Colm Hogan The Mind and Its Stories PDF

This page intentionally left blank

The Mind and Its Stories Narrative Universals and Human Emotion This book argues that there are profound, extensive, and surprising universals in literature that are bound up with universals in emotion. Professor Hogan maintains that debates over the cultural specificity of emotion have been misdirected because they have largely ignored a vast body of data that bear directly on the way different cultures imagine and experience emotion – literature. This is the first empirically and cognitively based isolation and discussion of narrative universals. Professor Hogan argues that, to a remarkable degree, the stories people admire in different cultures follow a limited number of patterns and that these patterns are determined by cross-culturally constant ideas about emotion. In formulating his argument, Professor Hogan draws on his extensive reading in world literature, experimental research treating emotion and emotion concepts, and methodological principles from the contemporary linguistics and the philosophy of science. He concludes with a discussion of the relations among narrative, emotion concepts, and the biological and social components of emotion. Patrick Colm Hogan is currently a professor in English, Comparative Literature, and Cognitive Science at the University of Connecticut. He is the author of nine books, including Colonialism and Cultural Identity: Crises of Tradition in the Anglophone Literatures of India, Africa, and the Caribbean (2000); The Culture of Conformism: Understanding Social Consent (2001); and the forthcoming Cognitive Science, Literature, and the Arts.

STUDIES IN EMOTION AND SOCIAL INTERACTION Second Series Series Editors Keith Oatley University of Toronto Antony S. R. Manstead University of Cambridge

This series is jointly published by Cambridge University Press and the Editions de la Maison des Sciences de l’Homme, as part of the joint publishing agreement established in 1977 between the Foundation de la Maison des Sciences de l’Homme and the Syndics of Cambridge University Press. Cette publication est publiée co-édition par Cambridge University Press et les Editions de la Maison des Sciences de l’Homme. Elle s’intègre dans le programme de co-édition e´ tabli en 1977 par la Foundation de la Maison des Sciences de l’Homme et les Syndics de Cambridge University Press. Titles published in the Second Series: The Psychology of Facial Expression James A. Russell and José Miguel Fernández-Dols Emotion, the Social Bond, and Human Reality: Part/Whole Analysis Thomas J. Scheff Intersubjective Communication and Emotion in Early Ontogeny Stein Br˚aten The Social Context of Nonverbal Behavior Pierre Philippot, Robert S. Feldman, and Erik J. Coats Communicating Emotion: Social, Moral, and Cultural Processes Sally Planalp Emotions across Languages and Cultures: Diversity and Universals Anna Wierzbicka Gender and Emotion: Social Psychological Perspectives Agneta H. Fischer (continued on page following the Index)

The Mind and Its Stories Narrative Universals and Human Emotion

PATRICK COLM HOGAN University of Connecticut

   Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press The Edinburgh Building, Cambridge  , United Kingdom Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521825276 © Cambridge University Press 2003 This book is in copyright. Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published in print format 2003 - isbn-13 978-0-511-06244-5 eBook (NetLibrary) - isbn-10 0-511-06244-3 eBook (NetLibrary) - isbn-13 978-0-521-82527-6 hardback -  hardback isbn-10 0-521-82527-X

Cambridge University Press has no responsibility for the persistence or accuracy of s for external or third-party internet websites referred to in this book, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

For Lalita

Contents

Acknowledgments

1 2 3 4 5 6

page xi

Introduction: Studying Narrative, Studying Emotion Literary Universals Emotion and Suggestion: Lexical Processes in Literary Experience Four Hypotheses on Emotion and Narrative Writing Beyond the Ending: A Problem of Narrative, Empathy, and Ethics Extending the Theory: Emotion Prototypes, Narrative Junctures, and Lyric Poetry Testing, Revision, and the Program of Research in Narrative Universals: Ainu Epic and the Plot of Sacrifice The Structure of Stories: Some General Principles of Plot Afterword: From the Emotional Nature of Narrative to the Narrative Nature of Emotion

Works Cited Index

1 17 45 76 122 152

172 202 239 265 285

Acknowledgments

An earlier version of Chapter 1 and a small part of the introduction were published as “Literary Universals” in Poetics Today: International Journal for Theory and Analysis of Literature and Communicac 1997 The Porter Institute for Poetics tion 18.2 (1997): 223–49. and Semiotics, Tel Aviv University. All rights reserved. An earlier version of Chapter 2, including one paragraph now in the introduction, was published as “Toward a Cognitive Science of Poetics: Anandavardhana, Abhinavagupta, and the Theory of Literature” in College Literature 23.1 (1996): 164–78. I am grateful to the editor, Kostas Myrsiades, for permission to reprint. An earlier version of Chapter 4 appeared as “The Epilogue of Suffering” in SubStance 94/95, Vol. 30, c 2001. Reprinted by permission of the University of Nos. 1 & 2. Wisconsin Press. These chapters, and my general thinking about literature and cognition, have benefited greatly from the comments of editors and referees at these journals. An earlier version of Chapter 3 was presented at the main panel of the forum on cognitive science at the 1998 Modern Language Association (MLA) convention. An earlier version of Chapter 4 was presented at the symposium on cognition, evolution, and the arts at the University of California at Santa Barbara in 1999. An earlier version of Chapter 5 was presented at the 1999 MLA convention. In each case, members of the audience and fellow panelists stimulated and challenged my thought on these issues.

xii

Acknowledgments

Keith Oatley and Norman Holland read the entire manuscript with care and gave insightful comments and helpful suggestions. The same may be said about the five anonymous referees for Cambridge University Press. Finally, I am grateful to Philip Laughlin, psychology editor at Cambridge University Press, for his masterful handling of this strange, cross-disciplinary manuscript, and (again) Keith Oatley who, as editor for the series “Studies in Emotion and Social Interaction,” made this book possible.

The Mind and Its Stories Narrative Universals and Human Emotion

Introduction Studying Narrative, Studying Emotion

When empirical researchers in the social sciences consider the nature of emotions and emotion concepts, they most often conduct anthropological interviews, send out surveys, analyze linguistic idioms, test stimulus response times, and so on. They may move toward medical and biological study as well, giving injections to test subjects, engaging in neuroimaging, and the like in order to gather as much relevant data as possible. But, with only a few exceptions, they almost entirely ignore a vast body of existing data that bears directly on feelings and ideas about feelings – literature, especially literary narrative.1 Stories in every culture both depict and inspire emotion. Indeed, the fact that some stories are highly esteemed in any given culture suggests that those stories are particularly effective at both tasks – representing the causes and effects of emotion as understood or imagined in that society and giving rise to related emotions in readers. Of course, one cannot assume that depictions of emotion accurately represent those emotions. This is the common, and quite reasonable, objection to treating literature as empirical data. However, we have very good reason to assume that widely admired depictions of emotions tell us something important about the way people in a 1

As just noted, there are exceptions here, especially among researchers influenced by psychoanalytic work, for psychoanalytic theory has drawn on literature since the time of Freud. A recent example is Labouvie-Vief, who combines developmental and empirical psychology with a study of myth from several Mediterranean cultures in order to discuss aging.

The Mind and Its Stories

given society think about emotions. In other words, we have a body of commonly enjoyed, elaborate, narrative portrayals of emotion scenarios. At the very least, these would seem to tell us far more about common emotion ideas than some verbal definition of an emotion term. Moreover, emotional reactions to literary works – the sorrow, anger, mirth felt and expressed by readers – clearly tell us something about what moves people in a particular culture, what touches them emotionally. Indeed, literary response is as close as we can usually get to a wide range of genuine and spontaneous human emotions that are most often concealed in private interactions. In these ways, the celebrated stories of any given society form an almost ideal body of data for research in emotion and emotion concepts. The central contention of this book is that anyone who pays attention to this body of literary data by examining it cross-culturally, cannot help but be struck by the uniformity of narrative structures and of the emotions and emotion ideas that are inseparable from those structures. More exactly, there are extensive and detailed narrative universals. These universals are the direct result of extensive and detailed universals in ideas about emotions that are themselves closely related to universals of emotion per se.

literature as a human act One reason literature has played such a limited role in cognitive study is that science seeks generalities while literature seems to be tied to narrow particularity. In connection with this, even humanists have been resistant to the idea that there are universal patterns in literature. The sharp contrast between literature and the stuff of empirical research seems to be one of the few things that most humanists and scientists agree on. Both literary critics and readers from other disciplines tend to think of literature in terms of nations and periods, genres, schools, and movements. Indeed, the tendency is much more pronounced among professional students of literature. Literary historians and interpreters categorize works of literature by groups, opposing the groups to one another and scrutinizing these groups for differences. What distinguishes Romanticism from Neo-Classicism?, Post-Modernism from Modernism?, European drama from Sanskrit theater?, western lyric from Chinese tz’u? These are the sorts of

Studying Narrative, Studying Emotion

questions asked in comparative literary study. Asking these questions leads one to find answers, and thus to find differences. When finding specific differences, one tends to exaggerate group difference. In other words, when one examines what separates this group from that group or this body of literature from that body of literature, distinguishing features become salient, while commonalities fade into the background. The result is a disproportionate sense of discrepancy and opposition. But in fact there are far more numerous, deeper, more pervasive commonalities than there are differences. As Donald Davidson has argued, even to understand and think about difference, we need to presuppose a vast range of similarities. In Davidson’s words, “disagreement and agreement alike are intelligible only against a background of massive agreement” (137). Put differently, literature – or, more properly, verbal art – is not produced by nations, periods, and so on. It is produced by people. And these people are incomparably more alike than not. They share ideas, perceptions, desires, aspirations, and – what is most important for our purposes – emotions. Verbal art certainly has national, historical, and other inflections. The study of such particularity is tremendously important. However, literature is, first of all and most significantly, human. It is an activity engaged in by all people at all times. As Paul Kiparsky put it, “literature is neither recent nor a historical invention. In fact no human community lacks a literature”; no group is “so wretched that it does not express its memories and desires in stories and poems” (“On Theory” 195–6). More recently, Mark Turner has argued that, “literary criticism has given us a concept of literature as the product of circ*mstances . . . not as a product of the capacities of the human mind. We do not ask, what is the human mind that it can create and understand a text? What is a text that it can be created and understood by the human mind? These questions are not at the center or even the periphery of our critical inquiries” (Reading Minds 16). The professional division of literature by nationality, ethnicity, and so on, tends to occlude this fundamental, human condition of verbal art. The following chapters address literature, then, not as the expression of an ethnic Weltanschauung, nor as evidence of an historical episteme, but rather as a human activity – something people do, and always have done, in all parts of the world, and at all times.

The Mind and Its Stories

At one level, this is, then, a study of literary universals. It is also a cognitive study.2 Indeed, it is my contention that literary universals are to a great extent the direct outcome of specifiable cognitive structures and processes applied in particular domains and with particular purposes. In this way, the study of literary universals is largely a subfield of cognitive research.3 Moreover, it is a crucial subfield for cognitive science. Cognitive science can hardly claim to explain the human mind if it fails to deal with such a ubiquitous and significant aspect of human mental activity as literature. In this way, cognitive science is not only important to the study of literary universals. The study of literary universals is equally important for cognitive science. Indeed, a small, but significant – and expanding – group of cognitive scientists has come to recognize the necessity of incorporating literary study into their domain as work by Steven Pinker, E. O. Wilson, John Tooby and Leda Cosmides, Howard Gardner, Keith Oatley, Jeffrey Saver (see Young and Saver), and others attests. But this is not a book on literary universals in general – which is, in any case, too large a topic for a single work. In Chapter 1, I do introduce general principles for the cognitive study of literary universals. However, the bulk of the volume focuses on the relation between two crucial elements of literature and the human mind – narrative and emotion. Despite the recent cognitive interest in literature, this is an area that has hardly been explored, leaving aside the work of one or two researchers, most importantly Keith Oatley – and even Oatley has not studied this nexus cross-culturally, in an attempt to isolate universal structures. Again, the general absence of attention to this is 2

I am, of course, not alone in linking literary study with cognitive science. The last decade has seen the growth of a significant movement in literary study based on cognitive science. Work by Turner, as well as Norman Holland (Brain), Ellen Spolsky, Paul Hernadi, Jerry Hobbs, Mary Crane, Alan Richardson, and a number of other writers, has provided a valuable alternative to recently dominant approaches to literature. This book is, to a certain extent, part of that movement. At the same time, however, there are some differences between my views and mainstream cognitivism, as I have discussed in On Interpretation. For the most part, these differences do not bear on the topics discussed in the following pages. Thus I shall leave them aside, except for a brief discussion in the Afterword. One important qualification here is that some literary universals do not seem to be a matter of psychology per se, but rather of social conditions, either changeable or permanent. We shall discuss some examples in Chapter 1 when treating implicational universals.

Studying Narrative, Studying Emotion

unfortunate, for it is an area of seemingly obvious value for cognitive science. Narrative is, of course, central to verbal art. Indeed, it is, as I shall argue, even more definitive and widespread than is commonly recognized. For example, it lies concealed in such apparently nonnarrative works as lyric poems. What is crucial for our purposes is that narrative is intimately bound up with emotion. Literary stories, especially the stories we most admire and appreciate, are structured and animated by emotions. Any coherent sequence of events might constitute a story. But the stories that engage us, the stories we celebrate and repeat – “paradigm” stories – are precisely stories that move us, most often by portraying emotions or emotionally consequential events. Conversely, the emotive impact of verbal art cannot be discussed separately from its narrative structure. Indeed, even real life emotion is bound up with narrative. As a number of writers have pointed out, our affective response to a situation, real or fictional, is not a response to an isolated moment, but to the entire sequence of events in which that moment is located, whether explicitly or implicitly. Consider someone’s death. This is narratively embedded, first of all, in the simple sense that we infer the person was alive and some causal sequence led to his/her death. But it is narrative also and more significantly in the sense that we respond to the death in terms of the narrative details through which we understand the person’s life. Suppose we learn that someone died in an automobile accident. We are likely to respond one way if we learn that the person was in the final stages of a terminal illness with only a few pain-filled weeks to live. We are likely to respond differently if we learn that the person was driving to his/her wedding. Skeptics might reply by arguing that narrative in these cases is simply a matter of causal inference and evaluative judgment, and thus is not narrative in any interesting or substantive sense. It is certainly true that what I am describing is in part ordinary causal analysis and evaluation. However, that is not all there is to it. One argument of the following pages is that our ideas about, evaluations of, and most crucially our emotive responses to all sorts of things are guided and organized by a limited number of standard narrative structures. Human thought, action, and feeling are not simply a matter of rational inference. They are also a matter of emplotment in a narrow, specifiable sense.

