or even an independent subplot, as they can both be treated as narrative spaces. But there is no reason to assume that compression does not work across sentences, and is only valid when more major narrative fragments are involved. True, when the text chooses to separate narrative chunks in terms of viewpoint, as in The Blind Assassin,
198
Stories in the mind
viewpoint compressions are more visible, but the process is equally valid in Eggers’s text, where numerous (but very short) escapes into imaginary scenarios also require compression to yield the general impact of the obsessive and frustrated narration. The narrative choices may be more or less obvious, but the process remains the same. This approach made my discussion much less focused on the distinctions which narratology relies on. Much ink was spilled to elucidate the distinctions between narration (as the mode in which the narrator ‘tells’ the events of the story) and various kinds of speech and thought representation. In my discussion of texts, I have assumed that these distinctions are primarily available at the lower levels of narration, since ultimately the entire text is in the scope of the SV-space and thus needs to be compressed to contribute to the narrative as a whole. I have also not attempted to rely on the idea of a ‘narrative event’ and assumed that a fantasy in the narrator’s mind is no different in its narrative status from a description of a character’s behavior: either one might be compressed into the story to add to the understanding of the character’s plight, and whether it was worded as a narrator’s description or a direct insight into the mind is a matter independent of the story-emergence mechanisms. The mechanisms were my main focus. 8.1.2
Constructions in the narrative
My approach relies heavily on the concept of a construction, but it uses the term in a sense which seems more appropriate in the context of the narrative. Narratives naturally rely on the constructions available in the colloquial context (such as the ditransitive construction or the caused-motion construction), just as they rely on the entire linguistic repertoire of the language in question. However, they also create constructions specific to the genre, where linguistic resources are deployed in a less strictly patterned way. I have referred to this phenomenon as constructional compositionality, but the narrative-specific ways in which linguistic forms are used may call for a discussion of constructions supporting specific features of the genre. One example of a construction which seems to be appearing primarily in dramatic narratives is the ‘vocative-cum-imperative’ construction in Early Modern drama. Both the vocative and the imperative are language forms available in English, but their function in the cases like Come, bitter conduct! is adjusted to the meaning of the scene and strictly dependent on the actions performed on stage. There are apparently many such constructions worth discussing. For example, I have not devoted much attention to the use of tense in the narrative, but there is clearly a need to elucidate what the present tense in narratives means, beyond being compressed to the SV-space to yield a less distanced narration. It definitely relies on the meanings available in colloquial discourse, but because it is not
8.1 The linguistics of literature
199
clearly correlated with a standard deictic ground, there are also differences in meaning. Questions can also be asked about the narrative use of demonstratives, articles, modals, and many other linguistic forms. There is much similarity in usage, but there are also differences. As I have shown, constructions may be specific to the linguistic ramifications of the genre. The ‘vocative-cum-imperative’ has been used in drama, while interesting uses of deictic pronouns can be found in poetry. Various types of narrative modes are available in novels and short stories, and some of them may be present in epic poems, but the deployment of constructional resources may be genre-specific and thus opens the door to an enquiry of the textual choices a genre makes available. These questions connect with the recent research on spoken and written genres. For example, studies of conceptual metaphor in discourse (cf. Semino 2008, 2010) suggest that different types of texts might call for a different use of conceptual metaphor. As Semino shows, scientific discourse may rely on elaborating an improbable source domain in order to allow for a new construal of a complex target. Also, a recent study by Moder (2010) shows that genres such as radio interviews may require constructions which rely on available resources (such as similes or copulative constructions) but yield a result geared into the context where the listeners have no access to visual contact with the guest in the studio. Narrative study confirms that specific discourse contexts use constructional resources of English in new ways. What seems to be required, then, is an understanding of how the particular type of contact between the recipient and the text may alter prototypical uses of linguistic forms. At the same time, the constructional and compositional approach would add to the linguistic study of genres, literary and non-literary. Possibly the most interesting constructional aspect of narratives is the use of deixis. My analysis has not fully covered the tense dimension of the use of deixis, but I have assumed that the SV-space constitutes a correlate of the deictic ground, so that the communication occurring in that context, even if written and directed at an unspecified addressee, assumes a default temporal dimension. Thus the SV-space time, like the time of any exchange, is present, and the story can either be distanced from it, which means told in the past tense, or compressed to it, which means that the choice is present tense. But the past itself is also an interesting choice, in that, on the one hand, it supports the default SV set-up and, on the other hand, it puts events in specific relations with each other, such as precedence. My analysis has been more specific in its treatment of pronouns, but these questions go far beyond the default deictic set-up. In various texts, pronouns have widely varying functions. They may open links to other narrative spaces (as in The Blind Assassin) or support decompressions for viewpoint (as in Raban’s work), or signal STR blends and other more unusual blends when a viewpoint of one space is compressed with that of another. They may also, as in
200
Stories in the mind
the case of Szymborska’s poems, signal unusual deictic set-ups in which participant roles are assigned in non-standard ways. What is systematic about all these options is that the choice of a deictic pronoun like I or you calls up a deictic centre (whether previously set up or not) and assigns communicative roles to the participants, even inanimate ones. Furthermore, once such a set-up is established, pronouns may serve functions which have much less to do with the flow of discourse and the resulting anaphoric or cataphoric relations, but instead connect various available spaces ‘at various angles.’ The range of pronoun use in the narrative is extremely broad and should be investigated further. The instability of constructional phenomena in the narrative is also correlated with the issue of narration versus speech and thought representation. As I argued, it is difficult to distinguish types of STR either as independent constructions or as instances of narration, speech, or thought, as the constructional correlates are not systematic enough. Also, any of the specific choices may have limited bearing on the overall interpretation: in the case of the reports included in the first paragraphs of The Blind Assassin, the specific words and thoughts of different participants are a stylistically interesting way of structuring two viewpoints, but there is a gap between the meanings specific to the constructions used and to the overall narrative effect. Furthermore, the fluidity of STR categories and their links to narration clearly supports the idea of compositionality, and poses questions which need investigation. Naturally, we can investigate the role of pronouns and tense in the management of viewpoint, and consider contextually interesting uses of adverbs from the point of view of their alignment with viewpoint spaces, but sentence types and complex sentences can also be viewed as compositionally interesting. If the form of a question can represent a character’s or on-stage narrator’s thought, then we need an explanation of how this effect emerges and what low-level aspects of questionhood had to be tweaked for the use to emerge. Similarly, while Verhagen (2005) has pointed out the intersubjective nature of reported speech and thought constructions, many questions still remain about intersubjective construals emerging out of different reporting choices. Practically any aspect of grammar has its interesting correlates in the narrative context. 8.2
The storyworld reality
The question of the status of the worlds structured through narrative discourse was another important recurring issue. I was arguing against the assumption that we need to surgically remove narratives from the scope of our ordinary assumptions about the world and language. Narrative worlds constructed linguistically (or visually, as in film) may depart from the reality of our experience in many ways, but they do not depart from it more than our everyday
8.3 Blending and narrative analysis
201
imagination might allow. As almost any paper on blending or metaphor shows, our ability to create imaginative construals of situations is not as limited as the strict division between reality and fiction assumes. My approach is thus closer to the work on text worlds than on possible worlds, since the former is much more of a discourse theory than the latter. I have also tried to show the ways in which reality participates in narrative construals. First, any text relies on our reality-based knowledge of frames: if we read about someone making a phone call, we understand it because we have an easily accessible frame of devices making communication over long distances possible. Frames and patterns of frame evocation constitute a reality-based conceptual network which underlies any construal a narrative may come up with. Even a science-fiction story has to rely to the same degree on accessible frames to understand how the imagined world questions the frames observed in reality (knowing gravity in real life is crucial to understanding what absence of gravity might mean). Second, the historical context of the text or the influence of the author’s experience should be treated as an integral part of the emerging meaning of the narrative, rather than as separate from it. Narratives could not be written coherently without being immersed in some cultural, social, or individual knowledge and experience. The reader may apply different frames, but the very reliance on frames is what makes narrative interpretation possible. Another dimension is the narrative rendering of characters’ experience of the world. I have talked about fictive experience, which might mean that narrative events and other construals are represented in the text as if they were seen or heard by someone, but there are more interesting patterns to explore if we consider various ways in which narratives, whether viewed as produced or received, rely on the experiential and embodied construals of experience. I have focused for a while on fictional minds and the representation of mental states, but questions of fictive experience reach far beyond these categories. 8.3
Blending and narrative analysis
In order to describe the fluid nature of narrative meaning construction I have selected a framework which seemed best suited to an analysis of meaning emergence – conceptual integration. The analysis suggests two conclusions. One, blending can indeed be applied in the analysis of more complex texts. And two, narratives may expand our understanding of how blending works. But the question of further implications of my theoretical choice remain difficult to answer. For example, researchers have asked an important question – what is the psychological reality of blending? Opinions vary. The proponents of blending (Fauconnier and Turner 2002) make strong claims about the role of doublescope blending in the evolution of modern human beings. Followers in many disciplines, from language through music, art, and performance to literature,
202
Stories in the mind
find blending a very helpful tool in reaching beyond structure-based theories. And yet it is possibly too early for experimental work in neuroscience or psychology which would tell us if blending is ‘real’ and can naturally be distinguished from other paths of conceptualization.1 However, as an interpretive tool, blending has become possibly the single most useful theory of meaning. It allows one to capture the so far unattainable scope of phenomena in a coherent way, and it is the first theory of meaning which is not encumbered by being restricted to this or that area of enquiry. It has uncovered hosts of observations on human processing of prompts and the economy of our meaning construction mechanisms. Blending may not become scientifically satisfying for a while, but it is very satisfying already from the point of view of language analysis. As regards stylistics and literary criticism, the complaints are (often ironically) different – couldn’t we say the same without blending? isn’t it too scientific to be useful to me? It is hard to imagine a theory which could satisfy all its critics, especially if they come from different academic cultures, with their specific expectations. But perhaps the fact that these divergent groups of researchers consider the pros and cons seriously is a signal that it is appealing enough to warrant more effort. I hope to have shown that there are merits to the blending framework, whichever academic culture one comes from. The examples analyzed throughout this book are useful in putting these questions in context. First, they show that meaning construction processes work in similar ways through different levels of language structure and that there is a natural way to connect sentence-level phenomena with textual-level processing. Second, blending allows one to incorporate complex cultural frames into the picture. Also, it does not restrict interpretive efforts unreasonably. Although it does require that more attention be paid to what the text does, it opens itself to any culturally or historically based set of questions. And it may still reach its experimental phase. Many of the concepts emerging out of cognitive linguistics research, such as embodiment or the metaphorical nature of abstract thought, have prompted experimental work which supports the validity of these research tools.2 Thus, while blending is so useful because it allows one to capture processes rather than static, complete conceptual structures, it is also difficult to test for, for the very same reason. Perhaps the material presented in this book might at least explain why it is so hard to do. Narratives evoke a number of complex research questions: the role of memory mechanisms which link language forms to representations of complex situations; the nature of linguistically prompted emotional responses; our embodied sense of how actions signal mental processes; the nature of mental processes involved in the construction, maintenance, and development of constructs called storyworlds; reliance on one’s own experience with various cultural artifacts, including narratives; the dynamic nature of text processing;
8.4 A bridge to the truth
203
the ways in which linguistic expressions can negotiate various forms of access to other people’s minds; and the evolutionary processes that led to the emergence of all these abilities. No less. So until and unless methods in psychology and cognitive science can help in answering these questions, blending will have to remain an interpretive tool and continue to inspire more difficult questions. It has achieved these goals when applied to the material in this book.