The Mind and Its Stories

Before going on, it is worth pausing for a moment over the notion of “paradigm” stories. Paradigm literary works are works that are widely shared by writers and readers within a tradition and that serve to establish evaluative standards and structural principles within a tradition. In the following pages, I shall refer repeatedly to prototypical literary works. These are not the same as paradigm literary works, though the latter are most often instances of the former. Specifically, prototypical works are works that share all our standard criteria for verbal art. They share all the properties we consider “normal” for literature. Thus, romantic novels and epic poems – including many that are unknown or even unpublished – are most often prototypical literary works. Riddles are not. Paradigm works usually share these prototypical properties, but add our collective familiarity and esteem. By “esteem” here, I mean esteem as literature. We may admire a work for many reasons. It may express courage in the face of political oppression. It may teach moral lessons that we find valuable. It may celebrate our national heritage. But we may admire a work for any of these reasons and still consider it a poor work of literature. In the following pages, I am concerned with works that are widely admired as good works of literature. Put differently, we tell and write stories every day. Some discussions of narrative are concerned with all these stories. Accounts of that sort are valuable. But they are different from an account that is concerned with prototypes and paradigms. The following analyses do not treat ephemeral stories (for example, what I tell my wife about how I had to go to three shops to get a particular spice). Ephemeral stories may be very engaging at the moment, but they are engaging for idiosyncratic and contingent reasons. What is important here are stories that have sustained interest within their respective traditions. A story that has sustained interest is unlikely to have its appeal for contingent reasons, due to the particular relationship of the speaker and addressee, or due to some unusual circ*mstance. As such, a story that has sustained interest is more likely to tell us something about the human mind. The following analyses, then, aim to begin the process of describing and explaining the remarkably detailed, cross-culturally universal, and interwoven patterns of our emotions, our ideas about emotions,

Studying Narrative, Studying Emotion

and our most enduring stories. One might refer to this project as an anthropology of world literature, in which it turns out that emotions are central – indeed, definitive, and formative. I undertake this task in relation to an encompassing research program in cognitive science.4

universalism and cultural study As we have noted, a handful of writers in cognitive science have recently become interested in literature. In some cases, this interest has extended to literary universals. Indeed, there has been a surprising increase in attention to the topic over the past few years, for the most part among cognitively oriented literary and film theorists. Except for the pioneering work of Roman Jakobson and Paul Kiparsky, for a long while there was little serious discussion of literary universals – hardly a mention, in fact. However, in the context of developing research programs in cognitive science, some scholars and theorists have begun taking the idea seriously. In addition to my own earlier efforts (see “Literary,” “Possibility,” “Beauty,” “Shakespeare,” and 286–95

It should go without saying that this analysis does not treat every aspect of narrative, not even every cognitive aspect. Thus it does not in any way preclude other cognitive approaches. Perhaps the most obvious omission is what narratologists call “discourse,” the mode of presentation of a “story.” The “story” is the events as they happened according to a particular narrative. The discourse is the way in which these events are presented. For example, in a murder narrative, the story begins with the murderer plotting the murder. It moves to commission of the murder, then the discovery of the crime, then the investigation. But murder narratives are not typically told this way. They usually begin with the discovery of the crime, then the investigation leads us to learn about the preceding events. Thus the discourse presents the events of the story out of chronological sequence. Discourse is clearly central to the emotional impact of a work. I have not discussed it simply because it is another topic, and a huge one. Readers interested in this topic should consult Brewer and Lichtenstein for empirical research and Tan for an extended and influential development in relation to film. More generally, there are many very useful ways in which narrative may be studied cognitively – and, indeed, has been studied cognitively. A particularly valuable cognitive treatment of narrative is David Bordwell’s Narration in the Fiction Film that addresses the film viewer’s cognitive construction of the story out of the discourse. The most influential cognitive examination of narrative in literary study is probably Mark Turner’s The Literary Mind that focuses on the mini-narratives of everyday life and their relation to conceptual blending. These, too, do not at all exhaust the possibilities.

The Mind and Its Stories

of Philosophical), a number of other writers have taken up the topic. These include well-established theorists, such as David Bordwell (“Convention”) and, in a very different way, Wendy Doniger, as well as younger critics, such as Alan Richardson and Joseph Carroll, and independent or extraacademic writers such as Ellen Dissanayake. In addition, cognitive scientists such as Steven Pinker, though they do not directly address the issue, clearly presuppose literary universals in their work on cognition and literature. Indeed, recently, the University of Palermo sponsored a website devoted to the topic of literary universals (http://litup.unipa.it). The aim of this website – the Literary Universals Project – is to bring together researchers from different fields in order to advance a research program in the area. With the continuing development of cognitive science, the study of literary universals is likely to expand in both breadth and depth. Still, literary universals remain a minority interest. Mainstream literary critics and theorists pay little attention to the topic. As we have already noted, in professional literary study, the focus of both theory and practice tends to be on difference, cultural and historical specificity, and so on. What Carl Plantinga said of film theorists applies equally to literary theorists: They tend to seek “explicit ways to link” literary phenomena “to particular historical conditions and to ideology” (450). In keeping with this, a self-evaluation by the American Comparative Literature Association worried that comparative literature “may well be left behind on the dustpile of academic history” if it does not incorporate the current trends variously referred to as “culture studies,” “cultural critique,” and “cultural theory.” Indeed, the authors of this study insisted that all work in comparative literature “should take account of the ideological, cultural, and institutional contexts in which . . . meanings are produced,” which amounts to an insistence that all comparatist study be focused on historical and cultural particularities (Bernheimer et al. 5,6). Again, this sort of work is undeniably important. Indeed, my own work (for example, Colonialism and Cultural Identity and The Culture of Conformism) has been, to a great extent, located within the field of culture study. But to say that such particularist study is valuable is not to say it is all that is valuable. When universalism is mentioned in humanistic writing, it is most often denounced as a tool of oppression. For example, in their

Studying Narrative, Studying Emotion

influential introduction to postcolonial literary study, The Empire Writes Back, Ashcroft, Griffiths, and Tiffin maintain that the notion of universality is “a hegemonic European critical tool” (149). There are exceptions, certainly, and not only among writers in cognitive ˜ ı science. For example, the important Kenyan Marxist novelist, Ngug˜ wa Thiong’o, has proclaimed himself “an unrepentant universalist” (xvii). However, there has been a general sense in literary study that attention to or advocacy of universals is somehow politically suspect. There is a fairly straightforward case against such political claims. After all, no racist ever justified the enslavement of Africans or colonial rule in India on the basis of a claim that whites and nonwhites share universal human properties or that their cultures share universal principles. On the other hand, in saying this, I do not want to fall into the opposite error of claiming that everything that goes by the name of “universalism” is politically good. Things are never that simple. Indeed, one does not have to look far to see how universalist claims have been used to support oppression. Typically, humanist criticisms of universalism refer back to those universalist claims that derive from and serve to further colonial, patriarchal, or other ideologies supporting unjust domination. However, as Kwame Appiah has noted, what anticolonial opponents of universals “are objecting to” in these cases “is the posture that conceals [the] privileging of one national (or racial) tradition against others in false talk of the Human Condition” (58). In other words, they are objecting to false and duplicitous claims of universalism, assertions of universality that are untrue and are, in addition, offered in bad faith. Appiah continues, “antiuniversalists . . . use the term universalism as if it meant pseudouniversalism, and the fact is that their complaint is not with universalism at all. What they truly object to – and who would not? – is Eurocentric hegemony posing as universalism” (58; see also Lalita Pandit 207–8). It is important to stress that this conclusion in no way detracts from the standard forms of particularist literary study. It responds, not to their positive worth, but to their exclusivity. More exactly, proponents of cultural and historical study sometimes seem to assume that the study of universals is opposed to or contradictory of cultural study. But to argue for the study of universals is not at all to argue against the

The Mind and Its Stories

study of culture and history. All reasonable students of literature – including those engaged in a universalist project – recognize that particularist research and interpretation are extremely valuable. Indeed, the study of universals and the study of cultural and historical particularity are mutually necessary. Like laws of nature, cultural universals are instantiated variously, particularized in specific cir˜ ı 26; see also King 33, 127, on the culturally c*mstances (cf. Ngug˜ “rooted” universalism of Derek Walcott and Wole Soyinka). Thus, to isolate universal patterns, we often require a good deal of cultural and historical knowledge. At the same time, in order to gain any understanding of cultural particularity, we necessarily presuppose a background of commonality (as, once again, Donald Davidson has argued forcefully [183–98]; see also Brown 151–2). In short, the study of universality and the study of cultural particularity are not contradictory, but complementary.5

universality and narrative: research programs, research methods Needless to say, these general comments do not establish that there actually are literary universals. They do not even indicate just how one might go about isolating universals. The first chapter takes up the nature of and criteria for universals. Specifically, Chapter 1 draws on work in linguistics – the field that has made the greatest advances in the study of universals – in order to explain what constitutes a literary universal and what counts as evidence for the existence of a universal. The second chapter considers the issue of literary emotion, drawing in particular on Indic literary theory and cognitive research to present an account of why literature moves us. Chapters 1 and 2 are, in a sense, preliminary to my main project, introducing basic principles about literary universals and literary emotion. Chapter 3 presents and defends my claims about narrative universals and their relation to emotion concepts. In that chapter, 5

Other writers have implicitly presented universalism and particularism as at least compatible, if not complementary. Good examples would include Zhang and ¨ Kovecses.

Studying Narrative, Studying Emotion

I isolate two universal narrative structures, heroic and romantic tragicomedy. (A number of Shakespeare’s histories and tragedies, such as King Lear – with its treatment of political usurpation, external invasion, and defeat of both the invaders and the usurpers – provide instances of the former structure; many of Shakespeare’s romantic comedies – with their conflicts between lovers and parents or society, separation and exile of lovers, and so on – are good examples of the latter.) I argue that these structures are not only to be found across unrelated traditions; they are, in fact, the dominant structures across traditions. In order to avoid misunderstanding, I should emphasize at the outset that the universality and prominence of these forms does not mean that they are the only forms of narrative. One can tell stories about anything. There are narratives about many different topics and with many different structures. However, other narrative structures simply do not have anything like the central, cross-cultural importance of these two. In addition, I argue that the centrality of these structures is due to another universal – a universal prototype for happiness, or, rather, two contextually dependent universal prototypes for happiness. My contention, then, is that our most prominent stories are generated from the prototypical structure of our emotion concepts. One of the main concerns in the following pages is with establishing the study of literary universals as a research program. The influential philosopher of science, Imre Lakatos, has argued that science does not proceed either by verification (as one might think intuitively) nor by falsification (as Karl Popper argued). Rather, theories encounter recalcitrant data all the time. But recalcitrant data do not simply falsify the theory. Theorists reformulate the theory in order to account for the data. This is what it means for a theory to be part of a research program. Theorists do not simply seek out confirmatory instances, cases that match their theory. Rather, they look for data that might contradict the theory. However, when they find such data, they do not simply toss the theory out the window. Rather, they try to reformulate the theory. Lakatos argues that theorists may reformulate the theory in a way that is ad hoc or in a way that is not ad hoc. An ad hoc reformulation merely isolates the recalcitrant data. It does not expand the explanatory capacity of the theory. A non-ad hoc reformulation makes predictions that go beyond the recalcitrant data. If these

The Mind and Its Stories

predictions turn out to be correct, then we have a robust research program. In other words, the recalcitrant data have not impeded, but advanced the research program. More generally, a successful research program is one that continually expands its explanatory scope and precision. Often it will do this through encountering disconfirmatory data and dealing with those data. In Chapters 4 through 6, I try to follow these Lakatosian principles. In Chapter 4, I raise some troubling issues about heroic tragicomedies. There are cross-cultural patterns that do not appear to fit the account of narrative and emotion presented in Chapter 3. I argue that the seemingly anomalous pattern is in fact predictable, given the principles of emotion put forth in Chapter 2. Moreover, I argue that this further development of the preceding hypotheses has very fruitful consequences for understanding the nature of ethical concerns in literary narrative. Thus, it ultimately expands the explanatory scope of the theory. Chapter 5 does not treat recalcitrant data. However, it does take up a set of data that seemed to lie outside the theory: lyric poetry. In this chapter, I seek to expand the scope of the theory by arguing that lyric poems most often imply narratives. In technical terms, they treat junctural moments in implied narratives. Moreover, the implied narratives are prominently (though, again, not invariably) those isolated in the third chapter – heroic and romantic tragi-comedy. Methodologically, my approach to this point had been fairly straightforward. The isolation of the prototypes for emotion concepts drew on empirical research performed by experimental and social psychologists over the past several decades. The data supporting claims about universals was gathered more slowly. For years, I had been reading important canonical works in all the major written traditions – European, Middle Eastern, South Asian, Chinese, and Japanese. Though concentrating on premodern works, I did include works of modern literature and film as well. More recently, I made a concerted effort to read highly regarded oral narratives from a broad range of regions – sub-Saharan Africa, South America, North America, Australia, and so on. I also did my best to read histories of the written traditions and scholarly outlines of structures and patterns to be found in these different traditions. In connection with this, I worked to familiarize myself with the major non-Western traditions

Studying Narrative, Studying Emotion

of literary theory as well.6 The isolation of heroic and romantic tragicomedy grew out of this reading.7 Experimental psychologists are likely to want laboratory tests and are likely to feel that one does not have a real research program without them. Laboratory research is undoubtedly of great potential importance here. However, my claims concern literature. Thus, the research program bearing on those claims will most directly address literature. In this respect, a research program in literary universals is 6

There are important traditions of literary theory associated with all the major written traditions. Most often, these are ignored by Euro-American writers. Even when mentioned, they are typically seen as outdated or as applicable solely to their own literary traditions. For example, writers who mention Sanskrit literary theory at all tend to see it as bearing only on Indian literature. To my mind, the value of nonEuropean literary theories is precisely in the degree to which they help point us in the direction of universal literary principles and structures. Insofar as we are trained in the theories of one literary tradition, we are likely to be more conscious of elements that conform to those theories. In other words, insofar as a theory organizes and guides our reading of and response to literary works, it will render salient certain aspects of those works, while partially occluding others. Insofar as foreign literary theories differ from those of our own familiar traditions, they will help bring to our attention different features of literary works in all traditions, including our own. Put simply, Aristotle will lead us to notice certain things about European, Indic, Chinese, and other literary works. A great precolonial South Asian theorist, such as Abhinavagupta, will lead us to notice other things. Moreover, as it turns out, some of these non-European theories are remarkably congruent with recent work in cognitive science. For this reason, I make particular use of such theories, primarily those from South Asia, in the following pages. In undertaking this project, I am indebted to my former teacher, Northrop Frye. However, my debt to Frye is perhaps not as obvious as it might seem. Specifically, some readers may be inclined to see my account of romantic tragi-comedy as Fryean. However, at the level where they overlap, neither account is greatly original. To a considerable extent, Frye took well-known facts about New Comedy and its progeny and integrated these into a larger, typological framework. I have drawn on the same well-known facts – along with less widely known, but no less well-established facts about literary works in other traditions – to make claims about literary universals and to integrate these into a quite different, explanatory framework. Leaving aside a few details, my greatest debt to Frye, then, is not in his articulation of the romantic plot. The influence of Frye’s work on this study is, rather, much broader. It is primarily a matter of adopting an inductive approach aimed at isolating recurrent literary structures through empirical study of actual literary works. On the other hand, it is probably true that, without reading Frye, I would not have been as sensitive to some of the specific structures discussed here. In connection with this, I have also benefited from the work of writers such as Hayden White who have extended Frye’s ideas in valuable ways, and from commentators on and critics of Frye’s theories, such as Paul Hernadi (see Beyond Genre 131–51) and Tzvetan Todorov (see Chapter 1 of The Fantastic).

The Mind and Its Stories

precisely parallel to a research program in linguistic universals. Laboratory experiments are important to linguistic study. However, the main research programs in linguistic universals have focused on the investigation of unrelated languages. I hope that experimental psychologists will take up the hypotheses articulated in the following pages and integrate them with laboratory study, leading to revision and improvement of the theory. For now, however, I would like to establish the possibilities for systematic study of literary universals in the way linguists established the possibilities for systematic study of linguistic universals many years ago. On the other hand, the experimental model does highlight the fact that my reading in world literature was not designed to test hypotheses. In Chapter 6, I set out to change this. I set out to study a body of literature precisely in order to evaluate my hypotheses. In other words, I set out to encounter potentially recalcitrant data in order to advance a research program. The first step was determining what would provide a good test case for the claims I had been making. A good instance would be a collection of narratives that are well preserved, highly esteemed within their culture, and isolated from traditions that I had already studied. An ideal instance would help compensate for two deficiencies in the previous data. First, the work would be oral, rather than written. Though I tried to read systematically in oral traditions, my coverage there was certainly much less adequate than in the case of written traditions. Second, and even more important, the work would be composed by women – not women working in a largely male tradition (such instances are readily available), but women working in a tradition that is not dominated by men. I came upon just such a body of work in the Ainu epics. They certainly included heroic and romantic tragi-comedies. However, there were some interesting variations. These variations did not quite fit the claims from Chapter 3. This led to a slight reformulation of the initial hypotheses. However, it was easy to see that the same variations could be found in other traditions as well. I had simply missed them. This was a small case of the sort of program discussed by Lakatos. But that was not all. I began to notice that a number of the Ainu epics treated sacrifice. Strictly speaking, this did not contradict my previous claims. Again, there are many possibilities for narrative.

Studying Narrative, Studying Emotion

Moreover, the heroic and romantic plots were clearly prominent in the Ainu material. However, reading the Ainu stories of sacrifice, I realized that the same narrative structure recurs across all those other traditions. Returning to those traditions, I found considerable material of the same sort (for example, in the Christian story of sin, exile, and redemption – the story of Adam, Eve, and Jesus). This led to the formulation of a third prominent narrative structure, sacrificial tragi-comedy, and to the related isolation of a third contextually dependent prototype for happiness. In this way, the explanatory scope of the initial theory came to be expanded by revision in response to new data – for these literary works provide data, however complex, in just the way natural languages do. Faced with these three genres, one might reasonably wonder what the connections among them might be. In the seventh chapter, having already treated these genres individually, I take up their interrelations. The three are remarkably similar in their organization and development. Chapter 7 sets out to isolate and explain the detailed structure that they share. The afterword returns us to the issue of emotion, now asking how the preceding account of narrative structures and emotion prototypes might relate to a broader theory of emotion per se. Specifically, it considers how the biological givens of emotion come to be bound up with social narratives – not only in fictional stories, but in our real emotional lives as well. In undertaking this study, my main hope is that its various descriptive claims and explanatory hypotheses will contribute to our understanding not only of literature, but of the human mind, specifically human emotion and the human conception of emotion (for example, the human imagination of happiness), with all that this entails. Again, I envision the following analyses as part of an ongoing research program – not only in literature, but in cognitive science and the psychology of emotion. They are not an application of cognitive principles to literary works, but a development of cognitive principles through the study of literature. At the same time, a work of this sort is not merely scientific. Understanding the breadth and depth of cultural universals – of literary and emotive commonality – is not politically inconsequential. As we

The Mind and Its Stories

have already noted, racial and cultural hierarchies are routinely and necessarily justified by an appeal to putative racial and cultural differences, even if these appeals are sometimes hidden behind universalist rhetoric (much as unequal treatment and double standards are often concealed behind rhetoric of equality and fairness). A research program that succeeds in uncovering genuinely universal principles of human feeling, expression, and interaction, principles that are not relative to race or culture, runs contrary to racism and ethnocentrism. Of course, we should not decide in favor of universalist hypotheses simply because they appear to be politically beneficial. False universals can be deeply pernicious. But an excessive readiness to accept universality seems an unlikely danger in the current intellectual climate, at least that of humanistic study, where a laudable emphasis on the value of examining cultural particularity is all too often viewed as incompatible with the study of universals. In short, I hope that readers of this book will come to recognize that universalism versus particularism is a false dichotomy. More generally, I hope that, at the end of this book, readers will be more ˜ ı, Marx (see Economic 114), Frantz inclined to follow Chomsky, Ngug˜ Fanon (10), Samir Amin (see especially the preface and final chapter), Kwame Appiah (58, 152), Aijaz Ahmad (316 and elsewhere), Edward Said (6 and elsewhere), and others, in recognizing both the intellectual and political value of studying universals – in this particular case, recognizing that our aspirations and emotions are fundamentally the same, no matter where we were born or what we look like, and that the stories we admire and preserve, stories about these aspirations and emotions, are most often mere variations on a handful of shared patterns.

1 Literary Universals

the structure of a theory of universals The first important point about literary universals is that they are not necessarily properties of all literary works. Indeed, such properties are rare, and often trivial (that is, a mere residue of our definition of a literary work). Rather, literary universals are properties and relations found across a range of literary traditions. What, then, constitutes “a range of literary traditions”? In linguistics, one counts a shared property as evidence of a universal only if the languages in question are genetically and areally distinct, which is to say, only if they have distinct origins and have not influenced one another with respect to this particular property. The basic idea is straightforward. If a shared property is the result of a common source – either because the languages in question have a common ancestor or because the property has been borrowed by one language from the other – then that property does not provide evidence of a universal. French, Italian, and Spanish do not provide three separate instances of a shared property, indicating that it occurs spontaneously in a range of languages. In all likelihood, they provide only a single instance of that property, because in all three cases that property probably derives from a common source. The same may be true of a property shared by, say, Spanish and Basque. Though Spanish and Basque are genetically distinct languages, there has been enough interaction between speakers of these languages that the property 17

The Mind and Its Stories

in question may well be the result of influence. In this case too, the shared property would not count as evidence of a literary universal. With some slight qualifications, the general principle of genetic and areal distinctness serves as an appropriate criterion for the isolation of literary universals as well.1

The qualifications bear on areal distinctness. It is rather difficult for one language to influence another. Contact between the language communities must usually be prolonged and intense. Moreover, linguistic influence is most often localized. For example, it is common for one language to introduce vocabulary items to another language, but it is rare for one language to introduce syntactic structures into another language. The problem with literature is that it is, in general, much easier to influence literary composition than to influence language. (For a sense of how areal influence operates in literature, see, for example, Edmonson 6–23.) A single story, transported across continents, could have a significant impact on a distinct tradition. Something like this could never happen in language. Even the transportation of a vocabulary item across continents will have only very local impact within the receiving language. But, then, the ease of influence in literature cuts both ways. It presents a problem, but at the same time suggests a partial solution. We need vocabulary items. Our inclination is always to add more words whenever they make distinctions lacking in our current system. For example, every time some group encounters a new species of bird or fruit, they want a name for it. There is no similar imperative in literature. If a single work, transported across continents, does indeed have a significant impact on another tradition – in the absence of political or related pressures – this seems to suggest that there was some sort of prior aesthetic propensity that this new work satisfies. Put differently, there is a common view in linguistic theory that all aspects of universal grammar are available to all speakers, even when those aspects of universal grammar are not instantiated in a given speaker’s language. For example, Kiparsky points out that “because the category of syllable onset is defined in universal grammar, words with identical syllable onsets are recognised as an equivalence class . . . whether or not the grammar of the language happens to contain rules referring to syllable onsets” (“On Theory” 192). In cases of literary influence where the input from the influencing source is minimal, we might infer that something similar is going on. The source literature has influenced the recipient literature because the recipients were already sensitive to the properties of the source work, because those properties were already universal. This is particularly likely in those cases where the relevant properties are complex and incorporated unself-consciously into the second tradition. In treating literary data, then, the ordinary linguistic criteria of genetic and areal distinctness seem inadequate. Specifically, they are not sufficiently fine grained. We can solve this problem if we simply turn to the ideas that underlie these criteria in linguistic study. To count as evidence for literary universals, literary properties should be found in genetically distinct traditions. Areal influence, however, should be further analyzed in relation to two criteria. First, we need to distinguish those cases where areal contact was intense and long-standing – and especially those cases where there was forced cultural hegemony – from cases where the contact was limited, sporadic, and unforced. If the areal influence was a matter of a single text

Literary Universals

But this does not fully determine what counts as “a range of literary traditions.” It gives us a criterion for distinguishing traditions, but it does not tell us what constitutes an appropriate “range.” We have already noted that a literary universal need not apply to all literary works. One might assume that it must apply to all traditions. But this is not the case. Linguists use the term universal to refer to any property or relation that occurs across (genetically and areally unrelated) languages with greater frequency than would be predicted by chance alone (see, for example, Comrie 10–12, 19–22). An absolute universal is merely a special case of this, a property or relation that occurs across all traditions. Universals with a frequency below 100% (but, of course, higher than chance) are referred to as statistical universals. In the following chapters, I will be concerned primarily with absolute universals or near absolute universals. Some confusion can arise from the fact that these absolute universals recur in all traditions, but (as already noted) they need not recur in all works. Rather, their presence in individual works only has to be greater than chance. In other words, there is a statistical element even in absolute universals. For example, I shall argue that romantic tragi-comedy is one of three predominant genres in all traditions. Thus I shall argue that it is an finding its way into a recipient culture, it seems overly stringent to discount sharped properties on the basis of areal influence. In contrast, prolonged or coerced cultural contact of the sort that one encounters in, say, colonial occupation, should lead us to discount shared properties. This is why postcolonial literature cannot have a place in a theory of literary universals. Finally, in addressing area influence, we must consider the degree to which the properties or structures in question are open to self-conscious choice. The simpler and more self-conscious the property in question, the more doubtful it is as an instance of universality. For example, suppose we wish to assert the universality of dramatic enactment, the portrayal of events by actors taking on the roles of literary characters. This is clearly something that is self-conscious. This is the sort of thing one tradition can easily take over from another. Specific patterns in background image pattern, however – or even the idea of using patterned imagery as background in a narrative – is an unlikely candidate for casual transferal. The former is much more akin to vocabulary; the latter is much more akin to syntax. Thus, even limited contact between literary traditions may lead us to question shared theatrical representation as evidence of a literary universal. However, limited contact would be a poor reason to dismiss shared patterning of background imagery. (By “background imagery,” I am referring to patterned imagery – of seasons, day and night, color, whatever – that bears directly on the themes of the work but is never explicitly connected with those themes, through for example simile. Rather, it is simply present as an apparently incidental part of the scenery.)

The Mind and Its Stories

absolute universal. However, I shall not argue that all works are romantic tragi-comedies. At minimum, I have to argue that it occurs among paradigm works in every tradition with a frequency greater than chance. Some readers may find it odd to refer to a universal as absolute if it does not apply to every work. But in fact this is a perfectly ordinary usage. Take a simple example. I suspect that hom*osexuality is an absolute universal. In other words, I suspect that it recurs across every culture. The absoluteness of this universal is unaffected by the fact that most people in every culture are not gay or lesbian.2 More exactly, a theory of literary universals includes a repertoire of techniques available to authors and a range of nontechnical correlations (that is, correlations that are not techniques) derived from broad statistical patterns. Nontechnical correlations comprise universal principles that are not devices we could use to make literature – though they may define a range of or limits on such devices. For example, standard line lengths appear to fall regularly between five and nine words. Clearly, a range of standard line lengths across different traditions is not a technique available to authors. Rather, it is a broad correlation across literatures. On the other hand, this universal correlation does presumably indicate a constraint on the techniques available to poets cross-culturally – or, if not a constraint, at least some sort of default tendency. Techniques include all universal matters of “form” and “content” – including poetic meters, rhetorical devices, and so on – that an author 2

This is clearly just an issue of terminology. A more serious question concerns not the words universal and absolute, but the possibility of calculating chance in the case of both statistical universals and absolute universals based on statistical properties. It is true that often one cannot calculate random probability with precision in literary cases. However, one can calculate it adequately for the determination of universals. Consider plot. There are uncountable topics and structures for stories in ordinary speech. For example, I might tell a story about how my lights went out and I had to go buy a fuse. Because of this, we cannot calculate the precise likelihood of romantic love turning up as the topic of story after story – say, in a third to half of paradigm works in tradition after tradition. But, given the huge number of topics for narrative, that likelihood is clearly near zero. Moreover, the impossibility of precise calculation is not a unique issue here. It is common in areas of psychological complexity. For example, David Rubin notes that he cannot make precise predictions using his theory of cues and constraints for memory in oral composition “because all possible alternatives to the words” used in particular cases “could not be listed” (302). But this does not affect the fact that the actual words used in poems undoubtedly conform to Rubin’s rules at a far greater rate than would occur by chance.

Literary Universals

may draw upon in composing a literary work. Most basic techniques used in English literature appear to be universal. A partial list would include symbolism and imagery, which are present in every tradition of which I am aware; assonance – found not only in Indo-European works, but in Japanese (Bownas lii), Babylonian (Sandars 17), and other verse (on the related phenomenon of rhyme, see Kiparsky 10); alliteration – important in Sanskrit (Miller 12), Japanese (Bownas lii), Ainu (Philippi 29), and so on; verbal parallelism – found in Tikopia (Firth with McLean 41), Igbo (Egudu and Nwoga throughout), Basotho (Kunene 68), Kuna (Sherzer 105), Chinese (Cooper 92 and Chapter XXXV of Liu Hsieh), Babylonian (Sandars 17), Hebrew (Sandars 17), and so on. There are also broader organizational devices that appear to be universal as well. An obvious case is foreshadowing. It is necessarily at least as widespread as the belief in omens, with which it is often connected (see, for example, BrewittTaylor, vol. I, 1–2 or 85). Another good case is plot circularity, that is, beginning and ending a plot in the same place or situation or in closely analogous places or situations, often with the repetition of specific phrases concerning those places or situations (for a striking in˜ ana´sa¯ kuntalam, see my “Beauty” 25–6). stance of this, K¯alid¯asa’s Abhijn¯ A particularly interesting device is structural assimilation. This is the explicit or implicit patterning of one work on the plot of another, often culturally central work. Obvious instances would include the common use of the story of the Fall as a model for Judeo-Christian works. It is also widespread in Muslim traditions, as shown by ¯ a) in his treatFerdowsi’s implicit use of the story of Moses (Mus¯ ment of the infant D¯ar¯ab (set afloat “in a casket on the Euphrates river” [221]) or his use of the story of Abraham (Ibr¯ah¯ım) and Ishmael (Ism¯a’¯ıl) in presenting Rost´am and Sohr´ab. In India, we find a promi˜ ana´sa¯ kuntalam on parts of nent case in K¯alid¯asa’s modeling of Abhijn¯ the R¯am¯ayan.a (for discussion, see my “Beauty,” 30). These techniques are organized into (explicit or implicit) schemas defining literary types and subtypes, such as “sonnet,” in English. Within schemas, techniques may be obligatory or optional. Obligatory techniques are techniques an author must use. For example, if a certain poetic genre requires the use of end rhyme, then end rhyme is obligatory in the schema for that genre. Optional techniques, on the other hand, are techniques that are available to an author, but are

The Mind and Its Stories

not obligatory. For example, the sonnet does not require alliteration, but it permits alliteration; thus, alliteration is an optional technique available to a poet composing a sonnet. We may also distinguish techniques that, while not strictly obligatory, are standard. In relation to alternatives, we could understand such standard techniques as the highest or default cases within a schema. A standard technique is employed unless the author makes an explicit choice to the contrary, or some concurrently operating principle or schema prevents the implementation of the default. Perhaps the best way to conceive of these schemas is by reference to cross-indexed entries in one’s mental lexicon. Some techniques are specified directly in the schema, which of course has its own lexical entry (for example, “sonnet”). Others are made accessible indirectly by reference to distinct lexical entries. These other entries may simply be coordinated (that is, fully distinct, though cross-indexed), but they may also be superordinate (encompassing the entry in question) or subordinate (encompassed by the entry in question). Thus the schema for “sonnet” might be structured in the following way. It would list features specific to the sonnet (number of lines, rhyme scheme), then add some reference to the overarching category “poem” (we could think of it as a “See ‘poem’” instruction). The entry for “poem” would be a superordinate category to “sonnet” and would include a list of techniques standard in poetry and available for use in sonnets (for example, alliteration). The precise nature of each of these techniques could be viewed as defined in coordinate categories, also cross-indexed (for example, with a “See ‘alliteration’” instruction). Organization of obligatory and optional techniques into schemas, crossindexing of techniques, and so on, all appear to be universal as well. The most broadly encompassing schema, and an absolute universal, is the minimally specified schema of verbal art itself. As Kiparsky has pointed out, all societies have verbal art (“On Theory” 195–6). This may seem a mere triviality, but it is not. There is no logical necessity in the existence of verbal art. In our own society, very few people actually produce verbal art. Why, then, should we expect it to appear in every society? As Chomsky has emphasized, one of the first tasks for researchers in the study of universals is to overcome habituation and recognize how surprising universals are. We often “lose sight

Literary Universals

of the need for explanation when phenomena are too familiar and ‘obvious’” (Language and Mind 25). Once we have recognized that our expectation of verbal art is a mere matter of habit, we come to see that, far from being trivial, it is in fact highly surprising that verbal art is produced in small nomadic groups and in vast, highly urbanized nations. Below this, in their most general forms, the three major genres of European literature – poetry, prose fiction, and drama – appear to be instances of larger universal categories as well. Thus it appears to be a universal that all or almost all societies have verse, which is to say a verbal art involving formalized cyclical organization of speech based on fixed, recurring patterns of acoustic properties. Tale telling also appears to be a literary universal. Probably in all societies, people articulate causal sequences of nonbanal events involving human agency (with banality defined relative to culturally specific expectations), and they do so at least in part for aesthetic enjoyment, itself based on identification, the patterned variation of emotional intensity, and so on. Finally, some form of enactment for such tales seems to be universal as well, though more limitedly. In some societies, this may be confined to brief episodes on festival occasions or in rituals. However, other societies – including all the major written traditions – have developed a form of extended drama. Thus, we find elaborated theatrical works in Europe, India, Southeast Asia, China, Japan, and the Middle East. (As the Middle East is often claimed to have had no precolonial theater, it is perhaps worth referring the interested reader to Moreh, Martinovitch, and Chapter 1 of Badawi.) More specific schematic patterns are isolable as well. To take a case we shall discuss in detail in Chapter 3, it appears that every tradition tells tales of conflict in two areas – love and political power. Moreover, these tales involve a wide range of common character types and motifs, fall into similar subgenres, and so on. Consider love stories. A romantic comedy, in its most minimal form, typically involves two lovers who are separated, then reunited after a period of uncertainty. It is already surprising that this structure should be found in drama from Greece, Rome, India, China, and Japan, and in stories from other regions as well. More surprising still is the fact that more particular patterns in this genre are also widely shared. For instance, the separation is typically a result of the lovers’ conflict