8.4
A bridge to the truth
In the motto to Chapter 1 of the book I gave a verse from Arabian Nights which claims that a tale is a bridge that leads to the truth. Perhaps “the truth” is an exaggeration, but I found the quotation appropriate to this project, because it highlights the fact that whatever understanding a reader might acquire, it is not contained ‘in’ the story, but can only be arrived at through the interaction with it. The understanding of the narrative as a kind of container into which a reader reaches to pull out the story seems to be an inheritance of the Conduit Metaphor, as described in Reddy (1979). It is a complex metaphor which conceptualizes linguistic expressions as containers into which meanings are packaged in order to be sent to the receiver, who unpacks them. The primary fallacy of this construal is that meaning is stable, bounded and possible to transmit intact. The conduit fallacy explains some of the unrealistic language-based assumptions about the meaning of fiction being contained in the language of the text, ready to be unpacked. Apparently, the fallacy has also led to a broad rejection of the view that the text matters in reading, beyond the basic meanings of words and the accessibility of syntactic forms. The rejection, in turn, has prompted the belief that meaning is so unstable that an analyst cannot presume to find it, though she or he may uncover some important elements, including those which the author may have never intended to include (this would go, among others, for the variety of interpretations intent on finding a feminist, racist, or other politically defined angle in the text). These approaches to narrative meaning tend not to look at the form of the narrative as a relevant criterion, and thus are also unrealistic, though in a different way. The approach I have argued for suggests that narrative meaning, as any other linguistic meaning, is prompted by linguistic forms, but emerges in the process of interpretation, guided by general linguistic as well as specifically narrative processes. Included in this approach is the idea that a narrative communicative act is in some ways like any other communicative act – the author, like the speaker, constructs the text, and the reader, like a hearer, interprets it, based on the textual prompts, available contextual ground, and general knowledge, and in an intersubjective context where other viewpoints are accessible. Much of this book was devoted to processes leading to the interpretation.
204
Stories in the mind
But the nature of the process, the length of narratives, and the vast variety of background knowledge authors and readers bring into the exchange are jointly responsible for the fact that readers of the same text may not arrive at the same reading. They may (though do not have to) uncover the sequence of events, but that alone does not constitute the content of the story. The story defined here is a much richer construct, in which dispersed identities are established and mental states constitute important narrative material. Representation of minds is as central to the narrative as representation of events, and it is possibly the aspect of texts which yields most innovative narrative techniques. Sequence of events is the result of the characters’ motivations and causal chains thus created, not the source of narrative coherence. My approach stresses the dynamic nature of the process, which is elegantly captured in the motto to this chapter: Life is a trace in the ocean of stories. Lives are depicted in narratives, but the depiction chooses a path through the maze of events affecting the main characters. The path may meander in strange ways, but that is also a part of the process, since finding coherence means constructing an emergent structure that puts some events in focus and backgrounds others. The story moves through events surrounding it, and emerges out of the links the reader’s mind adds. It is not built into the ocean of meaning we float in: it has to emerge as we move through the waves.
Notes to the text
C H A P T E R 1 L A N G U A G E A N D L I T E R A RY N A R R AT I V E S 1. For more discussion of this phenomenon, see Fleischman 1989; Dancygier and Sweetser 2005. 2. Cf. Fauconnier 1994 [1985], 1997; Fauconnier and Sweetser 1996; Fauconnier and Turner 1994, 1996, 1998a, 1998b, 2002; Turner and Fauconnier (1995, 1999); Coulson 2001; Coulson and Oakley (2000, 2005). Also, check for more sources at markturner. org/blending.html. 3. The work in cognitive stylistics has already made a significant contribution to our understanding of the ways in which literature and cognition are linked (see Tsur 1992, 2003; Stockwell 2002; Gavins and Steen 2003; Bex, Burke, and Stockwell 2000; Burke 2005; Freeman 2000, 2005; Semino and Culpeper 2002), and much new work is being done in that context. 4. There is indeed a significant amount of linguistic work within blending, too much perhaps to be sufficiently represented here. The relevant papers will be referenced as the need arises. 5. One volume on cognition and the arts, which holds articles on visual art, music, architecture, and many general questions of aesthetics, is Turner (2006b); work by theatre and performance scholars has been given some focus in Hart and McConachie (2006) and Cook (2007, 2010); work specifically devoted to music has been done primarily in numerous articles and books by Zbikowski (2002, 2006 are very representative examples); film was taken up in Rubba (2009). 6. The initial outline of the view of Literary Darwinism can be found in Carroll (1995, 2005), and Gottschall and Wilson (2005). Also, a special issue of Style (Style 2008) contains not only a programmatic article by Carroll, but also a number of (mostly critical) responses to the idea, with a concluding rejoinder by Carroll. 7. Some of the discussion on literature and ‘evolution’ is fleshed out in the papers in Gottschall and Wilson (2005). 8. Cf. Mar and Oatley (2008); Djikic, et al. (2009). 9. The concept of embodiment in the analysis of linguistic meaning is now one of the central issues of cognitive linguistics. Starting with the earliest work by Johnson (1987) on image schemas, and Lakoff and Johnson’s introduction of metaphor theory (1980, 1999), the idea that linguistic meaning emerges out of patterns of embodied experience has become the central claim in cognitive linguistics and has been applied to a broader and broader range of phenomena. Recent publications, such as Johnson (2007) and the papers included in Hampe (2005) and Frank et al. (2008), move the issues further into cognitive science, and into an exchange with work in philosophy, such as Gallagher (2005). 205
206
Notes to pages 25–64
10. Much of this work is reported in the papers gathered in Pecher and Zwaan (2005b). The relevant grounding research is also interestingly reported in the introduction (Pecher and Zwaan 2005a). 11. Any discussion of the framework is beyond the scope of this work, but a very clear introduction can be found in Richerson and Boyd (2005). 12. Work in evolutionary psychology has been publicized mainly through articles by John Tooby, Leda Cosmides, and their collaborators (cf. Barkow et al. 1992). Forthcoming broader volumes, such as Cosmides and Tooby (in press) and Tooby and Cosmides (in press) will soon become the definitive sources. C H A P T E R 2 B L E N D I N G , N A R R AT I V E S PA C E S , A N D T H E E M E R G E N T S TO RY 1. The work on mental spaces is now too broad to review here. Many interesting papers can be found in Fauconnier and Sweetser (1996); there is also interesting work on tense and aspect (Cutrer 1994; Fauconnier 1997), conditionals and coordinate constructions (Dancygier and Sweetser 2005), or constructions of intersubjectivity, including negation, adverbial conjunctions, complementation, etc. (Verhagen 2005). It might be worth noting that much of the work on mental spaces has gradually merged with research on blending. 2. Some examples are: Oakley’s paper (1998) on Art Spiegelman’s Maus, Rubba’s (2009) paper on Mulholland Drive, Sweetser’s paper on Cyrano de Bergerac (2006), Turner’s work on compression and representation in cartoons (2006a), and Dancygier’s (2006) discussion of the movie Annie Hall. 3. I first introduced narrative spaces in Dancygier (2005a), then developed the concept in Dancygier (2007) and (2008a); the concept owes much of its structure to the discussion by Sanders and Redeker (1996) of mental spaces in journalistic discourse. 4. The discussion of the narrative in terms of possible worlds can be found, among others, in Doležel (1976, 1998); Maitre (1983); Pavel (1986); Ronen (1994); and Ryan (1991). For an excellent, broad review of the framework, see Semino (2009). 5. Some of the more useful discussion can be found in Chatman (1978); Rimmon-Kenan (1983); Toolan (2001); Porter Abbott (2007). C H A P T E R 3 S TO R I E S A N D T H E I R T E L L E R S 1. The concept of the narrator has been discussed in much of the work in narratology and stylistics. Some general terminological distinctions can be found in Booth 1961; Chatman 1978; Toolan 2001; Rimmon-Kenan 1983; Herman 2007. 2. For a discussion of the views of authorship in Coetzee’s recent novels, see Dancygier 2010; also, some brief discussion of Powers’s Generosity, in Chapter 4. 3. The role of evaluative aspects of stance in the construction of narrative viewpoint is beyond the scope of this project. See Harding (2004; 2007) for interesting examples and analyses. 4. Since my focus throughout the book is on the elucidation of the viewpoint profiled in a narrative space, I will not discuss the various construals referring to the postulated recipient of the text, whether that subjectivity is termed ‘narratee,’ ‘implied reader,’ or the like. These concepts can naturally be built into the framework, but are not my focus here.
Notes to pages 69–118
207
5. I am thus rejecting Banfield’s (1987) concept of an ‘empty’ deictic center in the narrative. On the contrary, I assume that every narrative act assumes a space wherein there is a source of viewpoint, which can then take the form of a narrator. The onstage/off-stage distinction is meant to take care of the differences in the degree of profiling of the narrating subjectivity. 6. Now in past narratives can have other functions, to be discussed in Chapter 7. 7. Example (10) is earlier used in Chapter 2 on page 38. 8. Mental space embedding was best described in Sweetser (1996), and then developed in Dancygier (2002). The crucial point to be restated here is that the embedded space inherits the viewpoint, especially epistemic viewpoint, of the mother spaces. Both papers argue for this effect of embedding by looking at verb forms in conditionals, but in the context of the narrative the structure inherited by the embedded space may include much of the space topology.