The Mind and Its Stories

with social expectations and structures, often manifest in a conflict with parents. It is frequently resolved by some sort of recognition, leading to a reversal, in the standard Aristotelian manner. Moreover, this recognition not only reunites the lovers, it often involves the unexpected reunion of parents and children. It includes such characters as a hero, heroine, hero’s companion, hero’s parasite, heroine’s confidant, and so on. Indeed, the character typologies set out by Sanskrit literary theorists two millennia ago (for example, in Bharatamuni’s N¯a.tya´sa¯ stra) and those drawn from Greek and Roman New Comedy by Northrop Frye are similar to one another, and widely applicable beyond their own traditions, because of this cross-cultural consistency. Perhaps the most cross-culturally widespread version of the love plot is a particular variation on the comic love story. This variation, “romantic tragi-comedy,” in effect includes the tragic love story, where the lovers are separated, typically by death, often with a suggestion of literal or metaphorical reunion after death – as in Arabic ¯ story (for example, and Persian retellings of the Layl¯a and Majnun Niz.a¯ m¯ı’s twelfth-century poem), the R¯am¯ayan.a (100 b.c.e.), the love suicide plays of Chikamatsu (for example, Love Suicides in the Women’s Temple of 1708 [in Keene Major Plays]), and so on.3 Specifically, in romantic tragi-comedy, lovers are almost joined, then separated in a way that suggests death, then reunited in a sort of resurrection. The separation at least threatens not only to keep the lovers apart, but to prevent them from ever hearing of one another again (as in Zeami’s early fifteenth-century Lady Han or The Reed Cutter [both in Keene Twenty Plays]). This sort of separation is already akin to

Some readers might question my use of the term “tragi-comedy” here, complaining that tragi-comedy is necessarily a late development that synthesizes tragedy and comedy. I use “tragi-comedy” to refer to plots that pass through or closely approach an apparently tragic conclusion before resolving happily. Since these are comedies, I could simply have referred to them as such. I use the term “tragi-comedy,” however, to signal the complex structure and to emphasize the close relation of tragedy to comedy. In Chapter 3, I shall argue that tragi-comedy, in this sense, is the fullest and most widespread literary form cross-culturally. Moreover, I shall argue that tragedy is not a component of tragi-comedy, but a derivative of tragi-comedy – in effect, a shortening of tragi-comedy. Indeed, according to the analysis in Chapter 3, tragedy is only possible as a failed comedy, for the nature of narrative development is necessarily oriented toward comedy.

Literary Universals

death, but the link with death is typically more direct. Thus, the sep¯ aration often involves an apparent death – as in Chariton’s early Greek romance, Chaere¯as and Kallirrh˘oe¯; Shakespeare’s The Winter’s Tale; Bh¯asa’s fourth-century Vision of V¯asavadatt¯a (in Woolner and ¯ Sarup); Bhavabhuti’s eighth-century Uttarar¯amacarita; and Manohra (in Brandon, Traditional Asian Plays), the “earliest known drama in Thailand” (Brandon, Cambridge 234). If there is no literal death or near death, the separation is frequently represented in imagery closely as´ sociated with death (as in Sakuntal¯ a’s assumption into the heavens ¨ “soul leaving her body” in Chˆeng’s fourteenth-century or Chien-nu’s play [in Liu Jung-en]). In any case, the link with death is clear, consistent, and important in a wide range of literary traditions. In keeping with this, there is a regular association of the separation with imagery of seasonal demise (for example, winter) and the reunion with imagery of seasonal rebirth (for example, spring). Additionally, more general comic universals – conflict with society, recognition, reversal, reunion of separated parents and children, and so on – carry over as well, giving this schema remarkably detailed cross-cultural consistency. In Europe, this sort of story most obviously makes us think of Shakespeare, but it has been a standard part of European literature for millennia, with prominent instances including, for example, the ¯ “Erotici Graeci” of the early centuries c.e. Outside Europe, beyond the works already mentioned, we could list the first-century Toy Cart and the seventh-century Ratn¯aval¯ı (in Lal) from India; the thirteenthcentury Chang Boils the Sea (in Liu Jung-en) and the roughly contemporary story of Qiu Hu (see Dolby 155–6) from China; the final voyage of Sindb¯ad and the story of Aladdin (‘Al¯a’ al-d¯ın) from the Middle East (in Dawood, Tales, on dating see 8–9; see also Allen 176 and 168 on the provenance of these stories); Kan’ami’s fourteenthcentury Hanakatami (in Waley N¯o Plays) and the eighteenth-century Love Letter from the Licensed Quarter (in Brandon, Kabuki) from Japan – to take just two examples from each region that has produced a major written tradition. It is important to point out that the universality of either a (general) technique or a nontechnical correlation in no way implies the universality of any specification or instantiation of that technique or correlation, nor is putative universality falsified by differences in such specifications or instantiations. For example, the patterning of

The Mind and Its Stories

images seems to be an absolute universal. It is certainly found in the genetically distinct, and in part areally distinct, written traditions of Europe, the Middle East, India, China, and Japan; it is also found in the unrelated oral poetries of the Southern African Basotho (see Chapter 7 of Kunene), the Polynesian Tikopia (see Firth with McLean 36–9), the Yirrkalla of Arnhem Land (see Berndt 73–6), and so on. This does not, in and of itself, imply that implementations of the technique share further universal properties. For instance, it appears to be a universal that love is commonly associated with images of birds. Yet, if so, it is a distinct universal. Even if different cultures used widely different image patterns – some linking love with birds, others with tubers, others with types of fabric – the use of image patterns would still be universal. Note, in connection with this, that each level of a hierarchy of abstraction serves as a partial explanatory generalization of all elements on lower levels. Thus, we may find that the use of bird images for romantic love and the use of seasonal images for human life are far more common than chance, but not absolute. We may then find the more general use of image patterns to be absolute. In this case, the more abstract universal of imagery would partially explain the more specific avian and seasonal images. Other universals – literary and nonliterary – in combination with more specific accidental circ*mstances (such as environmental conditions), would yield a fuller explanation of these lower-level universals. For example, the crosscultural connection between romantic love and imagery of birds may derive in part from a universal, metaphorical, but not specifically literary correlation between “positive” emotions and the direction up (cf. Chapter 5 of Lakoff and Johnson). The more abstract universals may also be explained further. For example, the patterning of images may in turn be an instance of a still broader universal according to which the limited patterning of normal discourse is generalized to all levels of structure in literary art (as we shall discuss below). At the highest level, the complex of distinctively literary universals should indicate what is at the origin of the development of literature, what defines the human urge to make and experience verbal art. Of course, these overarching literary universals are not the end of the story either, for they too should be open to further explanation in terms of even more encompassing

Literary Universals

universals, such as those of psychology, sociology, history, and so on. (We shall return to this topic also in the following section.) Hierarchies of universals are defined not only by a receding series of explanatory abstractions, but also by a series of conditional relations. As a general methodological principle, linguists (like scientists in any other field) seek to redefine universals in such a way as to limit exceptions. Through repeated reformulation, they seek to bring statistical universals closer to absolute universals. Again, this may be done through abstraction, as when the (perhaps) absolute universal of image patterning is derived from the (perhaps) merely statistically universal patterns of bird imagery, season imagery, and so on. (In fact, bird and season imagery are probably absolute or near absolute universals. The point is illustrative.) Beyond this, however, the goal of absolute universality may be pursued through the delineation of specific conditions in which statistically universal techniques or correlations occur. In other words, statistical universals of the form “q” may be revised into implicational universals of the form “If p, then q.” Ideally, this reformulation would yield an absolute universal (that is, there would be no cases of p and -q). At least, it would limit the number of exceptions, bringing the universal closer to a frequency of 100%. Alliteration provides a good example of how universals may be redefined to limit exceptions. It is an obligatory technique in certain forms of poetry in certain societies – more than would be predicted by chance, but with a great many exceptions. One may abstract from this to some broader principle, such as the no doubt absolute universal that all poetry has some obligatory features relating to sound pattern. However, one may equally seek to formulate an implicational universal, determining conditions in which alliteration is or is not obligatory. Kiparsky argues that alliteration “seems to be found as an obligatory formal element only in languages where the stress regularly falls on the same syllable in the word, which then must be the alliterating syllable” (Kiparsky, “The Role” 9). While this does not fully fix conditions under which alliteration is obligatory, it does fully fix conditions under which it is not obligatory (that is, whenever syllable stress varies in the language). Thus, it yields the absolute implicational universal, “If syllable stress varies in a given language then alliteration is not an obligatory feature of poetry in that language.”

The Mind and Its Stories

It also yields a second, more complex absolute implicational universal: “If syllable stress does not vary in a language and if alliteration is an obligatory feature of poetry in that language then the stressed syllable is the alliterating syllable.” Much as unconditional universals may be subsumed into hierarchies of abstraction, implicational universals may be organized into typologies (on this and other aspects of universals in linguistic study, see Comrie and Croft). A typology consists in a set of mutually exclusive categories, each of which coordinates a number of implicational universals, forming them into a pattern that is more informative than any of the implications considered individually. Each type in a typology serves as a partial explanation of any given implication that it subsumes. In addition, the typology as a whole should come close to defining an absolute, disjunctive universal – that is, all, or almost all, literatures should fit under one or another type. The broad distinction between oral and literate composition is a case in point (on this distinction, see Ong). In effect, it sets up a broad implication of the form: “If a culture does not have writing, then its verbal art will be marked by frequent use of epithets, formulaic phrases, specific sorts of repetition, etc.,” or more generally “The degree to which a body of verbal art is marked by epithets, etc., is a function of the degree of literacy of the culture in which that verbal art is produced.” As the second formulation indicates, there is a spectrum of possibilities here with two general tendencies as endpoints, rather than a universal grid of discrete types; however, such a spectrum serves the same organizational and explanatory function. Though some writers have disputed the validity of this loose typological distinction (see, for example, Sherzer and Woodbury 9–10), it has been convincingly applied to a wide range of literatures: Slavic, Greek (see Lord), Sanskrit, Tamil (see Kailasapathy; though see also Hart, Poems), Ainu (see the introduction to Philippi), Xhosa (see Chapter 6 of Opland), and so on.4 4

William Brewer’s “The Story Schema: Universal and Culture-Specific Properties” treats some typological universals of orature. Brewer’s essay is valuable and insightful. At the same time, it illustrates some of the problems that occur in the study of universals, primarily the difficulty of noticing universals, a difficulty stressed by Chomsky, as we have already noted. For example, Brewer explains that “Conventionalized story openings occur throughout the world” (179). This would seem to be

Literary Universals

Indeed, this is a particularly interesting universal for our purposes as it makes reference to historical conditions. Humanists tend to think that the study of history and the study of universals are diametrically opposed. In fact, they are not. Marx’s isolation of historical/economic laws (Marx, Capital 8) is an obvious case in which historical change has been understood in relation to universals. Moreover, historical linguists have been no less inclined to isolate universals than have their colleagues in other areas of linguistics (see, for example, Chapter 10 of Comrie and Chapters 8 and 9 of Croft). In this case, the proposed typology not only involves historical contingency (the presence or absence of writing), it also involves historical development and thus specificity – though all in the context of universal principles. Another instance of this type is Arnold Hauser’s hypothesis that realism in literature is a function of social structure, the more realistic art being produced in urban, market economies, with feudal or tributary economies fostering stylization (see Hauser 49). Related to this, a particularly important distinction, relevant to historical and cultural variation, is that between indexical and nonindexical universals. Indexical universals are universals that cannot be fully particularized by reference to shared social or cognitive properties alone (for example, limits on the capacity of human working memory or the type of economic system in which works are produced). Rather, to be fully particularized, indexical universals must make reference to individually variable, subjective conditions (such as personal memories). For example, it appears to be a psychological universal that one’s conception of oneself is structured into a hierarchy of properties, such that properties important to one’s self-conception (such as sex) are high in the hierarchy and properties less important to that self-conception (such as ring size) are lower in the hierarchy. It seems that readers and auditors identify with a character on the basis of shared high-level properties in their self-conception and that they prefer works involving characters with whom they identify (see Klemenz-Belgardt 367–8 and citations).This (likely) literary a clear case of a universal. However, Brewer places this in the category of “culturespecific properties” (177) because the particular form of those openings varies. It is, of course, important to recognize that the precise conventions differ (that is, all cultures do not use “Once upon a time”). But it is at least as important to recognize that conventionalization of openings is itself a universal.

The Mind and Its Stories

universal is indexical because those high-level properties – and thus the associated literary preferences – vary from person to person (by sex, race, religion, and so on). One common preconception about literary universals is that they entail universal agreement in matters of taste. The existence of indexical universals shows that this preconception is false. Finally, we shall see in subsequent chapters that many difficulties in the study of universals may be overcome if we understand that universals are more likely to be based in prototypes than in necessary and sufficient conditions. Prototypes present us with standard cases. It may seem that prototypes must be less universal than necessary and sufficient conditions as they are more concrete or less abstract. But, in fact, it turns out that standard cases are far more likely to be universal than are limiting definitions. Put differently, cross-cultural variation (and, indeed, individual variation) is most often a matter of variation in marginal cases, limiting instances that bear on the formulation of necessary and sufficient conditions; it is much less frequently a matter of central or prototypical cases. For example, in looking at narrative and emotion, we will be considering prototype cases of narrative and prototype cases of emotion. We will not be trying to formulate necessary and sufficient conditions for narrative or emotion. Again, it is the former, not the latter, that appear most universal. In this way, the more we look to prototypes, the more likely we are to find absolute universals.

explaining literary universals: the nature of a research program One could draw further, consequential distinctions – for example, between universals bearing on aesthetic experience and universals bearing on evaluation outside aesthetic experience (for example, universals concerning canonization and dominant ideology). However, it is beyond the scope of this chapter, or this book, to present a final and complete listing of concepts germane to a theory of universals. Indeed, an empirically based theory – unlike the speculative theories common in the humanities – cannot arise in a fully developed and final form, ready only to be “applied” in explications of individual texts. Rather, an empirically based theory is always and necessarily

Literary Universals

part of an ongoing, broadly collaborative research program – in this case, a program involving the collaboration of a wide range of scholars in different fields, with different areas of literary expertise. For present purposes, the preceding analysis should provide an adequate idea of how a theory of literary universals can be structured initially, and what some parts of that theory might look like. Needless to say, a research program of this sort necessarily seeks not only descriptive adequacy, but explanatory adequacy (see Chomsky, Aspects 24–6). Descriptive adequacy is achieved when one formulates a generative account of the phenomena being studied (for example, language). A generative account presents rules that produce the phenomena. Suppose we are doing grammar. We discover that some languages place the object before the verb and some place it after the verb. We then discover that some languages place adpositions before the noun (making them prepositions), while others place them after the noun (making them postpositions). We go on discovering things of this sort. We then notice further patterns. A certain adposition/noun order is regularly correlated with a certain object/verb order and so on. We conclude that head/complement order is usually constant within a language and that there are just two options for head/complement order: head/complement, complement/head. Each time we isolate a pattern and formulate that pattern as a rule, we have a more descriptively adequate account of the phenomena under consideration. Note that in physics this is usually all we have to do. In physics, once we can generate the patterns from rules, we usually stop. But this is not true in linguistics or in other areas of cognitive study. In linguistics, we need to formulate a set of rules that generate the language, as in physics we formulate a set of rules that generate observable physical phenomena. But, in addition to this, we need to account for just how this set of rules gets into people’s minds.5 How does a child acquire this complex grammar? Linguistics and related fields of psychology must add this second step to a theory. This second step is “explanatory adequacy.” 5

Of course, the rules might, for the most part, get there genetically. For example, in Chomsky’s widely influential “principles and parameters” approach, the principles of grammar are innate and experience serves simply to set parameters (see, for example, Chapters 1 and 5 of Chomsky’s New Horizons). In this account, explanatory adequacy is achieved largely by innatism.