C H A P T E R 4 V I E W P O I N T: R E P R E S E N TAT I O N AND COMPRESSION 1. Focalization is the central concept in narratological enquiry, and it is not possible to give it justice here. The best introduction to the concept and its revisions can be found in Rimmon-Kenan 1983; Bal 1985; Van Peer and Chatman 2001; Herman 2002. 2. More discussion of ‘speech’ and ‘thought’ will be offered in Chapter 7. 3. ‘Represented speech and thought’ is yet another case of a concept which is central to narrative enquiry, and has thus inspired an impressive body of research. The most pertinent texts are Genette 1980; Leech and Short 1981; Banfield 1982; Bal 1985; Fludernik 1993; recent revisions, based on different methodologies, add much to the discussion (consider, for example, Sanders and Redeker 1996; Semino and Short 2004; Vandelanotte 2005 and 2009). For a good general review, consult Toolan 2001. 4. Cf. Lanser 1981; Rimmon-Kenan 1983; Ronen 1994; Prince 2001; Van Peer and Chatman 2001. 5. The idea of fictive motion was introduced in Talmy (1996), later developed in Talmy (2000); Matlock (2004); Langacker (2005). The ‘fictive’ idea has also been extended to ‘fictive interaction,’ introduced in Pascual (2002, 2006a, 2006b, 2008). 6. Some discussion of forms of STR and some of their constructional features will be included in Chapter 7. The discussion here is restricted to the possibility that STR phenomena are essentially an instance of the more general pattern of viewpoint compression.
CHAPTER 5 REFERENTIAL EXPRESSIONS AND N A R R AT I V E S PA C E S 1. For a general treatment of deixis, see papers in Rauh (1981). Wales (1966) is an excellent review of the issues posed by pronouns in general. The cognitively informed view of deixis can be found in Fillmore (1997). 2. It is not possible to review such approaches here, but let me briefly mention the framework developed in Van Hoek (1997), where referential relations are treated in terms of ‘reference points,’ a concept earlier established in cognitive grammar. 3. See Dancygier 2004a, 2005a, 2008b.
208
Notes to pages 119–172
4. In this discussion I am primarily interested in the role proper names play in story construction. They do, however, perform important discourse-organizing functions as well. For example, Van Vliet (2008) discusses the contrast between the choice of a pronoun or a proper name as important in signaling the episodic structure of narratives. Also, Vandelanotte (2004b, 2009) observes the role of proper names in signaling relations of accessibility relevant to the selection of speech and thought representation constructions. 5. See Chapter 7 for more discussion of this narrative technique. 6. Deictic Shift Theory has been proposed in the collection of papers edited by Duchan et al. (1995); see especially Galbraith (1995) and Wiebe (1955). It proposes a treatment of the specificity of textual deixis. 7. This chapter considers examples of pronoun use with the exception of constructions of represented speech and thought. These will be discussed in Chapter 7.
CHAPTER 6 FICTIONAL MINDS AND EMBODIMENT IN DRAMA AND FICTION 1. For a discussion of the approaches to the ‘ground,’ see Langacker (1990); Coulson and Oakley (2005b); papers included in the volume edited by Oakley and Hougaard (2008); and Verhagen (2005). 2. For a very basic introduction to the linguistics of drama, see Sanger (2001). 3. Theatre theorists include further divisions of the on-stage space into locus and platea, concepts emerging from medieval theatre and continued in Early Modern drama (see, among others, Weimann 2000); an in-depth discussion of these concepts is beyond the scope of this chapter, but it is interesting to note that even the earliest theatrical forms require different functional allocations of space. 4. I am relying here on the concept of a frame, as it is better matched with my overall framework. However, broader discussion of how objects carry meaning can be found in semiotics literature, e.g. Elam (1980). 5. Thanks to my colleague Patsy Badir for bringing this example to my attention. 6. Beyond the scenes discussed here, the Body/Mind blend and the vertical schema metaphors are used in almost every scene, including the central ‘seize the crown’ scene, where Richard and Bolingbroke are holding the crown together and Richard is describing his emotions through the metaphor of the bucket filled with tears, moving down with the weight of the sorrow. Further discussion of the scene would go beyond the scope of this chapter. 7. For more discussion in the context of theatre criticism, see Harris and Korda (2002). 8. Blackwell’s collection of essays (2007) explores it-narratives in some detail.
C H A P T E R 7 S P E E C H A N D T H O U G H T I N T H E N A R R AT I V E 1. There is a rich tradition (Leech and Short 1981; Semino and Short 2004) distinguishing a number of types of STR, including constructions specialized in thought or in speech (or indeed, writing) representation. 2. Some authors (cf. Brinton 1980; Adamson 1995) distinguish reported perception as a separate STR category, but the question remains open: how do we distinguish these STR modes in conceptual terms? In Chapter 4 I argued for a category of
Notes to pages 177–202
209
‘representation of conceptualization,’ rather than thought, and these comments apply here as well. 3. The discussion of quotations used as ‘selective depictions’ has further been advanced through Pascual’s concept of ‘fictive interaction’ (2002, 2006a), where stretches of discourse play the role of a ‘token’ of a given ‘type’ of an interactive move. 4. A review of the literature on STR cannot be undertaken here. Let me just refer the reader to the most comprehensive volumes, such as Banfield (1982); Leech and Short (1981); Fludernik (1993); Semino and Short (2004); and Vandelanotte (2009). 5. Fauconnier (1994 [1985]) uses the term Access Principle to describe a regularity in language use which depends on cross-connections between spaces. If two spaces are connected, speakers can choose an expression from one of the spaces to refer to an element in the other spaces. For example, if the book called Poetics, an element in a space of books, is cross-linked to the name Aristotle in the space of authors, one can talk about reading Aristotle without confusion.
CHAPTER 8 STORIES IN THE MIND 1. The search for the final experimental confirmation of the reality of blending still continues (see Gibbs 2000). Furthermore, there have been dismissive remarks also from the cognitive linguistics camp (as in an interview with George Lakoff reported in Sanchez Garcia 2003), suggesting that blending is not interesting precisely because it is present everywhere and it is in fact not more than neural binding – an unremarkable phenomenon. (The framework which Lakoff supports instead has become known as Neural Theory of Language [cf. Feldman 2006; Feldman and Narayanan 2003; Gallese and Lakoff 2005; Lakoff 2008].) If that, however, turned out to be true, then, as Mark Johnson commented in another interview (Pires de Oliveira and de Souza Bittencourt 2008), the role of blending is not diminished, because, unremarkable or not, neural binding is a hard fact. 2. See, for example, Spivey (2007), or papers in Pecher and Zwaan (2005b).
References
Adamson, Sylvia. 1995. From empathetic deixis to empathetic narrative: stylisation and (de)subjectivisation as processes of language change. In Dieter Stein and Susan Wright (eds.) Subjectivity and Subjectivisation: Linguistic Perspectives. Cambridge University Press, 195–224. Ariel, Mira. 1990. Accessing Noun Phrase Antecedents. London: Routledge. Bal, Mieke. 1985. Narratology: Introduction to the Theory of Narrative. University of Toronto Press. Banfield, Ann. 1982. Unspeakable Sentences, Narration and Representation of the Language of Fiction. Boston: Routledge and Kegan Paul. 1987. Describing the unobserved: events grouped around an empty centre. In Nigel Fabb, Derek Attridge, Alan Durant, and Colin MacCabe (eds.) The Linguistics of Writing: Arguments between Language and Literature. Manchester University Press, 265–285. Barkow, Jerome, Leda Cosmides, and John Tooby (eds.). 1992. The Adapted Mind: Evolutionary Psychology and the Generation of Culture. New York: Oxford University Press. Baron-Cohen, Simon. 1995. Mindblindness: An Essay on Autism and Theory of Mind. Cambridge, Mass.: MIT Press. Barsalou, Laurence W. 1999. Perceptual symbol systems. Behavioral and Brain Sciences 22: 577–660. Bex, Tony, Michael Burke, and Peter Stockwell. 2000. Contextualized Stylistics: In Honour of Peter Verdonk. Amsterdam/Atlanta: Rodopi. Blackwell, Mark (ed.). 2007. The Secret Life of Things: Animals, Objects, and ItNarratives in Eighteenth Century England. Lewisburg: Bucknell University Press. Booth, Wayne C. 1961. The Rhetoric of Fiction. Chicago University Press. Boyd, Brian. 2005. Evolutionary theories of art. In Gottschall and Wilson (eds.), 149–178. 2006. Fiction and theory of mind. Philosophy and Literature 30(2): 517–581. 2009. On the Origin of Stories: Evolution, Cognition, and Fiction. Cambridge, Mass.: Belknap Press of Harvard University Press. Bradley, Raymond and Norman Swartz. 1979. Possible Worlds: An Introduction to Logic and Its Philosophy. Oxford: Basil Blackwell. Brinton, Laurel. 1980. Represented perception: a study in narrative style. Poetics 9: 363–381. Bruner, Jerome. 1991. The narrative construction of reality. Critical Inquiry 18(1): 1–21. Burke, Michael. 2005. How cognition can augment stylistic analysis. European Journal of English Studies 9(2): 185–195. 210
References
211
Canning, Patricia. 2008.“The bodie and the letters both”: “blending” the rules of early modern religion. Language and Literature 17: 187–203. Carroll, Joseph. 1995. Evolution and Literary Theory. Columbia: University of Missouri Press. 2005. Literary Darwinism: Evolution, Human Nature, and Literature. Evanston, Ill.: Northwestern University Press. Chatman, Seymour. 1978. Story and Discourse: Narrative Structure in Fiction and Film. Ithaca: Cornell University Press. 1990. Coming to Terms: The Rhetoric of Narrative Fiction and Film. Ithaca: Cornell University Press. Cienki, Alan. 2008. Why study metaphor and gesture? In Cienki and Müller (eds.), 5–25. Cienki, Alan and Cornelia Müller (eds.). 2008. Metaphor and Gesture. Amsterdam: John Benjamins Publishing Company. Clark, Herbert H. and Richard J. Gerrig. 1990. Quotations as demonstrations. Language 66(4): 764–805. Cook, Amy. 2007. Interplay: The method and potential of a cognitive scientific approach to theatre. Theatre Journal 59: 579–594. 2010. Shakespearean Neuroplay: Reinvigorating the Study of Dramatic Texts and Performance through Cognitive Science. New York: Palgrave Macmillan. Cook, Guy. 1994. Discourse and Literature. Oxford University Press. Copland, Sarah. 2008. Reading in the blend: collaborative conceptual blending in the Silent Traveller narratives. Narrative 16(2): 140–162. Cosmides, Leda and John Tooby (in press). Universal Minds: Explaining the New Science of Evolutionary Psychology (Darwinism Today Series). London: Weidenfeld & Nicolson. Coulson, Seana. 2001. Semantic Leaps: Frame Shifting and Conceptual Blending in Meaning Construction. Cambridge University Press. Coulson, Seana and Todd Oakley (eds.). 2000. Conceptual blending (special issue). Cognitive Linguistics 11(3/4). (eds.). 2005a. Conceptual blending (special issue). Journal of Pragmatics 37. 2005b. Blending and coded meaning: literal and figurative meaning in cognitive semantics. Journal of Pragmatics 37: 1,510–1,536. Currie, Gregory. 1995. Image and Mind: Film, Philosophy, and Cognitive Science. Cambridge University Press. 2004. Arts and Minds. New York: Oxford University Press. 2007. Framing narratives. In Hutto (ed.), 17–42. Cutrer, Michelle. 1994. Time and tense in narratives and everyday language. PhD dissertation. University of California, San Diego. Damasio, Antonio. 1999. The Feeling of What Happens: Body and Emotions in the Making of Consciousness. New York: Harcourt Brace & Company. 2003. Looking for Spinoza: Joy, Sorrow, and the Feeling Brain. Orlando, Fla.: Houghton. Dancygier, Barbara. 1993. Interpreting conditionals: time, knowledge, and causation. Journal of Pragmatics 19: 403–434. 1998. Conditionals and Prediction: Time, Knowledge, and Causation in Conditional Constructions. Cambridge University Press. 2002. Mental space embeddings, counterfactuality, and the use of unless. English Language and Linguistics 6(2): 347–377.