The Mind and Its Stories

In the present study, my aim is to isolate universal patterns of story structure in such a way that we can see many specific stories as instances of these patterns (descriptive adequacy), then to explain these patterns by reference to well-defined cognitive principles (explanatory adequacy). Thus, having treated some broadly structural and descriptive aspects of a theory of literary universals in the preceding section, I should like, in this section, to consider the way in which the explanatory part of a research program in universals might proceed. To do this, I shall examine two specific cases, one concerning a complex of universal formal techniques and a principle that can be abstracted from those techniques, the other concerning a universal statistical correlation and its likely derivation from cognitive structure. As to the first, consider the list of formal devices that are used in a wide range of genetically distinct traditions – assonance, alliteration, parallelism, and so on. Again, these are best thought of as techniques available to writers in creating literary works. A first step in an explanatory research program would be to abstract some sort of principle from this list, a principle that indicates what these items share, what pattern they form, and that does so in such a way as to give this pattern some sort of function in verbal art. We could call such a “function-congruent” abstraction from empirically observable patterns a “secondary principle.” Paul Kiparsky has noted what is at issue in this particular case. He observes broadly that “it appears . . . all literary traditions . . . utilize the same elements of form” (“The Role” 11). Following Roman Jakobson, he goes on to suggest a reason for this: “[L]anguage allows certain ways of organizing sounds, and . . . poetic form must draw on this organization” (20). He concludes that the relations “between grammar and poetry account, at least in part, for the universality of poetic form” (22). The general connection between linguistic sounds and poetic sounds is plausible, and Kiparsky advances the discussion by drawing this link. However, this falls well short of an explanation, even such a minimal explanation as is given in a secondary principle, for it does not say anything about the specific use of the linguistic phenomena in literature. Part of the point of the literary devices is that they are different from the ordinary linguistic phenomena to which Kiparsky reasonably relates them. For example, onsets (the

Literary Universals

beginnings of syllables) are not used in the same way in ordinary language and in literature. The two uses are continuous, but not identical. We do not seek alliteration (or rhyme or assonance) in ordinary talk. We do in poetry. Relating the two may be part of an explanatory account, but necessarily only a part. What, then, is going on in literature? Literary theorists from different traditions regularly stress the unusual degree of structure and relevance in literature (see, for example, the valuable discussion in Bateson, especially 14). Presumably at least some of our enjoyment of verbal art is related to this. All speech is patterned. Whenever we speak, we try to make a coherent statement, present a coherent narrative, and so on. We choose our words so that they have the right sort of connotations. We try to avoid harsh or comic sequences of sounds. However, in the creation of verbal art, we do more of this and we do it more intensively. One could say that we seek to maximize this sort of patterning. In other words, we seek to render the causal sequence of the plot more rigorous, reenforcing it with foreshadowing and circularity. We seek to coordinate connotations and ambiguities of the words and phrases – including purely graphic connotations and ambiguities where these occur, as in Chinese (Cooper 68–72). We also seek to pattern the sounds through rhythm, assonance, alliteration, and so forth. One differentia of literature, then, would seem to be a maximization of patterning, a maximization that makes a wide range of features (narrative, semantic, phonetic, and so on) relevant to our literary experience.6

Because I use the word “relevance” here, a term associated with Sperber and Wilson, it is important to stress that my notion of relevance has no special relation to their work. Gibbs gives a nice summary of the Sperber/Wilson approach: “Relevance is defined in terms of contextual effects and processing effort. Contextual effects are achieved when a speaker’s utterance strengthens, contradicts, or denies an existing assumption or when a speaker’s utterance is combined with an existing assumption to yield some new contextual implications” (230). My use of “relevance” has no particular bearing on contextual effects or processing effort. In my usage, maximizing relevance is a matter of taking more and more properties and relations as germane to the experience of the literary work. Of course, these properties and relations may be germane in the sense that they “yield some new . . . implications.” But they need not be. Indeed, a wide range of poetically relevant features rarely bear on implications. More exactly, when I listen to a conversation, I simply do not attend to many features of the speaker’s language, such as assonance or patterns in syllabic stress (for example, iambs). Obviously, I pay attention to contrastive stress and other forms

The Mind and Its Stories

But this is only a first approximation. The Jakobson/Kiparsky hypothesis indicates that formal poetic devices are not merely a matter of maximizing relevance, but of maximizing a certain sort of relevance – linguistically specified relevance, in their view. As Kiparsky stresses, “certain patterns of considerable formal simplicity are never utilized in the construction of verse. . . . For example, no one thinks of filling in a stanzaic pattern on the principle that the last words of certain lines must contain the same number of sounds” (“The Role” 12–13). Why not? Kiparsky’s answer is that no linguistic rule involves counting sounds in this way and thus no poetic rule will do so. But this formulation is clearly too narrow. It covers the case at hand, but it does not cover, for example, imagery, foreshadowing, or other nonlinguistic patterns open to maximization. Many sorts of maximization are not narrowly linguistic and thus cannot be covered by Kiparsky’s principle. Moreover, it does not seem to provide an adequate explanation of the limitations on maximization even in the case of linguistic phenomena. One is left asking, “Why is there such a linguistic constraint?,” even in the relevant instances. So, Kiparsky is responding to a genuine problem, but he seems to be responding to it too narrowly. One way of trying to resolve this dilemma would be to consider what it is about Kiparsky’s account that allows it to solve the linguistic cases, then to see if this aspect of the account can be generalized. Consider again Kiparsky’s examples. One obvious and crucial difference between, say, onsets (or beginnings) of syllables (used both in language rules and in poetry) and number of speech sounds (used neither in language rules of semantically relevant emphasis, but that is different. Indeed, its difference is precisely the point. When I read poetry, I do attend to such features as assonance and recurring patterns in syllabic stress. They become relevant to my experience. But, unlike contrastive stress, assonance and patterns in syllabic stress do not necessarily, nor even typically, yield implications. The point is obvious as soon as one thinks of concrete examples. The presence of iambic pentameter is relevant to my experience of a huge variety of poems in English. However, with rare exceptions, it does not strengthen, contradict, or deny an existing assumption, nor does it yield any contextual implications, as is obvious from the fact that it is relevant to my experience of eulogies and satires, poems of victory and of defeat, poems of love and of hate, and so on. I do not find a poem of love contradicted by iambic pentameter – nor do I find a poem of hate contradicted by it. In short, the work of Sperber and Wilson, though important and insightful, just has no special bearing on my concept of maximizing relevance in literature.

Literary Universals

nor in poetry) is that we “hear” the former, but have to calculate the latter. More generally, any linguistic feature that is part of a linguistic rule is a feature we “hear” – not in the sense that we are conscious of it (typically we are not), but in the sense that it makes a difference to our experience. For the most part, we do not “hear” features not included in linguistic rules, but can at best calculate them. Indeed, when Kiparsky elaborates his hypothesis, he makes the point himself, arguing that the “faculty of language . . . equips” us with “modes of perceiving” certain features, but not others (Kiparsky, “On Theory” 191). Insofar as this notion can be generalized, it would seem to solve the problem we have been considering. And, as it turns out, the idea can be generalized easily. Indeed, it is well known in cognitive science. The linguistic “hearing” or “perceiving” of onsets (but not number of speech sounds) is simply a specific case of a more general cognitive mechanism, called “encoding.” Encoding, then, appears to be what is crucial in all these cases, both those that fit Kiparsky’s model (for example, alliteration) and those that do not (for example, foreshadowing). More exactly, whenever we perceive something, we do not perceive every aspect and relation of the thing. That would be impossible. Rather, we perceive some aspects and store them in memory while others escape us. The aspects we perceive and store are the aspects we “encode.” Holland, Holyoak, Nisbett, and Thagard give the following example: “[Y]ounger children often cannot learn about the rules underlying the behavior of balance beams” simply from observation because they “do not encode the distance of objects from the fulcrum.” However, when the distance is pointed out to them, they begin to encode the feature on new observations, and are able to induce rules (Holland et al. 55). Linguistic rules are just a particular case of cognitive principles that allow for encoding. This leads us to reformulate our earlier principle. Now we would say that a wide variety of formal literary techniques (alliteration, assonance, circularity, foreshadowing, and so on) function to maximize relevance or patterning across encoded properties or relations. These literary techniques could be thought of as particularizations of a relevance/patterning schema in which the properties or relations in question are values of variables, with the variables necessarily confined to the class of encodable values. Thus, when applied to onsets

The Mind and Its Stories

(which are encodable), this schema yields alliteration. When applied to speech rhythm (which is encodable), it yields meter and related forms of organization. When applied to the background constituents of a scene (for example, landscape), it yields a form of image patterning. Note also that encoding can be learned, though there are limits to what can be encoded. This is important because it seems clear that greater experience of and training in literature increases one’s sensitivity to certain sorts of patterns. This is only to be expected when the process is understood in terms of the general cognitive process of encoding. It is less obvious that Kiparsky’s hypothesis could accommodate this sort of development. On the other hand, not all of our problems are yet solved. Our revised formulation does not allow for limits on maximization; it treats maximization as a good in and of itself. Literature, however, does not absolutely maximize patterns (cf. Bateson 18). It does so only to a certain point. One degree of alliteration is aesthetic, but even a little more may be comic. This, too, is true cross-culturally. Some traditions may employ more alliteration than others, but, with rare exceptions, they do not equate more alliteration with a better poem (other things being equal), as this formulation would appear to imply. Rather, alliteration, rhyme, and so on reach a sort of ceiling, after which they detract from aesthetic effect. To explain this, we need to extend our analysis of perception. Again, in perceiving any object, we fail to encode some features at all; other features we do encode. Of those features we encode, some have, so to speak, forced themselves on us, drawing our attention. A well-known case of this concerns background conversation. Typically we are not paying attention to background conversation when we are engaged in a conversation ourselves. However, if we hear our name mentioned, our attention will suddenly shift to the background conversation, entirely independent of conscious decision (see Johnson-Laird 148), and perhaps even against our will. Miller and Johnson-Laird note that “surprising stimuli” have this obtrusive or attention-forcing effect as well (133). More generally, we might say that any perceptual feature has a certain degree of saliency for a particular perceiver in particular circ*mstances. (Features that are not encoded could be thought of as having

Literary Universals

a saliency of zero.) Various qualities of the feature, the perceiver, and the context determine that degree of saliency. As just noted, unexpectedness of the feature is one such quality. For temporally ordered occurrences, frequency of repetition would be another. Once the degree of saliency (resulting from these qualities) goes above a certain threshold, it automatically draws our attention. We can refer to this as the “threshold of forced attentional focus.” Consider the quality of frequency. Suppose I use the phrase “of course” more than other speakers. If I use it once every other paragraph, the attention of readers is unlikely to be drawn to this usage (that is, they are unlikely to “notice”). As frequency increases, however – suppose I use it in every sentence – the usage will become obtrusive, “drawing attention to itself.” In other words, it will eventually cross the threshold of forced attentional focus. This threshold probably varies somewhat both culturally and individually. However, it is also no doubt governed by broad cognitive constraints. We may further refine our general principle, drawing on this distinction as follows: A wide variety of formal literary techniques (alliteration, assonance, foreshadowing, circularity, and so on) function to maximize relevance or patterning across encoded properties or relations with a normative limit at the point where such maximization would surpass the threshold of forced attentional focus. There are still exceptions to this formulation. They typically involve an extension of forced attentional focus to the point where use of the feature comes to be seen as humorous or as manifesting a sort of cadenza-like virtuosity on the part of the poet. The precise nature of these exceptions would be further examined in an ongoing research program. However, this should adequately illustrate the abstraction of secondary principles and their refinement in such a program. Therefore, I should now like to turn to another aspect of such a research program, considering a nontechnical universal with a more direct relation to a particular aspect of cognitive structure. As we have already noted, standard line lengths for poetry in a wide range of traditions tend to fall between five and nine words. Standard line length is of course defined not in terms of words, but in terms of some acoustic property. However, it typically puts the number of words per line in this range – and that is what is crucial

The Mind and Its Stories

for achieving descriptive adequacy in this case.7 Thus, in Chinese, a monosyllabic language, one standard line length has only five syllables, which equals five words, while another has seven syllables and thus seven words (see Cooper 63). The Yirrkalla poems quoted by Berndt (Appendices 1 and 2), the Dinka songs cited by Deng, the Basotho verses quoted in Kunene, many of the Hawai’ian poems in Pukui and Korn, and the Babylonian creation poems discussed by Sandars (17) also fit this pattern reasonably well. As to European literatures, the first twenty lines of The Canterbury Tales contain 144 words, about seven per line; the first twenty lines of Paradise Lost average a bit under eight words per line; twenty lines taken at random from the Aeneid have just under seven words per line; and twenty from the Odyssey have almost exactly seven words per line. Blake’s Songs of Innocence have unusually short lines, but the first poem of the sequence (Kazin 83), which is in no obvious way formally different from those that follow, has six words per line. French lines tend to be unusually lengthy, but the first twenty lines of Racine’s Ph`edre include just under nine words per line and thus still do not exceed the range. Nonetheless, there are many exceptions. In considering these, we need, first of all, to determine which counterexamples fall within the range of phenomena we are seeking to characterize. Clearly, our concern here is with literary forms defined by fixed, recurrent, phonetic patterns, such as iambic pentameter. Noting this makes free verse irrelevant and thus eliminates a large number of possible counterexamples. Moreover, we are, as already noted, speaking about standard line lengths. Nonetheless, there remain a number of recalcitrant cases. One option would be to say that the universal is statistical and limited in application – well above chance, certainly, but with many 7

Different traditions formulate their criteria for poetic line length in different ways. The lines bear a much greater similarity than the stated criteria would indicate. To achieve descriptive adequacy in such a case, we need to specify the similarity among the lines as closely as possible. To do that, we have to set aside the official formulations (for example, the definition of the standard English line as iambic pentameter), because these do not reflect the cross-cultural pattern. In fact, they obscure it. I have attempted to achieve descriptive adequacy, capturing the crosscultural similarity, by formulating the pattern by reference to the number of words. This formulation also points directly to an explanatory account of the pattern in terms of cognitive structure, as I discuss as follows.

Literary Universals

exceptions. This is, of course, possible. But it is not a solution as we must go on to determine what gives rise to the exceptions anyway. Another option is to change the predicate of the universal. In every exception of which I am aware, the standard line length is less than five words, not more than nine. One way of dealing with these exceptions would be to revise the universal to say that all standard lines are less than nine words. However, the five- to nine-word spread of line lengths fits well with the structure of rehearsal memory or, equivalently, working memory (see Garman 322; for a fuller discussion, see Gathercole and Baddeley). Rehearsal memory allows us to “keep in mind” strings of information, cycling them through our attention as we try to complete some encompassing cognitive task. Specifically, rehearsal memory is structured in such a way as to include a limited amount of information – typically, five to nine words – at any given time. There are several reasons why this correlation is theoretically appealing. Most importantly, poetry in all traditions demands a sort of plenary attention, for which rehearsal memory seems ideally designed. More exactly, as the great tenth-century Indian theorist, Abhinavagupta, put it, we “savor” poetry: “Aesthetical experience takes place . . . by virtue, as it were, of the squeezing out of the poetical word. Persons aesthetically sensitive, indeed, read and taste many times over the same poem. In contradiction to practical means of perception, that, their task being accomplished, are no more of any use and must then be abandoned, a poem, indeed, does not lose its value after it has been comprehended” (quoted in Gnoli xxxii). Abhinavagupta’s view here is almost a necessary consequence of the maximization of unobtrusive patterning. As more features become relevant to our experience of a literary work, we are less and less able to appreciate it without “savoring.” More features must be encoded in our experience of poetry than in our experience of ordinary speech. The most obvious way of ensuring this is through rehearsal memory, which does, in effect, allow us to “savor” segments of a poem. It makes a great deal of cognitive sense that the unit of savoring would be the poetic line. Or, rather, given this need for savoring, it makes sense that the recurring unit of poetry would develop in accordance with the structure of rehearsal memory. Moreover, this is true not only receptively, but productively. A poet composing a

The Mind and Its Stories

poem is generating short, repeatable, nonsyntactic units. He/she has to revise and “polish” these units – to satisfy metrical and other constraints and the broader criterion of nonobtrusive maximization of patterning. Given the structure of human cognition, one would expect that any such unit would almost necessarily be structured in accordance with rehearsal memory. This is even more obvious when one takes into account the oral, bardic composition which is at the origin of poetry (on the nature of oral poetic composition, see Lord, Ong, and Rubin). Without the aid of writing, the recurring unit of poetic form would almost necessarily be structured by rehearsal memory. In short, rehearsal memory seems to provide a good explanation of the universal as initially stated. Indeed, it not only accounts for the five to nine words, but relates this to the maximization of relevance, and to other aspects of the experience and creation of poetry. Thus, we have broad theoretical reasons not to adopt a formulation of the line length universal that would dissociate it from rehearsal memory. Of course, if lines are only shorter and not longer, then it is still possible to link the poetic line with rehearsal memory. Longer lines would not fit into the limited space available for rehearsal memory; shorter lines fit, even though there is room left over. We do often use rehearsal memory in this way. On the other hand, the nature of any link with rehearsal memory is less clear if we accept the reformulated universal (“less than nine words”). In this case, it would seem that rehearsal memory does not structure the line, but merely limits its extent. Is there a separate, structuring principle, then? The majority of standard lines appear to fall within the five- to nine-word range and thus to be structured by rehearsal memory. How are we to explain this? One option that preserves the straightforward relation to rehearsal memory is to reconsider the notion of the line. Indeed, there seems no necessary reason to identify the unit at issue with what is, at least to some extent, a convention of writing and printing (a similar point has been made by Frederick Turner [73–4]). In the great majority of cases, the printed line is indeed the unit we want. In the case of the alexandrine or iambic pentameter or a wide range of other patterns, the printed line rightly defines the recurring nonsyntactic unit. In other cases, however, this is less clear. One of the most obvious exceptions to the universal as initially formulated is the Japanese haiku, a seventeen-syllable poem divided into sections of five, seven,

Literary Universals

and five syllables. We typically conceive of this as a three-line poem. One way of reconciling the haiku with our universal and with the structure of rehearsal memory would be to conceive of the poem as a single line with a single caesura. This fits perfectly well, yielding roughly six to nine words per unit (see, for example, the haiku in Keene, Anthology, 361–9). An alternative account that also preserves the structural function of rehearsal memory, might incorporate silent beats, akin to musical rests. This would fit with certain aspects of Japanese aesthetic theory. Moreover, the general phenomenon is well attested. As Rubin points out, in certain English poems, “beats must be allowed to fall on rests. Although our conventional writing system ignores rests . . . their presence in verse can easily be heard by trying to recite” certain poems “without including the appropriate rests” (252–3n.3). On the other hand, a further source of exceptions may be found in such languages as Kuna and Dyirbal, which tend to have between two and four words per line (see Sherzer 107–10 and the poems in Dixon and Duwell), and where the problem does not appear to be one of defining the recurring rhythmic unit. (There is also no obvious motivation for positing silent beats in these cases.) These counterexamples rather lead us to a reconsideration of the nature of rehearsal memory. Both Kuna and Dyirbal are highly morphologically complex languages. It may be that rehearsal memory is not appropriately measured in terms of words at all, but in terms of, say, morphemes. Even in European languages, it seems odd to count “habeo” (“I have”) as one word, but “j’ai” (“I have”) as two, or “et” (“and”) as one word, but encl*tic “-que” (“and”) as none. Perhaps “et virum” (“and man”) takes up two slots in rehearsal memory and “virumque” (“and man”) takes up only one, but this seems counterintuitive. In any case, at this point, the research project of literary universals abuts the broader research project of cognitive psychology. Before the literary project can proceed, the broader psychological study of rehearsal memory has to proceed further. Moreover, as this example indicates, aspects of this broader study can be inspired by literary questions, data, and hypotheses. The problems of Kuna and Dyirbal point to further areas for cognitive research – specifically, the examination of whether rehearsal memory is best understood in terms of words, morphemes, or something else. Should it turn out that it is

The Mind and Its Stories

best understood in terms of words, then the Kuna and Dyirbal data will be difficult to account for and we may have to weaken or modify our hypothesis. Should it turn out that it is morphemes, this will solve the Kuna and Dyirbal problems. On the other hand, it may render Chinese problematic. Should it turn out to be something else, this too will probably produce anomalies. Indeed, some research indicates that the relevant parameter, or one relevant parameter, is something we have not considered – subvocalization time. Rehearsal memory, in this view, should not be measured in words (five to nine), but in seconds (about two; see Gathercole 20).8 If correct, this solves the problem of highly morphologically complex languages, since the subvocalization time for each word in such a language would be much longer than for a word in a language such as English. It may suggest that we not treat a Haiku as a single line, for an entire Haiku would seem to take longer than two seconds for rehearsal. By the same token, however, it makes our alternative account of Haiku more plausible, for it indicates how pauses or silent beats could occur in a cycle of syllabic verse. On the other hand, the Chinese data now become more difficult to explain (assuming that the subvocalization time of Chinese lines is as short as it appears). Once again, there is no contradiction as the Chinese lines do not exceed the limit. But we are still left with the question of why the Chinese standard line would apparently be so short. 8

Six years after writing the first version of this argument, I came upon Frederick Turner’s essay, “The Neural Lyre.” In that essay, Turner sets out to explain the universality of line length in poetry. His approach is based on articulation time, which he finds to be about 3 seconds in most traditions – though his tabulation does include instances ranging from 2.2 seconds, for the Chinese four-syllable line, up to 3.9 seconds, for the Latin Alcaic strophe (76). As just indicated, Turner may have been right to emphasize time rather than words or some other variable. But his temporal account would need to be reworked in a research program, and not only in order to treat the anomalous data just cited. First, he almost certainly needs to consider subvocalization time rather than vocalization time. Second, he probably should address the issue of standard or, equivalently, prototypical line lengths. Again, my own hypotheses have concerned standard line lengths. It seems to me very unlikely that rules governing line length can be formulated in terms of necessary and sufficient conditions (that is, in such a way as to cover all line lengths, rather than standard – thus prototypical – line lengths). Finally, Turner offers “the three-second present moment of the auditory information-processing system” as the source of universal line length and connects this with the effects of poetry, which in his view operates “to improve the memory, and to promote physiological and social harmony” (103). This at least requires further clarification and development, not to mention empirical study.

Literary Universals

Another option is to view the structure of working memory as more complex than any of these accounts allows. Perhaps the number of morphemes, or the extent of phonological information (for example, the use of tone in Chinese) combines or interacts with subvocalization time in the structure of working memory. Indeed, we know that rehearsal memory is affected by a number of variables, not just one. As Kintsch, Healy, Hegarty, Pennington, and Salthouse, put it, “there is no single, all-encompassing capacity-limiting factor for working memory” (420; see also Miyake and Shah, “Toward” 447). For example, features of sound (especially rhyme) affect rehearsal memory, as does the lexical status of items in the rehearsal string (roughly, words are easier to recall than nonwords). (For a discussion of this research, see Gathercole.) A further possibility, consistent with the final point, is to set aside working memory as a distinct part of cognitive architecture and account for the relevant data – such as the five- to nine-word capacity for rehearsal – by reference to other structures and processes. For example, in Barnard’s theory, there are acoustic, articulatory, morphonolexical, and other subsystems that bear on the phenomena of “working memory.” Thus, the “properties of performance on working memory tasks,” including the five- to nine-word pattern, “must be attributed to process-mediated interactions among multiple subsystems of cognition” (313). An account such as this opens the possibility that several of the preceding suggestions have bearing on the determination of standard line length, with morphology, phonology, subvocalization time, and other factors interacting in complex ways. Clearly, there is a great deal of work here – not only in one research program, but in several competing programs. The maximization of unobtrusive patterning and the relation between rehearsal memory, line length, and aesthetic experience nicely illustrate the process of descriptive and explanatory study in a theory of literary universals.9 However, as such, they remain mere starting points for research, hypotheses to be modified, elaborated, replaced. 9

The importance of explanatory adequacy is not confined to the study of absolute universals. An excellent illustration in the study of typological universals is provided by David Rubin. Rubin articulates a powerful cognitive analysis of oral composition in terms of cues for and constraints on memory. The development of epithets and formulae is directly predicted by his theory, as is the typological difference between oral and written composition.

The Mind and Its Stories

Again, the study of literary universals, like the study of linguistic universals, is a project that can progress only through the cooperative efforts of a broad range of researchers engaged in an ongoing process of empirical reevaluation of theories and theoretical reorientation of empirical research. As the concluding discussion in particular illustrates, such a program of research could be greatly valuable, not only for our understanding of those cognitions and affections that generate and sustain literary art, but for our broader understanding of the human mind as well.

conclusion In sum, there are many types of universals – absolute, statistical, implicational, typological, and so on. There are also well-defined criteria for determining universality – recurrence across a higher percentage of genetically and areally distinct traditions than would be predicted by chance. By these criteria, there appear to be many literary universals. These may be roughly organized into techniques and nontechnical correlations. Many literary universals may be partially understood through the abstraction of secondary principles, such as the maximization of unobtrusive patterning. Others may have a more direct relation to a cognitive (or other) structure or process, such as the capacity of working memory. In each case, the study of literary universals seeks both descriptive and explanatory adequacy through the development of a research program in cross-cultural, comparative literary study and the (mutually beneficial) integration of this program with research programs in cognition, as well as history and other related fields.

2 Emotion and Suggestion Lexical Processes in Literary Experience

The study of non-European literary theories is particularly valuable for the isolation of literary universals. Such theories often carry insights into universal patterns that are occluded by European theories, thus patterns that readers may miss, if their attention and understanding are too guided by European approaches. Indian theories are unusually significant for the present study as they are particularly compatible with recent developments in cognitive psychology. Perhaps this is due to the fact that Indian literary theory developed partially in relation to Indian linguistic theory, which is widely considered to have close connections with Chomskyan developments (see, for example, Kiparsky’s P¯an.ini). In any case, starting more than two millennia ago, and extending over a millennium, Sanskrit writers developed an elaborate theory of poetics. This theory reached a culmination and a sort of theoreti¯ cal impasse in the writings of Anandavardhana and Abhinavagupta. With the aid of cognitive science, we may recognize the value of Sanskrit poetics and develop it beyond the impasse it had reached in the eleventh century. I hope to show that, when redeveloped in the context of cognitive science, the work of these writers provides us with a plausible and productive cognitive theory of poetic feeling, a

The Mind and Its Stories

theory that is not a conclusion, but the start of a larger research program in the field.1 More exactly, I shall begin with a broad outline of Sanskrit poetics ¯ from its beginnings through Anandavardhana (for a more detailed discussion, the reader may wish to consult a history of Sanskrit poet¯ ics, for example that of Gerow). Roughly speaking, Anandavardhana sought to develop and systematize previous ideas in Sanskrit theory in order to provide an adequate description of poetic effects. ¯ Abhinavagupta took Anandavardhana’s descriptive ideas and sought to provide an explanatory framework for them. The second section is consequently devoted to the explanatory views of Abhinavagupta. In the third section, I turn to contemporary cognitive science, first presenting what I take to be a plausible, partial theory of the internal lexicon (or mental dictionary/encyclopedia), then going on to reformulate Abhinavagupta’s views in terms of this theory. The fourth section responds to some possible misunderstandings of the resulting account, by distinguishing among the objects, causes, and sources of emotion. The chapter concludes with a brief interpretive illustration from Hamlet.

anandavardhana and suggestion ¯ Histories of Sanskrit poetics generally divide early Indic theories of verbal art into two broad traditions: alam . k¯ara” . k¯ara and rasa. “Alam or “ornamentation” refers to a range of poetic devices and rhetorical figures from alliteration to metaphor. In itself, the alam . k¯ara tradition is of limited interest, for it hardly extends beyond taxonomy (for example, listing and describing the varieties and subvarieties of figures such as simile). On the other hand, without this often tedious program of analysis and categorization, including its rudimentary lit¯ erary semantics, Anandavardhana could never have formulated his seminal theory of dhvani or “suggestion,” to which we shall turn in a moment. ¯ Anandavardhana’s work is, however, a culmination and synthesis, not only of alam . k¯ara analysis, but of rasa theory as well. Rasa 1

Indeed, since the first version of this chapter was published, such a project has made at least a tentative beginning – most significantly, in Keith Oatley’s work (see “Emotions”).

Lexical Processes in Literary Experience

is usually translated as “sentiment.” It is distinguished from bh¯ava or “emotion,” to which it is, nonetheless, closely related. Specifically, bh¯ava is what we feel in ordinary life – love, sorrow, happiness, anger, and so on. Rasa, in contrast, is what we feel in experiencing a work of art. It is akin to emotion, but not identical with it. (In the rest of this chapter, I shall use the English word “emotion” in place of the Sanskrit “bh¯ava” as the two are close enough in meaning for present purposes; I shall, however, most often use “rasa” rather than “sentiment” as “rasa” is a technical term with no precise ordinary language equivalent in English.) Specifically, when I watch a romantic play, I do not actually love the hero or heroine (as I love my wife), but I do experience some sort of feeling. Moreover, this feeling is related to love in a way that it is not related to sorrow or anger. Thus the Sanskrit theorists say that I am experiencing the “erotic rasa,” not the emotion of love per se. The earliest extant development of this theory is in the N¯a.tya´sa¯ stra or Treatise on Dramaturgy attributed to Bharatamuni, but composed by a number of authors between the second century b.c.e. and the sixth century c.e. This foundational volume lists eight primary emotions and corresponding rasas: love/the erotic, sorrow/the pathetic, and so on. (Lest this seem too restrictive, they acknowledge a wide range of ancillary feelings also.) Each literary work, in order to be aesthetically effective, was required to have one dominant rasa. This is not to say that other rasas could not enter; they not only could, but must. However, any additional rasas should function to further the dominant rasa. For example, suppose that the dominant rasa of a work is the erotic. Then it makes perfect sense to bring in the pathetic rasa. The pathetic may be part of the erotic, just as sorrow may be part of love. (Indeed, another way of thinking about the relation between rasa and bh¯ava is that the characters experience the bh¯avas, such as love and sorrow, while the readers/spectators experience the rasas, such as the erotic and the pathetic.) But the pathetic rasa in the work must operate to contribute to the erotic rasa. The same constraint holds if the dominant rasa is the pathetic and the erotic is subsidiary. If the subsidiary rasa does not contribute to the dominant rasa, the experience of the reader/spectator will push in different directions and the overall aesthetic experience will be weakened; it will not be a satisfactory experience of the erotic or of the pathetic.