212
References
2004a. Identity and perspective: the Jekyll-and-Hyde effect in narrative discourse. In Michael Achard and Suzanne Kemmer (eds.) Language, Culture, and Mind. Stanford: Center for the Study of Language and Information, 363–375. 2004b. Visual viewpoint, narrative viewpoint, and mental spaces in narrative fiction. In Augusto Soares da Silva, Amadeus Torres, and Miguel Gonçalves (eds.) Cultura e cognicão: estudos de linguistica cognitiva (2 vols.) Coimbra: Livraria Almedina, vol. 2, 347–362. 2005a. Blending and narrative viewpoint: Jonathan Raban’s travels through mental spaces. Language and Literature 14(2): 99–127. 2005b. Constructional compositionality and blending: the case of Polish SLVF constructions. In the Proceedings of the Thirtieth Annual Meeting of the Berkeley Linguistics Society. Berkeley, Calif., 456–467. 2006. Preface: what can blending do for you? Language and Literature 15(1): 5–15. 2007. Narrative anchors and the processes of story construction: the case of Margaret Atwood’s The Blind Assassin. Style 41(2): 133–152. 2008a. The text and the story: levels of blending in fictional narratives. In Oakley and Hougaard (eds.), 51–78. 2008b. Personal pronouns, blending, and narrative viewpoint. In Andrea Tyler, Yiyoung Kim, and Mari Takada (eds.) Language in the Context of Use: Cognitive and Functional Approaches to Language and Language Learning. Berlin: Mouton de Gruyter, 167–182. 2008c. Review of Arie Verhagen’s Constructions of Intersubjectivity: Discourse, Syntax, and Cognition (Oxford University Press 2005), Linguistics 46(3): 658–667. 2009. Genitives and proper names in constructional blends. In Vyvyan Evans and Stephanie Pourcel (eds.) New Directions in Cognitive Linguistics. Amsterdam and Philadelphia: John Benjamins Publishing Company, 161–184. 2010. Close encounters: the author and the character in Elizabeth Costello, Slow Man and Diary of a Bad Year. In Graham Bradshaw and Michael Neill (eds.) J. M. Coetzee’s Austerities, Farnham, UK: Ashgate Publishers, 231–252. in press a. Language, cognition, and literature. In Mark Collard and Edward Slingerland (eds.) Creating Consilience. Oxford University Press, 409–423. in press b. Narrative time, sequence, and memory: a blending analysis. In Ralf Schneider and Marcus Hartner (eds.) Blending and the Study of Narrative. Berlin/New York: Mouton de Gruyter. Dancygier, Barbara and Eve Sweetser. 2005. Mental Spaces in Grammar: Conditional Constructions. Cambridge University Press. Dancygier, Barbara and Lieven Vandelanotte. 2009. Judging distances: Mental spaces, distance, and viewpoint in literary discourse. In Geert Brône and Joeren Vandaele (eds.) Cognitive Poetics: Goals, Gains, and Gaps. Berlin: Mouton de Gruyter, 319–370. Deacon, Terrence. 1997. The Symbolic Species: The Coevolution of Language and the Brain. New York: W. W. Norton. 2006. The aesthetic faculty. In Turner (ed.), 21–56. Decety, Jean and Thierry Chaminade. 2003. Neural correlates of feeling sympathy. Neuropsychologia 41: 127–128. Dennett, Daniel. 1997. Consciousness Explained. New York: Little, Brown. Djikic, Maja, Keith Oatley, Sara Zoeterman, and Jordan B. Peterson. 2009. On being moved by art: how reading fiction transforms the self. Creativity Research Journal, 21: 24–29.
References
213
Doležel, Lubomir. 1976. Narrative modalities. Journal of Literary Semantics 5: 5–14. 1998. Heterocosmica: Fiction and Possible Worlds. Baltimore and London: Johns Hopkins University Press. Donald, Merlin. 1991. Origins of the Modern Mind: Three Stages in the Coevolution of Culture and Cognition. Cambridge, Mass.: Harvard University Press. 2001. A Mind So Rare: An Essay on Faculty Psychology. Cambridge, Mass.: MIT Press. 2006. Art and cognitive evolution. In Turner (ed.), 3–20. Duchan, Judith, Gail Bruder, and Lynn Hewitt (eds.). 1995. Deixis in Narrative: A Cognitive Science Perspective. Hillsdale: Lawrence Erlbaum Associates. Dutton, Denis. 2009. The Art Instinct: Beauty, Pleasure, and Human Evolution. New York: Bloomsbury Press. Elam, Keir. 1980. The Semiotics of Theatre and Drama. London/New York: Methuen. Emmott, Catherine. 1997. Narrative Comprehension: A Discourse Perspective. Oxford: Clarendon Press. 2002. “Split-selves” in fiction and medical “life stories”: cognitive linguistic theory and narrative practice. In Semino and Culpeper (eds.), 153–182. 2003. Reading for pleasure: a cognitive poetic analysis of “twists in the tale” and other plot reversals in narrative texts. In Gavins and Steen (eds.), 145–160. Emmott, Catherine, Anthony J. Sanford, and Eugene J. Dawydiak. 2007. Stylistics meets cognitive science: studying style in fiction and reader’s attention from an interdisciplinary perspective. Style 41(2): 204–227. Evans, Brad. 2007. Introduction: rethinking the disciplinary confluence of anthropology and literary studies. Criticism 49(4): 429–445. Evans, Vyvyan. 2004. The Structure of Time: Language, Meaning and Temporal Cognition. Amsterdam: John Benjamins. Fauconnier, Gilles. 1994 [1985]. Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge University Press. 1997. Mappings in Thought and Language. Cambridge University Press. Fauconnier, Gilles and Eve Sweetser (eds.) 1996. Spaces, Worlds, and Grammar. University of Chicago Press. Fauconnier, Gilles and Mark Turner. 1994. Conceptual Projection and Middle Spaces. San Diego: University of California, Department of Cognitive Science Technical Report 9401. 1996. Blending as a central process of grammar. In Adele Goldberg (ed.) Conceptual Structure, Discourse, and Language. Stanford: Center for the Study of Language and Information, 113–130. 1998a. Conceptual integration networks. Cognitive Science 2(22): 133–187. 1998b. Principles of conceptual integration. In Jean-Pierre Koenig (ed.) Discourse and Cognition: Bridging the Gap. Stanford: Center for the Study of Language and Information, 269–283. 2002. The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic Books. Feldman, Jerome. 2006. From Molecule to Metaphor: The Neural Theory of Language. Cambridge, Mass.: MIT Press. Feldman, Jerome and Srini Narayanan. 2003. Embodied meaning in a neural theory of language. Brain and Language 89: 385–392.
214
References
Fillmore, Charles J. 1985. Frames and the semantics of understanding. Quaderni di Semantica 6: 222–254. 1990. Epistemic stance and grammatical form in English conditional sentences. CLS 26: 137–162. 1997. Lectures on Deixis. Stanford, Calif.: CSLI Publications. 2006. Frame semantics. In Dirk Geererts (ed.) Cognitive Linguistics: Basic Readings. Berlin: Mouton de Gruyter, 373–400. Fillmore, Charles J. and Paul Kay. 1999. Grammatical constructions and linguistic generalizations: The What’s X doing Y? construction. Language 75(1): 1–33. Fillmore, Charles J., Paul Kay, and Mary Catherine O’Connor. 1988. Regularity and idiomaticity in grammatical constructions: the case of let alone. Language 63(3): 501–538. Flaherty, Alice W. 2004. The Midnight Disease: The Drive to Write, Writer’s Block, and the Creative Brain. Boston/New York: Houghton Mifflin Company. Fleischman, Suzanne. 1989. Temporal distance: a basic linguistic metaphor. Studies in Language 13(1): 1–50. 1990. Tense and Narrativity: From Medieval Performance to Modern Fiction. London: Routledge. Fludernik, Monika. 1993. The Fictions of Language and the Languages of Fiction. London and New York: Routledge. 1994. Second-person narrative as a test case for narratology: the limits of realism. Style 28(3): 445–480. 1997. Towards a “Natural” Narratology. London and New York: Routledge. 2003. Natural narratology and cognitive parameters. In Herman (ed.), 243–270. Fodor, Jerry A. 1983. The Modularity of Mind: An Essay on Faculty Psychology. Cambridge, Mass.: MIT Press. Frank, Roslyn M., Rene Dirven, Tom Ziemke, and Enrique Bernardez (eds.). 2008. Body, Language, and Mind (2 vols.), vol. 2: Sociocultural Situatedness. Berlin / New York: Mouton de Gruyter. Freeman, Margaret H. 2000. Poetry and the scope of metaphor: toward a cognitive theory of literature. In Antonio Barcelona (ed.) Metaphor and Metonymy at the Crossroads: a Cognitive Perspective. Berlin: Mouton de Gruyter, 253–281. 2005. The poem as complex blend: conceptual mappings of metaphor in Sylvia Plath’s “The Applicant.” Language and Literature 14(1): 25–44. Galbraith, Mary. 1995. Deictic shift theory and the poetics of involvement in narrative. In Duchan, Bruder, and Hewitt (eds.), 19–59. Gallagher, Shaun. 2005. How the Body Shapes the Mind. New York: Oxford University Press. Gallagher, Shaun and Daniel D. Hutto. 2008. Understanding others through primary interaction. In Zlatev et al. (eds.),17–38. Gallese, Vittorio, and George Lakoff. 2005. The brain’s concepts: the role of the sensorimotor system in conceptual knowledge. Cognitive Neuropsychology 22: 455–479. Gavins, Joanna. 2003. Too much blague? an exploration of the text worlds of Donald Barthelme’s Snow White. In Gavins and Steen (eds.), 129–144. 2007. Text World Theory: An Introduction. Edinburgh University Press. Gavins, Joanna and Gerard Steen (eds.) 2003. Cognitive Poetics in Practice. London and New York: Routledge.
References
215
Genette, Gérard. 1980. Narrative Discourse: An Essay in Method. Translated by J. E. Lewin. Ithaca: Cornell University Press. Gerrig, Richard. 1993. Experiencing Narrative Worlds. New Haven: Yale University Press. Gerrig, Richard and Giovanni Egidi. 2003. Cognitive psychological foundations of narrative experiences. In Herman (ed.), 33–55. Gibbs, Raymond W. Jr. 1994. The Poetics of Mind: Figurative Thought, Language, and Understanding. New York: Cambridge University Press (also translated into Italian, Korean, Spanish, and Japanese). 1999. Intentions in the Experience of Meaning. New York: Cambridge University Press. 2000. Making good psychology out of blending theory. Cognitive Linguistics 11(3/4): 347–358. 2005a. Embodiment and Cognitive Science. Cambridge University Press. 2005b. The psychological status of image schemas. In Hampe (ed.), 113–136. Gibbs, Raymond W. Jr. and Teenie Matlock. 1999. Psycholinguistics and mental representations. Cognitive Linguistics 10(3): 263–270. Goldberg, Adele E. 1995. Constructions: A Construction Grammar Approach to Argument Structure. University of Chicago Press. Gottschall, Jonathan. 2008a. The Rape of Troy: Evolution, Violence and the World of Homer. Cambridge University Press. 2008b. Literature, Science, and a New Humanities. New York: Palgrave Macmillan. Gottschall, Jonathan and David Sloan Wilson (eds.). 2005. The Literary Animal: Evolution and the Nature of Narrative. Evanston, Ill.: Northwestern University Press. Graver, David. 1997. The actor’s bodies. Text and Performance Quarterly 17(3): 221–235. Haiman, John. 1980. The iconicity of grammar: isomorphism and motivation. Language 56: 515–540. Hamilton, Craig. 2002. Conceptual integration in Christine de Pizan’s City of Ladies. In Semino and Culpeper (eds.), 1–22. Hampe, Beate (ed.). 2005. From Perception to Meaning: Image Schemas in Cognitive Linguistics. Berlin: Mouton de Gruyter. Hanks, William F. 1990. Referential Practice: Language and Lived Space among the Maya. Chicago University Press. Harding, Jennifer Riddle. 2004. Simple regrets: counterfactuals and the dialogic mind. PhD dissertation, University of Maryland. 2007. Evaluative stance and counterfactuals in language and literature. Language and Literature 16(3): 263–280. Harnad, S. 1990. The symbol grounding problem. Physica 42: 335–346. Harris, Jonathan and Natasha Korda (eds.). 2002. Staged Properties in Early Modern English Drama. Cambridge University Press. Hart, Elizabeth and Bruce McConachie (eds.). 2006. Performance and Cognition: Theatre Studies and the Cognitive Turn. New York: Routledge. Herman, David. 2002. Story Logic: Problems and Possibilities of Narrative. Lincoln, Nebr., and London: University of Nebraska Press. 2003a. Stories as a tool for thinking. In Herman (ed.), 163–193. (ed.) 2003b. Narrative Theory and the Cognitive Sciences. Stanford: Center for the Study of Language and Information.
216
References
2006. Genette meets Vygotsky: narrative embedding and distributed intelligence. Language and Literature 15(4): 357–380. (ed.) 2007. The Cambridge Companion to Narrative. Cambridge University Press. Hiraga, Masako. 2005. Metaphor and Iconicity. New York: Palgrave Macmillan. Hogan, Patrick Colm. 2003. The Mind and Its Stories: Narrative Universals and Human Emotion. Cambridge University Press. Humphrey, Robert. 1954. Stream of Consciousness in the Modern Novel. Berkeley: University of California Press. Hutchins, Edwin. 1995. Cognition in the Wild. Cambridge, Mass.: MIT Press. 2005. Material anchors for conceptual blends. Journal of Pragmatics 37: 1,555–1,577. Hutto, Daniel D. (ed.). 2007a. Narrative and Understanding Persons. Cambridge University Press. 2007b. The Narrative Practice Hypothesis: origins and applications of folk psychology. In Hutto (ed.), 43–68. Iser, Wolfgang. 1974. The Implied Reader: Patterns of Communication in Prose Fiction from Bunyan to Beckett. Baltimore: Johns Hopkins University Press. Jackendoff, Ray. 1975. On belief contexts. Linguistic Inquiry 6(1). Jackson, Tony E. 2005. Explanation, interpretation, and close reading: the progress of cognitive poetics. Poetics Today 26(3): 519–532. Johnson, Mark. 1987. The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason. Chicago University Press. 2007. The Meaning of the Body: Aesthetics of Human Understanding. Chicago University Press. Klein, Kitty. 2003. Narrative construction, cognitive processing, and health. In Herman (ed.), 56–84. Labov, William. 1972. Language in the Inner City. University of Philadelphia Press. Labov, William and Joshua Waletzky. 1967. Narrative analysis: oral versions of personal experience. In June Helm (ed.) Essays on the Verbal and Visual Arts. Seattle: University of Washington Press, 12–44. Lakoff, George. 1996. Sorry, I’m not myself today: the metaphor system for conceptualizing self. In Fauconnier and Sweetser (eds.), 91–123. 2008. The Neural Theory of Metaphor. In Raymond Gibbs (ed.) Cambridge Handbook of Metaphor and Thought. New York: Cambridge University Press, 17–38. Lakoff, George and Mark Johnson. 1980. Metaphors We Live By. University of Chicago Press. 1999. Philosophy in the Flesh: The Embodied Mind and Its Challenge to Western Thought. New York: Basic Books. Lakoff, George and Mark Turner. 1989. More than Cool Reason: A Field Guide to Poetic Metaphor. University of Chicago Press. Langacker, Ronald W. 1990a. Concept, Image, Symbol: The Cognitive Basis of Grammar. Berlin: Mouton de Gruyter. 1990b. Subjectification. Cognitive Linguistics 1: 5–38. 1991 [1987]. Foundations of Cognitive Grammar, 2 vols. Stanford University Press. 2001. Discourse in cognitive grammar. Cognitive Linguistics 12: 143–188. 2005. Dynamicity, fictivity, and scanning: The imaginative basis of logic and linguistic meaning. In Pecher and Zwaan (eds.), 164–198.
References
217
Lanser, Susan Sniader. 1981. The Narrative Act: Point of View in Fiction. Princeton University Press. Leech, Geoffrey N. and Michael H. Short. 1981. Style in Fiction: A Linguistic Introduction to English Fictional Prose. London: Longman. Lehrer, Adrienne and Eva Feder Kittay (eds.). 1992. Frames, Fields, and Contrasts: New Essays in Semantic and Lexical Organization. Hillsdale, NJ: L. Erlbaum Associates. Lermitte, Janet. 2010. Faith and art, frames and anchors: Annie Dillard’s Holy the Firm. Paper presented at the PALA Conference in Genoa, Italy. Lupton, Christina. 2006. The knowing book: authors, it-narratives, and objectification in the eighteenth century. Novel 39(3): 402–420. Lyons, John. 1977. Semantics. Cambridge University Press. Maitre, Doreen. 1983. Literature and Possible Worlds. London: Middlesex Polytechnic Press. Mar, Raymond A. and Keith Oatley. 2008. The function of fiction is the abstraction and simulation of social experience. Perspectives on Psychological Science 3: 173–192. Margolin, Uri. 1996. Telling our story: on “we” literary narratives. Language and Literature 5(2): 115–133. 2003. Cognitive science, the thinking mind, and literary narrative. In Herman (ed.), 271–294. Matlock, Teenie. 2004. Fictive motion as cognitive simulation. Memory and Cognition 32: 1,389–1,400. McGregor, William B. 1997. Semiotic Grammar. Oxford: Clarendon Press. McNeill, David (ed.). 2000. Language and Gesture. Cambridge University Press. Michaelis, L. 2003. Word meaning, sentence meaning, and syntactic meaning. In Hubert Cuyckens, Rene Dirven, and John Taylor (eds.) Cognitive Approaches to Lexical Semantics. Berlin/New York: Mouton de Gruyter, 163–210. Moder, Carol. 2010. Two puzzle pieces: fitting discourse context and constructions into Cognitive Metaphor Theory. English Text Construction 3(2): 294–320. Moschovakis, Nicholas. 2006. Topicality and conceptual blending: Titus Andronicus and the case of William Hacket. College Literature 33(1): 127–149. Moss, John and Tobi Kozakiewich (eds.). 2006. Margaret Atwood: The Open Eye. University of Ottawa Press. Murakami, Haruki and Philip Gabriel. 2008. What I Talk about When I Talk about Running. New York: Alfred Knopf/Random House. Nikiforidou, Kiki. 2010. Viewpoint and construction grammar: The case of past+now. Language and Literature 19(3): 265–284. Noh, Eun-Ju. 2000. Metarepresentation: A Relevance-Theory Approach. Amsterdam: John Benjamins. Nuñez, Raphael E. and Eve Sweetser. 2006. With the future behind them: convergent evidence from Aymara language and gesture in the crosslinguistic comparison of spatial construals of time. Cognitive Science 30: 401–450. Oakley, Todd. 1998. Conceptual blending, narrative discourse, and rhetoric. Cognitive Linguistics 9(4): 321–360. 2009. From Attention to Meaning: Explorations in Semiotics, Linguistics, and Rhetoric. Berne: Peter Lang AG. Oakley, Todd and Anders Hougaard. 2008. Mental Spaces in Discourse and Interaction. Amsterdam and Philadelphia: John Benjamins.