The Mind and Its Stories

¯ Again, Anandavardhana combined the two strains of Sanskrit poetics, developing them into a unified theory of aesthetic response, in part based on the concept of dhvani or “suggestion.” By the time ¯ of Anandavardhana, Sanskrit theorists of “figures of thought” had isolated several varieties of meaning. At this level, the distinctions with which they were operating were roughly the same as those operative in western poetics/semantics today. They had concepts of literal meaning, idiomatic meaning, and various metaphorical and related meanings (encompassing simile and other figures); they had also isolated a number of ways in which nonliteral meanings could be manifested, recognized, and interpreted (for example, through some explicit marker, such as English “like” and “as” in similes). We would consider much of this fairly straightforward today. “V¯ar¯an.as¯ı is on the Ganges” does not literally mean that V¯ar¯an.as¯ı is on the Ganges; it means (idiomatically) that V¯ar¯an.as¯ı is on the bank of the Ganges. “Moon-faced beauty” does not literally mean that the woman’s face is the moon, but that it is rounded and fair, and so on. More interestingly, Sanskrit theorists had isolated a variety of complex poetic and other implications – or, rather, nonlogical “implicatures,” to use Paul Grice’s term – both literal and nonliteral. For example, a message sent by a woman to her lover, “The lion, they tell me, does not prowl at the riverbank,” would involve a complex metaphor/implicature to the effect that, if the woman and her lover meet at the riverbank, they will not be caught. ¯ Anandavardhana systematized this work by distinguishing varieties of dhvani that encompass these nonliteral meanings. Most importantly for our concerns, however, he maintained that there was one sort of meaning that is not part of this typology and is most appropriately the referent of the term “dhvani.” This “dhvani proper” is not an idiom or metaphor (or simile, metonymy, and so on); nor is it an implicature (cf. Amaladass 92–3; unless otherwise noted, I shall use “dhvani” to refer to this “dhvani proper”). Ex¯ tending Anandavardhana somewhat, we may say that it is, rather, a nonparaphrasable suggestion of a word, phrase, sentence, topic, or (linguistically constructed) situation (see Amaladass 105). To say that the dhvani of a given utterance is nonparaphraseable is not to ¯ say that one cannot say anything about it (as Anandavardhana emphasizes [671]). Quite the contrary. One can say many things about

Lexical Processes in Literary Experience

it. It is not unparaphraseable because of being ineffable. There are, rather, three reasons why dhvani is unparaphraseable. First, it is infinitely ramified. We can never enumerate all the suggestions, even all the relevant suggestions, of a given work (cf. Abhinavagupta, Locana 206). This is not, however, a distinctive property of dhvani. For example, some metaphors are, roughly, paraphraseable in the sense that one can list all the relevant information conveyed by the metaphor (especially those metaphors that are close to idioms). But many metaphors are not fully restatable in this sense. And at least for larger texts, even the relations among literal meanings cannot be ennumerated exhaustively. The second reason that dhvani is not paraphraseable is more important and more distinctive. Indeed, it indicates that, in a sense, dhvani is not even partially paraphraseable. In the language of analytic philosophy, the dhvani of a text cannot be substituted for the text with a preservation of truth value (cf. Abhinavagupta, Locana 81). Indeed, dhvani is usually not statable in the form of a proposition that might have or not have a truth value. Consider a metaphor partially explicit in Derek Walcott’s highly mythic play about AfroCaribbean identity, Dream on Monkey Mountain: “Makak is a lion.” As a metaphor, this means that Makak is brave, ferocious to his foes, and so on. (For the purposes of the example, we can assume that it is not fully paraphraseable and thus that this list will not come to an end.) But note that each of these interpretations is substitutable for “a lion” with a preservation of truth-value. “Makak is brave” and “Makak is ferocious to his foes” are just as true as “Makak is a lion.” (I leave aside the issue of how to specify “metaphorical truth.” Clearly, at least at a certain point in the play, this sentence is [ambiguously] true in a way that, say, “Makak is a gazelle” or “Makak is a unicorn” is not.) In contrast, consider the standard example of “a village on the Ganges.” This phrase, the Sanskrit commentators emphasize, suggests holiness due to the holiness of the sacred river Ganges. However, there is no way in which the word “holiness” can be substituted for the phrase or for part of the phrase. Take, for example, the beginning of a story, “R¯am was the headman of a village on the Ganges.” The truth value of this statement is not preserved with the relevant substitution: “R¯am was headman of holiness.” Moreover, there is no way that this suggestion can be turned into a proposition. For

The Mind and Its Stories

example, the suggestion of holiness does not imply or implicate that the village itself is holy or that any of the villagers (for instance, R¯am) is holy. Returning to Walcott, we might say that the name “Makak” (or “macaque”) suggests a whole range of west African mythological motifs concerning the divinity of the monkey, and at the same time a range of contrary European attitudes in which “monkey” serves as a racial insult. These are not substitutable for “Makak.” Nor do they imply or implicate any propositions. Even at this level, the notion of dhvani seems to be substantially different from standard concepts of literary semantics in the western tradition. For example, though “connotations” are not viewed as truth preserving, they are typically conceived of as involving some sort of assertion. Western theorists may well say that “a village on the Ganges” connotes “holiness,” but I suspect that they would unreflectively think of this as part of a (perhaps ironic) assertion – that the village is or should be holy, that the villagers have a special duty to be holy, and so on. For this reason, it would seem somewhat odd to say that various Yoruba beliefs are a “connotation” of the name “Makak,” for, again, none of these beliefs is asserted (or denied) by this “connotation.” But this is not all that distinguishes dhvani from common western notions – and this brings us to the final and most important reason why dhvani is not paraphraseable. Dhvani is not purely semantic. It is ¯ affective as well. Specifically, in the strictest sense, Anandavardhana tells us, dhvani is rasadhvani, the dhvani of rasa – not the intellectual implication of some sentiment, but the “suggestion” of a rasa as an affective experience (see Abhinavagupta’s comments on ¯ Anandavardhana in Locana 70). In other words, dhvani is not paraphraseable, most importantly because it is not solely, nor perhaps even primarily, a meaning; as rasadhvani, the “truest” form of dhvani is bound up with feeling. It is, in short, an experience of rasa. A literary portrait of a village on the Ganges involves a full rasadhvani of holiness – or, more aptly, sacred peace – only if it gives us a feeling of that holiness or sacred peace along with the more narrowly semantic suggestion. Indeed, the feeling is what is crucial. The name “Makak” communicates not only the idea of west African beliefs and customs carried by slaves to the new world, but feeling as well – perhaps sadness (the pathetic rasa) over the partial loss of these

Lexical Processes in Literary Experience

beliefs and customs, or anger (the furious rasa) over their colonial denigration. In sum, aesthetic response is a matter of the experience of rasas. These rasas are evoked in a reader by words, sentences, topics, and so on, presented in a literary work. This is, of course, in part the result of literal meanings. But it is also, and crucially, a function of the clouds of nondenumerable, nonsubstitutable, nonpropositional suggestions that surround these texts – what Keith Oatley has recently called the “Suggestion Structure” of the work (“Emotions” 45, 51–9). Finally, it is important to add that, in the overwhelming majority of cases, these rasas are evoked without our having any explicit awareness of suggested meaning. We do not, in other words, self-consciously infer some semantic suggestion, then feel the rasa. Rather, we experience the rasa as we watch the play or read the poem. Indeed, if we have to stop and work through the suggestions, we may not experience the rasa at all.

abhinavagupta, memory traces, and aesthetical feeling Abhinavagupta turned his attention away from the linguistic and ¯ related abstractions that had preoccupied even Anandavardhana, focusing his attention instead on the human mind, specifically the mind of the reader or viewer of a literary work. The first step in Abhinavagupta’s project involved the, at least tacit, recognition that the theory of rasadhvani could not be understood as a theory of abstract linguistic structure. Rather, it only made sense as a theory of the way people respond to literature. In other words, rasadhvani had to be conceived in psychological terms. Broadly speaking, Abhinavagupta was a transcendental realist whose philosophy involved a strong empiricist component. (A Hindu theologian himself, he was one of the harshest opponents of the Ved¯antist view that perceptual reality is illusion or m¯ay¯a; see, for example, B.N. Pandit 23.) His theory of mind was, consequently, a realist theory. Unsurprisingly, it was broadly similar to other (eastern and western) theories of mind in isolating perception, memory, and other components or faculties. For our purposes, the most important part of Abhinavagupta’s theory of mind is his theory of

The Mind and Its Stories

memory, both storage and recollection. For Abhinavagupta, all experiences – perceptual, verbal, emotional – leave “traces” in the mind (see Aesthetic Experience 79). Drawing out the implications of Abhinavagupta’s analysis, we may understand these traces as having two components: one representational, one emotive. The emotive part is not the abstract recollection of one’s having had an emotion – such a recollection would be representational. It is, rather, a sort of reexperiencing of the emotion. It is not, in other words, remembering that one was sad, happy, or frightened at a given time and place, but actually feeling again, in some degree, that sadness, happiness, or fright. The point is most obvious with respect to strong emotions. For example, if one recalls a deceased friend or relative, one will have certain visual and other impressions, one will remember certain facts, and one will probably experience again, in a more or less attenuated form, the sense of loss, the sorrow that one felt at his/her death. These traces, Abhinavagupta tells us, are usually latent in our minds. At times they are fully activated – that is, at times we recall these memories and reexperience the emotions. This is all commonsensical enough. But, Abhinavagupta continues, there are also times when these traces are neither latent nor fully activated – and these are the most crucial. In other words, there are times when these traces are in some sense activated, but are not brought into self-reflective consciousness. More exactly, there are times when we are not selfreflectively aware of the representational content of the trace, and yet feel some hint of associated affect. This is, I think, common in our experience, though we may not immediately recognize the fact. I enter a building and am suddenly sad; I ask myself why and then recall an embarrassing or unpleasant event that occurred the last time I was in the building. Clearly, the memory had in some sense been activated (that is, it was not fully latent) and clearly the associated affect had bled into my conscious experience (that is, I felt sad), but the representational content of the memory was initially not conscious. This sort of analysis is what allows Abhinavagupta to explain the ¯ phenomena isolated and described by Anandavardhana. Specifically, Abhinavagupta indicates that rasadhvani operates in the following manner. Through dhvani, the literary work activates memory traces in the mind of the reader (see Abhinavagupta, Locana 81) without bringing these memories into consciousness. Again, these traces may

Lexical Processes in Literary Experience

be activated by words, phrases, topics, events, or whatever. Thus stories of suffering will activate memories of suffering, stories of romantic love will activate memories of romantic love, and so on. Once these traces are activated, the associated emotions seep into consciousness – again, not as ideas, but directly as feelings. The experience of the rasa of a literary work is precisely the experience of these feelings. More exactly, developing Abhinavagupta’s ideas, we may say that all speech involves the activation of memory traces with their associated affect. However, most often, these traces are activated in a haphazard and noncumulative manner, or else they are fully recalled. What makes aesthetic experience distinctive is that such activations are not fully recalled, but they are patterned; they are focused on traces of a specific type, which is to say, traces that bear to one another a certain similarity in both representation and affect (for example, in being memories of death and feelings of sorrow). While any given activation is likely to produce only very limited, indeed imperceptible and fleeting experiences of affect, this sort of repeated, patterned activation should result in a more pronounced and continuous experience. As Abhinavagupta puts it, “the basic emotion is put to use in the process of relishing [a work of art]: through a succession of memoryelements it adds together a thought-trend which one has already experienced in one’s own life” (Locana 117; see also 182 – “the relishing of beauty arises in us from our memory bank of mental states which are suitable” to the “basic emotions” of the characters – and Aesthetic Experience 112). Abhinavagupta extends this idea somewhat further when he argues that aesthetical pleasure results from the “generalization” of emotion in rasa, which is to say, its removal from the self-interest that is part of the link between the affect and the representational content in memory traces. In other words, when we fully remember a trace, the emotion that we experience is tied to self-interest (for example to our own personal loss of the loved one whose death we are recalling). However, through literature, we experience a version of the affect removed from its direct link with any particular (egocentric) representation in memory, and thus at least partially removed from self-interest (see Abhinavagupta, Aesthetic Experience 86–7 and 96–7). In this way, rasa may be redefined as emotion isolated from such self-interest and may even be compared, in Abhinavagupta’s view,

The Mind and Its Stories

with the experience of religious enlightenment or moks.a, where such self-interest is entirely extinguished (see, for example, Locana 226). (Indeed, for Abhinavagupta, the dominant rasa of a successful work always resolves itself into s´ a¯ ntarasa, the rasa of peace, a temporary and partial version of the endless and perfect peace that accompanies moks.a.) Another way of putting the general point is to say that the feeling of rasa is empathic. To some extent, I have explicated and developed Abhinavagupta’s ideas in a manner coherent with and partially derivative of contemporary cognitive science. However, I have not, I believe, fundamentally altered his claims or basic concepts. Indeed, it is my view that current theoretical and empirical work allows us to understand those claims and concepts more thoroughly than could Abhinavagupta himself, because he lacked a psychological theory adequate to his literary insights. However, I have refrained thus far from giving a full blown, reductive account of Abhinavagupta’s ideas within a broader theory of cognition. In the following section, I shall attempt something along these lines: a translation of Abhinavagupta’s theoretical premises into one version of a theory of mind that I take to be plausible in light of recent research in cognitive science. As I have already indi¯ cated, cognitive science has little to add to Anandavardhana’s and Abhinavagupta’s ideas about literary response. My concern, then, is to situate these ideas in a current theoretical framework that is part of ongoing research programs.

cognition and lexical structures Due to the nature of Abhinavagupta’s hypotheses, I shall be concentrating on the storage and accessing or recollection of representational knowledge (that is knowledge about something, including memories, and so on, as opposed to procedural knowledge, knowledge of how to do something). I shall assume that all representational information is stored in and accessible through a single long-term memory “unit.” This unit includes a broad range of information which, in everyday life, we would be inclined to divide between a dictionary (meanings of words), an encyclopedia (general facts about things), and a biography or personal archive (propositional and perceptual memories of individual history). This is a controversial assumption.

Lexical Processes in Literary Experience

However, the following analyses could be rewritten – with some loss in elegance – in terms of two or three separate, but closely interrelated units. Thus, nothing in this analysis is contingent on there being just one unit. In any case, following one standard usage, I shall refer to this unit as the “lexicon.” I should emphasize that this term should not be taken to imply a sort of unidirectional organization and access: from dictionary/meaning through encyclopedia/belief to personal archive/memory. In fact, lexical “entries” (for example, that for “monkey”) are structured to allow access not only from words to memories, but from perceptions (for example, seeing or hearing a monkey) to words, from memories to beliefs, and so on. Indeed, just as there are words for which we have no beliefs or memories, and some for which we do not have even basic meanings (“I’ve heard that word before; what does it mean?”), there are also uncategorized perceptions and perceptual memories (“I’ve seen that before; what is it?”). Thus the mental lexicon is no more a “lexicon,” in the common sense of “dictionary,” than it is a “mnemonicon” or an “aistheticon” (or an encyclopedia). Fortunately, however, these complications do not directly affect our concerns, which are with texts, and thus roughly fit the commonsense model of organization from “meanings” to beliefs and memories. Were we to extend this study to aesthetic response in, say, the visual arts, however, we would have to be more careful not to allow the prejudicial nature of the term to mislead us. In the present context, then, we may think of the lexicon as structured into “lexical entries,” which is to say, meanings, perceptions, and so on, clustering around lexical items, such as “monkey,” “dhvani,” “chant,” “compose,” “saffron,” and so on. These entries are multiply cross-indexed such that each entry is part of a number of networks that allow access across entries. Thus, “monkey” is linked with “ape” and “chimpanzee” in one network (of related species), “Africa” and “India” in another network (of habitats), and so forth. These networks may themselves be categorized by some lexical item (for example, “primate”), but they need not be – though, of course, one can always construct a category for a given network (for example, “habitats of monkeys”) and, indeed, we often use such ad hoc categories in actual cognition. As this indicates, our mental lexicon – a structure that exists only in individual people’s minds and thus varies somewhat from person

The Mind and Its Stories

to person – is different from a dictionary or encyclopedia, not only in content, but in structure. Perhaps most importantly, it does not have a single strict organizing principle (such as alphabetical order). Rather, our mental lexicon can be reordered in a variety of ways to suit our accessing needs. For example, we may access words by first consonant (as in a dictionary, more or less) or by some other phonetic property, such as final syllable – if, say, we are writing an alliterative, rhymed poem. In most contexts, however, we are likely to access items by topic. One way of thinking about this is in terms of the networks just mentioned. When discussing animals, we reorder our lexicon to make, for example, “monkey” more directly or swiftly accessible than when discussing computer technology (see Garman 293). Moreover, this effect is cumulative. If we are discussing Africa and animals, then “monkey” will be more accessible than if we are discussing animals alone. This “extra” accessibility of related terms, concepts, and so on, is standardly interpreted in terms of “priming.” In this view, the introduction of one item “primes” cross-indexed items (for example, “monkey” will prime the entries for “ape,” “primate,” and so on). At this level, priming can be understood as a reordering of the lexicon, such that the primed items are those that are placed highest in the order of a lexical search. For example, when we read or hear the syllable “mon” at the beginning of a word, we begin searching the lexicon for a “fit.” When the topic is Wall Street, we will reach “money” first; when the topic is animals related to apes, we will reach “monkey” first. (In fact, the situation is somewhat more complicated than this, but the point is adequately valid for present purposes.) On the other hand, priming has complex effects not only on access, but on other aspects of comprehension and response. In this way, it is not simply a matter of reordering the lexicon. There seems to be a change of status in the primed network. Primed items are, in effect, brought out of long-term memory, though they are not accessed directly in consciousness. When speaking of primates, I will have part of one lexical entry in consciousness, and parts of several lexical entries in rehearsal memory, which may be understood as circulating through consciousness. I also have a whole range of material stored in long-term memory (for example, the meaning of “ambidextrous,” visual images of my first day at school, and a vast range of other currently irrelevant stuff). A network of primed lexical entries is in