218
References
Palmer, Alan. 2003. The mind beyond the skin. In Herman (ed.), 322–348. 2004. Fictional Minds. Lincoln, Nebr., and London: University of Nebraska Press. Panther, Klaus-Uwe. 2006. Metonymy as a usage event. In Gitte Kristiansen, Michel Achard, Rene Dirven, and Francisco Ruiz de Mendoza Ibañez (eds.) Cognitive Linguistics: Current Applications and Future Perspectives. Berlin and New York: Mouton de Gruyter, 147–186. Pascual, Esther. 2002. Imaginary Trialogues: Conceptual Blending and Fictive Interaction in Criminal Courts. Utrecht: LOT. 2006a. Fictive interaction within the sentence. Cognitive Linguistics 17(2): 245–267. 2006b. Questions in legal monologues: fictive interaction as argumentative strategy in a murder trial. Text & Talk 26(3): 383–402. 2008. Fictive interaction blends in everyday life and courtroom settings. In Anders Hougaard and Todd Oakley (eds.) Mental Spaces in Discourse and Interaction. Amsterdam: John Benjamins, 79–107. Pavel, Thomas G. 1986. Fictional Worlds. Cambridge, Mass., and London: Harvard University Press. Pecher, Diane and Rolf Zwaan. 2005a. Introduction. In Pecher and Zwaan (eds.), 1–8. (eds.) 2005b. Grounding Cognition: The Role of Perception and Action in Memory, Language, and Thinking. Cambridge University Press. Perrakis, Phyllis. 2006. Negotiating with the looking glass: Atwood, her protagonists, and the journey to the dead. In Moss and Kozakiewich (eds.), 349–359. Pinker, Steven. 1994. The Language Instinct. New York: William Morrow and Company. 2002. The Blank Slate: The Modern Denial of Human Nature. New York: Penguin. Pires de Oliveira, Roberta, and Robson de Souza Bittencourt. 2008. An interview with Mark Johnson and Tim Rohrer: from neurons to sociocultural situatedness. In Frank et al. (eds.), 21–52. Porter Abbott H. 2007. Story, plot, and narration. In Herman (ed.), 39–52. Pratt, Mary Louise. 1977. Toward a Speech Act Theory of Literary Discourse. Bloomington: Indiana University Press. Prince, Gerald. 2001. A point of view on point of view or refocusing focalization. In Van Peer and Chatman (eds.), 43–64. Prinz, Jesse J. 2002. Furnishing the Mind: Concepts and Their Perceptual Basis. Cambridge, Mass.: MIT Press. 2004. Gut Reactions: A Perceptual Theory of Emotion. New York: Oxford University Press. 2005. Passionate thoughts: the emotional embodiment of moral concepts. In Pecher and Zwaan (eds.), 93–114. Propp, Vladimir. 1968 [1928]. Morphology of the Folktale. Translated by Laurence Scott. Austin: University of Texas Press. Rauh, Gisa. 1981. Aspects of deixis. In Gisa Rauh (ed.) Essays on Deixis. Tübingen: Gunter Narr Verlag, 9–60. Reddy, Michael. 1979. The conduit metaphor – a case of frame conflict in our language about language. In Andrew Ortony (ed.) Metaphor and Thought. Cambridge University Press, 284–297. Richerson, Peter J. and Robert Boyd. 2005. Not by Genes Alone: How Culture Transforms Human Evolution. Chicago University Press. Rimmon-Kenan, Shlomith. 1983. Narrative Fiction: Contemporary Poetics. London: Methuen.
References
219
Ritchie, David. 2004. Lost in “conceptual space”: metaphors of conceptual integration. Metaphor and Symbol 19(1): 31–50. Rohrer, Tim. 2005. Mimesis, artistic inspiration, and the blends we live by. Journal of Pragmatics 37: 1,686–1,716. Ronen, Ruth. 1994. Possible Worlds in Literary Theory. Cambridge University Press. Roy, Wendy. 2006. The body of/as evidence: Margaret Atwood, The Blind Assassin, and the feminist literary mystery. In Moss and Kozakiewich (eds.), 361–371. Rubba, Jo. 1996. Alternate grounds in the interpretation of deictic expressions. In Fauconnier and Sweetser (eds.), 227–261. 2009. The dream as a blend in David Lynch’s Mulholland Drive. In Vyvyan Evans, and Stephanie Pourcel (eds.) New Directions in Cognitive Linguistics. Amsterdam and Philadelphia: John Benjamins Publishing Company, 465–498. Ryan, Marie-Laure. 1991. Possible Worlds, Artificial Intelligence, and Narrative Theory. Bloomington and Indianapolis: Indiana University Press. 2007. Toward a definition of narrative. In Herman (ed.), 22–38. Sanchez Garcia, Jesus. 2003. Language and cognition: George Lakoff on some internal and external complexities. An interview. In Francisco J. Ruiz de Mendoza (ed.), Annual Review of Cognitive Linguistics 1: 233–267. Sanders, José and Gisela Redeker. 1996. Perspective and representation in speech and thought in narrative discourse. In Fauconnier and Sweetser (eds.), 290–317. Sanders, Ted, José Sanders and Eve Sweetser. 2009. Causality, cognition, and communication: a mental space analysis of subjectivity in causal connectives. In Ted Sanders and Eve Sweetser (eds.) Causal Categories in Discourse and Cognition. Berlin/New York: Mouton de Gruyter, 19–60. Sanger, Keith. 2001. The Language of Drama. London: Routledge. Scalise Sugiyama, Michelle. 2005. Reverse-engineering narrative: evidence of special design. In Gottschall and Wilson (eds.), 177–195. Scalise Sugiyama, Michelle and Lawrence S. Sugiyama (in press). “‘Once the child is lost he dies’”: Monster stories vis-à-vis the problem of errant children. In Mark Collard and Edward Slingerland (eds.) Creating Consilience. Oxford University Press, 363–371. Schnepf, Jennifer. 2006. Role-value mappings and narrative structure in Atwood’s The Blind Assassin. Paper presented at the Literature and Cognitive Science Conference, Storrs, Conn., April 6–9, 2006. Semino, Elena. 1997. Language and World Creation in Poems and Other Texts. London: Longman. 2002. A cognitive stylistic approach to mind style in narrative fiction. In Semino and Culpeper (eds.), 95–122. 2003. Discourse worlds and mental spaces. In Gavins and Steen (eds.), 83–98. 2006. Blending and characters’ mental functioning in Virginia Woolf’s Lappin and Lapinova. Language and Literature 15(1): 55–72. 2007. Mind style 25 years on. Style 41(2): 153–203. 2008. Metaphor in Discourse. Cambridge University Press. 2009. Text worlds. In Geert Brône and Joeren Vandaele (eds.) Cognitive Poetics: Goals, Gains, and Gaps. Berlin/New York: Mouton de Gruyter, 33–72. 2010. Unrealistic scenarios, metaphorical blends, and rhetorical strategies across genres. English Text Construction 3(2): 250–274.
220
References
Semino, Elena and Jonathan Culpeper (eds.). 2002. Cognitive Stylistics: Language and Cognition in Text Analysis. Amsterdam: John Benjamins. Semino, Elena and Michael Short. 2004. Corpus Stylistics: Speech, Writing, and Thought Presentation in a Corpus of English Writing. London: Routledge. Sharifian, Farzad. 2008. Distributed, emergent cultural cognition, conceptualization and language. In Frank et al. (eds.), 109–136. Short, Mick. 2007. Thought presentation twenty-five years on. Style 41(2): 227–240. Simpson, Paul. 1993. Language, Ideology, and Point of View. London and New York: Routledge. Slingerland, Edward. 2005. Conceptual blending, somatic marking, and normativity: a case example from ancient Chinese. Cognitive Linguistics 16(3): 557–584. 2006. What Science Offers the Humanities: Integrating Body and Culture. Cambridge and New York: Cambridge University Press. Sofer, Andrew. 2003. The Stage Life of Props. Ann Arbor: University of Michigan Press. Spivey, Michael J. 2007. The Continuity of Mind. New York: Oxford University Press. Spivey, Michael J., Daniel C. Richardson, and Monica Gonzalez-Marquez. 2005. On the spatial and image-schematic underpinnings of real-time language processing. In Pecher and Zwaan (eds.), 246–281. Stockwell, Peter. 2002. Cognitive Poetics: An Introduction. London: Routledge. Style. 2008. Style, 42 (2/3), special issue on literary Darwinism. Surma, Szymon. 2009. The significance of ghosts in nō and kabuki: an analysis of stories and forms. Unpublished manuscript. Sweetser, Eve. 1990. From Etymology to Pragmatics. Cambridge University Press. 1996. Mental spaces and the grammar of conditional constructions. In Fauconnier and Sweetser (eds.), 318–333. 1999. Compositionality and blending: semantic composition in a cognitively realistic framework. In Theo Janssen and Gisela Redeker (eds.) Cognitive Linguistics: Foundations, Scope, and Methodology. Berlin: Mouton de Gruyter, 129–162. 2000. Blended spaces and performativity. Cognitive Linguistics 11(3–4): 305–334. 2004. “The suburbs of your good pleasure”: cognition, culture, and the bases of metaphoric structure. The Shakespearean International Yearbook 4, 24–55. 2006. Whose rhyme is whose reason? Sound and sense in Cyrano de Bergerac. Language and Literature 15(1): 29–54. Talmy, Leonard. 1988. Force dynamics in language and thought. Cognitive Science 12: 49–100. 1996. Fictive motion in language and “ception”. In Paul Bloom, Mary A. Peterson, Lynn Nadel, and Merrill F. Garrett (eds.) Language and Space. Cambridge, Mass.: MIT Press, 211–276. 2000. Towards a Cognitive Semantics (vols. 1 & 2). Cambridge, Mass.: MIT Press. Tobin, Vera. 2006. Ways of reading Sherlock Holmes. Language and Literature 16(1) 73–90. 2010. Joint attention, To the Lighthouse, and modernist representations of intersubjectivity. English Text Construction 3(2): 185–202. Tomasello, Michael. 1999. The Cultural Origins of Human Cognition. Cambridge, Mass.: Harvard University Press. 2008. Origins of Human Communication. Cambridge, Mass.: MIT Press.
References
221
Tooby, John and Leda Cosmides (in press). Evolutionary Psychology: Foundational Papers. Cambridge, Mass.: MIT Press. Toolan, Michael. 2001. Narrative: a Critical Linguistic Introduction. (2nd edn.) London and New York: Routledge. Tsur, Reuven. 1992. Toward a Theory of Cognitive Poetics. Amsterdam: Elsevier. 2003. On the Shore of Nothingness: A Study in Cognitive Poetics. Charlottesville: Imprint Academic. Turner, M. 1991. Reading Minds: The Study of English in the Age of Cognitive Science. Princeton University Press. 1996. The Literary Mind: The Origins of Thought and Language. New York: Oxford University Press. 2003. Double-scope stories. In Herman (ed.), 117–142. 2004. The ghost of anyone’s father. The Shakespearean International Yearbook 4: 72–97. 2006a. Compression and representation. Language and Literature 15(1): 17–29. (ed.) 2006b. The Artful Mind: Cognitive Science and the Riddle of Human Creativity. Oxford University Press. Turner, Mark and Gilles Fauconnier. 1995. Conceptual integration and formal expression. Metaphor and Symbolic Activity 10(3): 183–203. 1999. A mechanism of creativity. Poetics Today 20(3): 397–418. Vandelanotte, Lieven. 2002. But forced to qualify: distancing speech and thought representation as a symptom of uninformedness in Larkin. Leuvense Bijdragen (Leuven Contributions in Linguistics and Philology) 91(3–4): 383–426. 2004a. Deixis and grounding in speech and thought representation. Journal of Pragmatics 36(3): 489–520. 2004b. From representational to scopal “distancing indirect speech or thought”: a cline of subjectification. Text 24(4): 547–585. 2005. Types of speech and thought representation in English: syntagmatic structure, deixis and expressivity, semantics. PhD dissertation. University of Leuven. 2009. Speech and Thought Representation in English: A Cognitive-Functional Approach. Berlin: Mouton de Gruyter. 2010. “Where am I, lurking in what place of vantage?” The discourse of distance in John Banville’s fiction. English Text Construction 3(2) 203–225. Van Hoek, Karen. 1997. Anaphora and Conceptual Structure. University of Chicago Press. Van Oort, Richard. 2003. Cognitive science and the problem of representation. Poetics Today 24: 237–295. 2007. The culture of criticism. Criticism 49(4): 459–479. Van Peer, Willie and Seymour Chatman (eds.). 2001. New Perspectives on Narrative Perspective. Albany: State University of New York Press. Van Vliet, Sarah. 2008. Proper nouns and pronouns: the production of referential expressions in narrative discourse. PhD dissertation. Publications of the Netherlands Graduate School of Linguistics. Vargas Llosa, Mario. 1997/2003 [1997]. Letters to a Young Novelist. Translated by Natasha Wimmer. New York: Picador Books. Verhagen, Arie. 2005. Constructions of Intersubjectivity: Discourse, Syntax, and Cognition. Oxford University Press.
222
References
2008. Intersubjectivity and the architecture of the language system. In Zlatev et al., (eds.), 307–332. Wales, Katie. 1996. Personal Pronouns in Present Day English. Cambridge University Press. Weber, Jean Jacques. 2004. A new paradigm for literary studies, or: the teething troubles of cognitive poetics. Style 38(4): 515–523. Weimann, Robert. 2000. Author’s Pen and Actor’s Voice: Playing and Writing in Shakespeare’s Theatre. Cambridge University Press. Werth, Paul. 1999. Text Worlds: Representing Conceptual Space in Discourse. London: Longman. Wetmore, Kevin J., Jr. 2008. “Avenge me!”: ghosts in English Renaissance and Kabuki revenge dramas. In Wetmore (ed.) Revenge Drama in European Renaissance and Japanese Theatre: From Hamlet to Madame Butterfly. New York: Palgrave Macmillan, 75–90. Wiebe, Janyce. 1995. References in narrative text. In Duchan et al. (eds.), 263–285. Wilson, Sharon. 2006. Blindness and survival in Margaret Atwood’s major novels. In C. A. Howells (ed.) The Cambridge Companion to Margaret Atwood. Cambridge University Press, 176–190. Yachnin, Paul. 2002. Wonder-effects: Othello’s handkerchief. In Harris and Korda (eds.), 316–334. Zbikowski, Lawrence. 2002. Conceptualizing Music. New York: Oxford University Press. 2006. The cognitive tango: music, mind, and brain. In Turner (ed.), 115–132. Zlatev, Jordan. 2008. The co-evolution of intersubjectivity and bodily mimesis. In Zlatev et al., (eds.), 215–244. Zlatev, Jordan, Timothy P. Racine, Chris Sinha, and Esa Itkonen (eds.). 2008. The Shared Mind: Perspectives on Intersubjectivity. Amsterdam and Philadelphia: John Benjamins Publishing Company. Zunshine, Lisa. 2006. Why We Read Fiction: Theory of Mind and the Novel. Columbus: Ohio State University Press.
Literary works cited
Amis, Martin. 1991. Time’s Arrow. New York: Vintage International. Atwood, Margaret. 2000. The Blind Assassin. Toronto: McClelland and Stewart. Austen, Jane. 2009. Pride and Prejudice. London: Penguin Classics. Ballard, J. G. 1984. Empire of the Sun. New York: Simon and Schuster. 1994/1995. Rushing to Paradise. London: Flamingo. 2008. The Miracles of Life. London: Fourth Estate. Bulgakov, Mikhail. 1966/1997. The Master and Margarita. Translated by Richard Pevear and Larissa Volokhonsky. London: Penguin Books. Calvino, Italo. 1979/1981. If on a Winter’s Night a Traveler. Translated by William Weaver. Toronto, Canada: Lester & Orpen Dennys. Chatwin, Bruce. 1988. Utz. New York: Penguin Books. Coetzee, J. M. 2003. Elizabeth Costello. London: Secker and Warburg. 2005. Slow Man. London: Secker and Warburg. 2007. Diary of a Bad Year. London: Harvill Secker. Dillard, Annie. 1977/1988. Holy the Firm. New York: Harper Perennial. Eggers, Dave. 2000. A Heartbreaking Work of Staggering Genius. Toronto: Vintage Canada. Fallaci, Oriana. 1979/1980. The Man. Translated by William Weaver. New York: Simon and Schuster. Fielding, Henry. 2000. The History of Tom Jones. Cologne: Köneman. Fitzgerald, F. Scott. 2004 [1925]. The Great Gatsby. New York: Scribner. García Márquez, Gabriel. 2006 [1970]. One Hundred Years of Solitude. Translated by Gregory Rabassa. New York: Harper Perennial. Garland, Alex. 1999. The Tesseract. New York: Riverhead Books. Grass, Günter. 2009. The Tin Drum. Translated by Breon Mitchell. Boston/New York: Houghton Mifflin Harcourt. James, Henry. 1963 [1881]. The Portrait of a Lady. London: Penguin Books. Joyce, James. 2000. Dubliners. Oxford University Press. Marlowe, Christopher. 2007. Dr Faustus (edited by Michael Keefer). Toronto: Broadview Editions. McEwan, Ian. 2001/2002. Atonement. Toronto: Vintage Canada. Melville, Herman. 1998 [1851]. Moby-Dick. New York: New American Library. Miller, Henry. 1941/1958 [1941]. The Colossus of Maroussi. New York: New Directions Books. Murakami, Haruki. 2007/2008. After Dark. Translated by Jay Rubin. Toronto: Vintage Canada Nabokov, Vladimir. 1991[1955]. Lolita (annotated edition). New York: Vintage Books. 223
224
Literary works cited
Paretsky, Sara. 1999. Hard Time. New York: Dell Publishing. Philip, Neil. 1994. The Arabian Nights. New York: Orchard Books. Potocki, Jan. 1995. The Manuscript Found in Saragossa. Translated by Ian Maclean. London: Penguin Books. Powers, Richard. 2009. Generosity: An Enhancement. New York: Farrar, Straus and Giroux. Proulx, Annie. 1993. The Shipping News. New York: Scribner. 1985. Foreign Land. London: Picador Books. 1986. Coasting. London: Picador/Pan Books. 1987 [1979]. Arabia through the Looking Glass. London: Picador Books. 1991. Hunting Mister Heartbreak. New York: Harper Collins. Raban, Jonathan. 1998 [1981]. Old Glory: An American Voyage. New York: Vintage Books. 2000 [1999]. Passage to Juneau. New York: Vintage Books. Richardson, Samuel. 1747/1985. Clarissa. London: Penguin Books. de Riojas, Fernando. 1962. The Celestina. Translated by Lesley Byrd Simpson. Berkeley and Los Angeles: University of California Press. Roth, Philip. 1997/1998. American Pastoral. New York: Vintage International. 2000. The Human Stain. New York: Vintage International. Rushdie, Salman. 1980. Midnight’s Children. New York: Penguin Books. Shakespeare, William. 1994. The Complete Works. The Shakespeare Head Press Oxford Edition. New York: Barnes & Noble Books. (Macbeth, Romeo and Juliet, Richard II, Othello) Shields, Carol. 1997. Larry’s Party. Viking Penguin. Stephenson, Neal. 1992/1993. Snow Crash. New York: Bantam Books. Stoppard, Tom. 1993. Arcadia. London/New York: Faber and Faber. Szymborska, Wisława. 1997. Nothing Twice: Selected Poems / Nic Dwa Razy: Wybór Wierszy. Translated by Stanisław Barańczak and Clare Cavanagh. Cracow: Wydawnictwo Literackie. Theroux, Paul. 1978. The Picture Palace. New York: Penguin Books. Tolkien, J. R. R. 1954. The Fellowship of the Ring. Boston: Houghton Mifflin Company. Tyler, Anne. 1995. Ladder of Years. New York: Fawcett Columbine. 1998. A Patchwork Planet. Toronto: Penguin Books. Woolf, Virginia. 1953 [1925]. Mrs. Dalloway. New York: Harcourt Brace Jovanovich Publisher.
Index
Acting/Speaking blend, 156, 158 agency, 137, 163 Atwood, Margaret, 10, 40, 50 authorship, 48, 58, 79, 82, 83, 84, 135 Bakhtin, Mikhail, 165 Ballard, J. G., 22, 23, 128 blending, 7, 31–35, 38, 56, 189, 201 compression, 18, 59, 70, 93 optimality constraints, 57 selection, 58 vital relations, 56 vs embedding, 190 Body/Mind blend, 152, 158, 159 Booth, Wayne, 74, 87 Boyd, Brian, 21 Bulgakov, Mikhail, 65, 81, 83, 86 Calvino, Italo, 76 Canning, Patricia, 9 Carroll, Joseph, 27 Chatman, Seymour, 53 Chatwin, Bruce, 60 Chekhov, Anton, 143 Cienki, Alan, 151 Clark, Herbert, 177 Coetzee, J. M., 79, 80, 119, 168 cognitive linguistics, 6, 31, 195 conceptual distance, 72–73, 74, 75, 99 conceptual integration, See blending constructional compositionality, 8, 10, 33, 75–76, 193, 198 constructions, 20, 88, 105, 121, 140, 166, 194, 198 vocative-cum-imperative, 153–160, 198 cross-input projection, 40, 42, 43, 44, 46, 52 Currie, Gregory, 29, 115 Damasio, Antonio, 14 de Riojas, Fernando, 165 Deacon, Terrence, 7, 14, 24, 28, 50
decompression for viewpoint, 100–102, 118, 120, 133 default communicative context, 19–21, 31, 65, 129, 139–141 deictic ground, 20, 31, 64, 68, 76, 117, 136, 139–140, 165, 183, 199 Deictic Shift Theory, 129, 140 deixis, See deictic ground Dennett, Daniel, 20 Dillard, Annie, 43 discourse blends, 182 DIST, See speech and thought representation distanced discourse, 193 distributed cognition, 14, 18, 109 Donald, Merlin, 14, 16, 28 drama, 162–163 and blending, 155–159 and mental spaces, 141–145 and supernatural phenomena on stage, 160–164 and traditional Japanese theatre, 162–163 body posture, 149–152 discourse of, 152–160 narration, 147–148 spatial construals, 142–144, 148–152 vs early novels, 164–168 dual voice, 88, 111 Eggers, Dave, 38, 60, 62, 73, 74, 173, 177, 186 embodiment, 25, 26, 29, 113, 151, 158, 168–169 emergence of literature, 16–19 Emmott, Catherine, 30, 51, 54 epistemic stance, 22, 55, 89, 90 epistemic viewpoint, 48, 59, 61, 62, 65, 70, 85, 132 evidentiality, 17, 22 evocation, 44, 47, 49, 99, 124, 143, 146 evolutionary psychology, 11, 13, 28, 109 eye contact, 144
225
226
Index
fact vs fiction, 22–23 Fallaci, Oriana, 76 Fauconnier, Gilles, 31, 33, 201 Faulkner, William, 74 Feeling/Speaking blend, 157 fictive interaction, 182 fictive motion, 103, 117, 128 fictive vision, 102–103, 104, 107, 131, 201 Fielding, Henry, 64, 191 Fillmore, Charles, 32 FIST, See speech and thought representation flashbacks, 6, 37, 54, 63, 100, 189 Fludernik, Monika, 6, 78, 88 focalization, 87, 108 frame evocation, 46, 157 frame metonymy, 33, 50, 177 frames, 32–33, 35, 43, 48, 50, 51, 52, 54, 101, 120, 121, 135, 138, 143, 156, 201 Gallagher, Shaun, 113 García Márquez, Gabriel, 119 Garland, Alex, 40, 107 Gavins, Joanna, 36 Genette, Gérard, 60, 87, 91 Gerrig, Richard, 18, 54, 177 gesture, 149, 151, 155, 158 Gibbs, Raymond, 59 Grass, Günter, 132, 168 Graver, David, 147 Ground, 19 grounding problem, 24, 25 Herman, David, 6, 15, 23, 24, 27, 53, 54, 109 Hutchins, Edwin, 158 Hutto, Daniel, 113, 114 image schemas, 24, 26, 27, 28 imperative, 154, 193 intentional stance, 20 internal monologue, 67, 133, 167, 179 intersubjectivity, 112–116, 140, 141, 164, 200 Iser, Wolfgang, 58 it-narratives, 166 James, Henry, 66 Johnson, Mark, 55, 149 joint attention, 164 Joyce, James, 191 Labov, William, 56 Lakoff, George, 55, 100, 149 Lanser, Susan, 115 Leech, Geoffrey, 103
Literary Darwinists, 13 literary interpretation, 9–11 literature and art, 14–16 Lupton, Christina, 166 Margolin, Uri, 130 Marlowe, Christopher, 159 material anchors, 158, 163, 166 material objects, 136–138, 146–147, 152–161, 166, 197 McEwan, Ian, 84 meaning construction, 4–6, 10, 25, 35, 145, 202 Melville, Herman, 166, 172, 197 mental spaces, 32, 35, 36, 39, 94, 141–145 and theatre, 141–145 metaphor, 149, 151, 153, 155, 158, 164 metaphors of time, 55 metonymy, 149, 150, 151, 162, 168 micro/macro viewpoint, 62–63, 87, 133 Miller, Henry, 187 mind style, See narrator, unreliable mirror neurons, 113 Moder, Carol, 199 Monterroso, Augusto, 37 Murakami, Haruki, 21, 131 mythic past, 17 Nabokov, Vladimir, 119 narratee, 68, 76 narration, 53, 63, 64, 69, 73, 75, 111, 167, 191, 200 narrative anchors, 42–50, 82, 144, 147, 158, 163, 166, 197 narrative and cognition, 23–27 narrative and evolution, 27–28 Narrative Practice Hypothesis, 114, 115 narrative spaces, 36–37, 38, 40–42, 51, 53, 59, 61, 63, 73, 74, 76, 81, 83, 84, 86, 93, 127, 136, 140, 164, 172, 173, 188, 192 and meta-representation, 108–112 embedding, 84–86, 175, 178–183, 190 types and relations, 38, 63–67 narrative transport, 18 narrative uptake, 59 narrative viewpoint, 49, 61, 63, 67, 81, 84, 86, 88, 93, 115, 196 experiential viewpoint, 103–106, 111, 128 visual viewpoint, 91–94, 128, 131, 161 narrator as character, 60, 62, 63, 71, 78, 132, 135, 186 as epistemic viewpoint, 61, 64–70, 136 intrusive, 64, 68 multiple narrators, 80–84 off-stage, 66, 69, 174, 188 omniscient, 64, 66
Index on-stage, 65, 166, 167, 174, 180, 186, 189 types of, 60–67 unreliable, 61, 73 narratorship, 21, 31, 48, 58, 59, 74, 80, 82, 86, 133 natural narratology, 6 Nikiforidou, Kiki, 190 now, 69, 93, 125, 190–193 Oakley, Todd, 164 Oatley, Keith, 21 Odyssey, 54 omniscience, 59, 61, 64, 68–70 oral stories, 17–19, 56, 114 Palmer, Alan, 21, 29, 115 Paretsky, Sara, 110 Pascual, Esther, 182 performative blend, 154 person, See tense and person choices personal pronouns, 31, 51, 52, 64, 130, 133–136, 182, 184–189 Pinker, Steven, 11, 12, 13, 27 poetry, 136–138 possible worlds, 36 Potocki, Jan, 80, 84, 86 Powers, Richard, 133 Prinz, Jesse, 25 proper names, 34, 48, 52, 118–121 props, 147, 157 Proulx, E. Annie, 169 Raban, Jonathan, 91, 96, 118, 199 reader response, 58 Reddy, Michael, 203 referential expressions, 42, 51–53, 117–128 representation blend, 45–46, 92–96, 101, 102, 178 representational stance, 20, 50 Richardson, Samuel, 165 Rimmon-Kenan, Shlomith, 87, 90 role-descriptors, 119, 121–128 role-value mappings, See role-descriptors Roth, Philip, 120, 178, 180 Rubba, Johanna, 117, 139, 193 Rushdie, Salman, 168 Scalise Sugiyama, Michelle, 17, 18 scene vs summary, 90 science vs humanities, 12–13 semantic roles agent, 138, 157 experiencer, 138, 156, 162 instrument, 138 patient, 138
227 Semino, Elena, 74, 103, 199 sentences as spaces, 37–39 sequence of events, 37, 51, 53, 54–56 Shakespeare, William, 148, 152 Shared Mind theory, 112–115 Shields, Carol, 96 Short, Mick, 103 Slingerland, Edward, 12 Sofer, Andrew, 147 speaking for thinking, 174–178 speech and thought representation, 70, 88–91, 108, 110, 111, 159, 164, 169, 171–194, 200 constructional phenomena, 183–194 direct discourse, 88, 126, 141, 174, 181 Distancing Indirect Speech and Thought (DIST), 187 free indirect discourse, 69, 88, 107, 111, 183, 185 Free Indirect Speech and Thought (FIST), 187 indirect discourse, 183, 185 speech vs thought, 88, 103, 126, 171–178 Spivey, Michael, 27, 35 split-self metaphors, 100 Stoppard, Tom, 143 stories as tools for thinking, 15, 23 story construction, 37, 38, 39, 42, 50, 51, 53, 56–57, 68, 86, 91, 143, 152, 164, 188, 197 stream-of-consciousness, 70, 103, 169 structuralist narratology, 6 Sweetser, Eve, 8, 9, 10, 33, 55, 154, 193 symbolic representation, 25, 26, 27, 50 Szymborska, Wisława, 136, 200 Talmy, Leonard, 117, 162 telling vs showing, 87 tense and person choices, 8, 41, 63, 67, 68–79, 85, 132–133, 190 text vs story, 53–54 text world theory, 36, 51 theory of mind, 21, 28, 29, 58, 109, 112, 165 Theroux, Paul, 94 Thinking/Speaking blend, 158 time vs tense, 75 Tobin, Vera, 164 Tolkien, J. R. R., 128 Tomasello, Michael, 17, 20, 29, 114, 116, 164 Toolan, Michael, 87 Turner, Mark, 9, 24, 26, 27, 28, 31, 162, 201 Tyler, Anne, 66, 69, 120, 172, 181, 183
228
Index
van Oort, Richard, 13, 24, 26, 28 Vandelanotte, Lieven, 74, 88, 103, 171, 187, 193 Vargas Llosa, Mario, 22 Verhagen, Arie, 19, 114, 140, 141, 200 vertical integration, 12 viewpoint and focus, 39 viewpoint compression, 40, 70, 96–100, 106–108, 110, 112, 141, 175, 176, 178, 183, 189, 196 visual viewpoint, See narrative viewpoint
we, 53, 104, 130–132 Werth, Paul, 36, 51 Wetmore, Kevin, 162 Woolf, Virginia, 64, 67, 69, 75 Yachnin, Paul, 158 you narratives, 53, 76–79, 133 Zlatev, Jordan, 112 Zunshine, Lisa, 21, 28, 58, 109, 112, 165