Lexical Processes in Literary Experience

a different mental state from either the conscious/rehearsal material or the material stored in long-term memory. Consider the following sentence: “Animals commonly associated with Africa would include lions, tigers, elephants, and apes.” When I am directly conscious of the word “ape,” the rest of the sentence is in rehearsal memory, and ¯ ınus”) are in long-term various irrelevant entries (for example, “Plot¯ memory. The lexical entry for “monkey,” however, is primed and is therefore in a state different from “apes” and “lions,” on the one ¯ ınus,” on the other. The primed entries are, we might hand, and “ Plot¯ say, placed temporarily in a sort of buffer between long-term memory and consciousness (or between long-term memory and rehearsal memory), a buffer that operates in part to allow access with minimal search. This gives us a basis for rearticulating and reunderstanding Abhinavagupta’s notion that traces may be activated while not consciously recalled. Specifically, we may conceive of dhvani as lexical networks primed and stored temporarily in the memory buffer. Note that these networks may “decay” rapidly, which is to say, drop quickly out of the buffer if they are not repeatedly primed. (The rapid decay of priming effects is well documented; see, for example, Garman 294.) However, when repeatedly primed, they would yield a pattern of unstated suggestions of precisely the sort we discussed previously as defining dhvani. To get a more detailed idea of how rasadhvani might work, however, we need to consider not only the relations between lexical entries, but their internal structures as well. Here, as elsewhere, cognitive scientists are not unanimous in theories or even terminology. I will use the terms “schema,” “prototype,” and “exemplum” to define three types of substructure within a lexical item. By “schema,” I mean a hierarchy of principles defining a lexical item. The hierarchy is based on “definitiveness” or “centrality” of the properties; the most central or definitive properties are at the top of the hierarchy, with increasingly peripheral properties listed in descending order. For example, being organic is more central to our conception of a human than is having two arms – a person with one arm would count as a human, but a statue with two arms would not. (This roughly recapitulates the distinction between the dictionary meaning and encyclopedic factual beliefs, but the distinction is one of degree, not of kind; some elements

The Mind and Its Stories

are “more definitive” than others, rather than one set of elements being “the definition” and another being “the empirical beliefs.”) As virtually all cognitive scientists emphasize, this hierarchy involves a number of “default” options. These are properties or relations that we assume unless we are told otherwise. Thus, “human” includes “having two arms” unless we are given information about a birth defect or amputation. By “prototype,” I mean, first of all, a sort of concretization of the schema with all default values in place, including those that are relatively unimportant in our schematic hierarchy (cf. Johnson-Laird and Wason 342). Again, these vary somewhat from person to person, but probably all of us have a prototype of, say, “man” as having two arms, two legs, and so forth. In addition, the prototype will have some “average” properties as well, properties that we would not ordinarily think of as “default assumptions.” For example, one person’s prototype human might be brown haired (that is, not bald and not grey, blonde, or black haired), clean shaven, of medium height, and so on. This is probably not a simple averaging, but a “weighted” averaging, with certain instances counting more than others – in part due to salience, but due to other factors as well. An important part of weighting is contrast (cf. Tversky, Ortony “Beyond,” and Barsalou 212). When forming our prototype of men, we tacitly weight more heavily those individual men who contrast most strongly with women. Thus our prototypical man will probably be more “manly” than the average man. Likewise our prototypical case of sadness will be sadder than the average. Related to this, certain prototypes have a normative element, as several authors have noted (see, for example, Kahneman and Miller 143). This is probably a function of the degree to which the term itself is considered normative. For example, our prototypical man or woman is probably better looking than the average man or woman and our prototypical surgeon is probably more skillful than the average surgeon.2 2

For my purposes, it is not terribly important to determine just how prototypes arise – what particular cognitive architecture they involve and just what cognitive processes and structures constitute them. I tend to speak in a “representational” idiom. Someone else might wish to adopt a purely connectionist idiom (for a connectionist treatment of prototypes, see McLeod, Plunkett, and Rolls 84–8, summarizing work by McClelland and Rumelhart; to be made consistent with the preceding

Lexical Processes in Literary Experience

Finally, by “exemplum,” I mean any specific instance of a category. Thus any man I know is an “exemplum” of man. The more common term here is “exemplar.” However, use of this term has been confusing. Hampton points out that “Exemplar models are . . . frequently account, this particular connectionist network would have to be revised slightly to allow some properties to weigh more heavily than others). Barsalou presents an intermediate account that involves representationalist categories in a spreading activation network. Specifically, Barsalou argues that prototypes are well-established networks of connection among concepts, instances, and properties (212–13; this account too may require some tinkering to incorporate ideals). For my purposes, choosing a particular account is not important at this point. On the other hand, there are properties that such an account should have. Specifically, George Lakoff has stressed that experimental research gives us only “prototype effects,” not prototypes per se. The point is, of course, just a generalization of all scientific inference. Studies of falling objects, the tides, and planetary motion do not give us some direct experience of gravity. Rather they give us data that we try to explain. We might refer to these data as “gravitational effects.” In addition to previously established data or prototype effects, the present study adds cross-cultural constancy in emotion concepts and in narrative genres. Thus any treatment of prototypes – or any alternative explanations – will have to treat these “prototype effects” as well. In the appendix, I suggest some ways in which these particular constancies may be explained. In addition, my description of prototypes indicates that they cannot be generated in any simple way (for example, by some sort of straightforward averaging). Moreover, they are not fixed or univocal. For example, in the next chapter, I shall discuss how they vary with context. I have argued elsewhere that they are affected by a range of variables, including, for example, one’s mood – even to such an extent that normative elements may shift radically, from positive to negative or the reverse (see Culture 126–31; for obvious reasons, the point is particularly consequential with ethnic, racial, and related prototypes). The purposes for which we invoke a prototype have bearing here as well. Barsalou argues that ad hoc categories are generated due to purposes. We have aims (for example, to raise money in a garage sale) and we generate categories (for example, “items that are appropriate to sell in a garage sale”) in order to achieve those aims. I would argue that even our well-entrenched prototypes are sensitive to purposes. That sensitivity, too, allows for contextual variation, leading to contrast effects, weighting, and the incorporation of ideals depending on the degree to which a given term is normative. Whatever the cognitive architecture, an alogrithmic level account of prototypes should capture this complexity. One account of “prototype effects” that does not cohere well with the present discussion is that of Lakoff. With work, the two could probably be reconciled. However, I find Lakoff’s account implausible for other reasons. Lakoff sets out to explain prototype effects by reference to a broad range of structures. Thus, despite his repeated return to the notion of prototypes, he in effect disputes the category. But the structures he isolates do not seem to be of the same logical type, nor do they seem clearly distinct in the way that theoretical categories need to be distinct. He begins by isolating “radial structures” that have a central case and a gradient of less central/more peripheral cases. He then isolates a series of “metonymic models” in which “a subcategory . . . is used to stand for the entire category” (84). First, it is not clear

The Mind and Its Stories

vague about just what an exemplar is. There is an ambiguity between types and tokens” (98). Moreover, the term “exemplar” suggests a “good” or “definitive” case of a category. In part, to avoid this final problem, and the associations that have attached to “exemplar”

that this differentiates metonymic models from radial structures where the central cases would seem to have just this function. In any case, metonymic models include “social stereotypes,” “typical examples,” “ideals,” “paragons,” “generators,” “submodels,” and “salient examples.” First, “paragons,” and “salient examples” are instances. (Despite the name, “typical examples” are not.) Thus, they do not fall under the general prototype category to begin with. (For discussion of these, see the following note.) It is very difficult to see how social stereotypes differ from typical examples, except that the term “social stereotypes” implies that the examples are both mistaken and socially harmfull. It is certainly important to distinguish true from false prototypes. And it is important to discuss the social consequences of prototypes. However, it is not clear that a “typical example” picked up from seeing many birds, reading about birds, hearing people talk about birds, and so on, is cognitively different from a “social stereotype” picked up from seeing many representations of some ethnic minority on television, hearing people talk about that minority, and so on, especially when combined with contrast effects, the biasing effect of norms, and so forth. As to ideals, these appear to be a function of the degree to which the relevant category is normative. “Submodels” are a strange category that seems to include some typical examples, some ideals, and some other things that do not involve prototype effects at all. Lakoff’s example is body temperature. But this is an actual norm, and it involves a continuous gradient of numbers. Not everything that involves a norm and a gradient is a prototype effect. The same point seems to hold for generators. I may simply have missed the organizing principle here. However, it seems that these categories are partially overlapping and partially irrelevant to prototype effects. Lakoff is right to point to the varieties of prototype effect. However, these seem best treated through an account that recognizes the complexities and functions of prototype generation, as discussed previously. Finally, I should reemphasize that I see the lexicon as having a complex structure that is not reducible to prototypes (understood as distinct and stable representations, connectionist circuits, or whatever). Thus, the present account does not suffer from the problems associated with purely prototype-based accounts of concepts. For example, Fodor has argued that such accounts run aground on semantic compositionality. Specifically, they have no way of explaining such simple concepts as “pet fish” (see Fodor’s Chapter 5). This is not a problem for an account that includes schemas and exempla along with prototypes, and that recognizes schematic complexity involving defaults, alternatives, and so on. (A full account of a lexical entry would include ideals, common beliefs, and other elements.) Indeed, compositionality is itself highly complex, and involves its own prototype effects. It seems likely, therefore, that a complex understanding of lexical entries, including prototypes, provides the best way of accounting for compositionality. For example, Fodor is right that a goldfish is not a good case of either pet or fish, but it is the best case of a pet fish. Of course, this means that “pet fish” has an associated prototype. But that prototype is also related to the prototype for “pet” in that a goldfish is probably the best case

Lexical Processes in Literary Experience

through various theories, I have adopted the term “exemplum” (plural, “exempla”) for simple instances. (“Exemplar” may then be reserved for “good” instances, instances conforming to the prototype.)3 It is important to stress that all three substructures operate in cognition.4 We do not merely interpret or respond to the world and to other people in terms of abstract schemas. Indeed, we are more likely to understand and respond by reference to prototypes or particularly salient exampla (on the last, see, for example, Chapter 3 of Ross and Nisbett; on the use of instances and larger, schematic abstractions, see, for example, Klein, Loftus, Trafton, and Fuhrman). The cognitive role of exempla is particularly important to literary response for a number of reasons. Most significantly, it appears to be the case that exempla – which are regularly primed and accessed along with

of a pet from among the set of fish. In accounting for compositionality in this case we need to account not only for the fact that a pet trout would still count as a pet fish. We also need to account for the fact that the prototype for a pet fish is related to the prototype for pet and to the set of fish concepts, but not to the prototype for fish. Here we return to the issue of prototype effects. In this case, such effects suggest the value of a multilevel account of lexical entries, an account in which there is at least some sort of hierarchical structure corresponding to a distinction among exempla, prototypes, and schemas (again, however these are ultimately specified in terms of cognitive architecture). In this terminology, Lakoff’s “paragons” are exemplars of highly normative categories. His “metonymic model” of “salient examples” combines instances that contribute disproportionately to the formation of prototypes (that is, highly weighted exempla) with instances that are highly nonexemplary (that is, exempla that contradict the prototype) and override the prototype in our thought. Thus when people refuse to fly on a DC-10 after a crash (Lakoff’s example [89]), they are allowing the specific case to substitute for their broader prototype of “airplane.” Here too it seems that Lakoff’s categories are not well formulated. As with prototypes, however, Lakoff’s analysis does indicate the complexity of exempla and exemplars. This complexity should be explicable by reference to weighting, contrast effects, the incorporation of ideals depending on the degree to which terms are normative, utility, and perhaps other factors. There is some disagreement as to whether it is necessary to have three lexically distinct types of structure in order to account for the three distinct levels of structure. Thus some theorists try to account for the prototype level by reference to schemas or exempla (see Shanks). Of course, to be descriptively adequate, any such reduction would have to incorporate all the complexities outlined previously. I am skeptical of attempts at genuine eliminative reduction. Even if it turns out that prototypes may be understood as relatively stable networks of exempla, then those networks themselves still form the intermediate of three levels. In other words, they are the prototypes, and they are not identical with any individual exemplum.

The Mind and Its Stories

schemas and prototypes – often have considerable affective force. Clearly, many such exempla (for example, the memory of a close relative’s death) cause one to feel strong emotion when one recollects them consciously.

cognition, memory, and rasadhvani This, then, gives us a way of translating and redeveloping Abhinavagupta’s notion that the affective component of memory traces is the source of rasa. Specifically, the dhvani of a text may now be understood as the schemas, prototypes, and exempla primed or placed in a buffer between long-term memory and rehearsal memory. The exempla include not only representational content, but affective force. As Schacter puts it, “we do not store judgment-free snapshots of our past experiences but rather hold on to the meaning, sense, and emotions these experiences provided us” (5). When an exemplum is sustained in the buffer, its affective force should lead to precisely the sorts of effect hypothesized by Abhinavagupta when he explained rasa in terms of memory traces. Specifically, we have every reason to expect that the affective force of an exemplum would bleed into consciousness without our being aware of its associated representational content, which is to say, the perceptual or propositional aspect of the exemplum. Or, rather, we have every reason to expect this when a set of affectively and representationally related exempla (for example, sorrowful exempla of love in separation) are maintained in the buffer through repeated priming due to the patterned dhvani of a text. (It is important that the exempla are related both affectively and representationally because the representation, so to speak, gives definition to the affect; for example, sorrow over a relative’s death is not precisely the same as sorrow over personal failure.) Schacter puts the general point simply, “someone might experience emotions and affects that result from incidents that are not recollected explicitly,” such that “implicit effects of past experiences . . . shape our emotional reactions” to new situations (232–3), presumably including literary situations. The repeated priming to which we have referred a number of times is the consequence of two factors. First, it is one direct result of the maximization of relevance or patterning. We not only assume that

Lexical Processes in Literary Experience

a literary work is maximally relevant, authors make works that are maximally relevant. In doing this, they shape not only sound, but suggestion and feeling as well. In other words, one thing an author does is work through the precise phrasings, the exact formulation of events, in order to keep the rasadhvani consistent, thus in order to prime semantically and emotionally relevant memories. Note, by the way, that an author does not need to have any idea whatsoever that he/she is doing this. All that is necessary is that the author read his/her own work, evaluating its emotive impact, and revising in order to produce the desired impact. Being a person like his/her readers, the author will experience some comparable priming of memories, and so on, whether he/she knows it or not. But this rasadhvani effect is not solely the result of authorial efforts. It is also the result of readers’ cognitive propensities. To understand how this occurs, we might consider the discussions of social judgment by Bower, Forgas, and others that present a virtual mirror image of the analysis we have just set out. Specifically, rather than primed memories leading to emotion, they discuss ways in which emotion serves to prime memories – and thus to sustain the initial emotion. The general point is well known. For example, Oatley explains that “There is now substantial empirical evidence to indicate that when happy, happy memories come to mind, and when sad, sad memories come to mind” (Best Laid Schemes 201). But the “algorithmic level” accounts by Bower and Forgas – the detailed, explicit, step-by-step accounts of cognitive processing – fit the present analysis much more specifically (on the different levels of cognitive scientific analysis, see Dawson 9; on the algorithmic level in particular, see his Chapter 5). Bower argues that the “Activation of an emotion node,” which is to say, the activation of some lexical cluster linked by an emotion (for example, a cluster linked via romantic love, fear, or anger), “spreads activation throughout the memory structures to which it is connected, creating a subthreshold excitation at those event nodes” (“Mood” 135). Put differently, when a particular “emotion is aroused . . . activation will spread out along its connections, thus priming and bringing into readiness . . . associated ideas and memories” (“Affect” 389). This not only produces priming of relevant memories, but leads us to “focus on affect-consistent rather than -inconsistent information” (Forgas, “Affect” 244), thus further

The Mind and Its Stories

perpetuating the feeling. Bower discusses the ways in which particular emotions lead to selective attention and other mechanisms that tend to be mood-reenforcing as they lead us to notice and remember particular aspects of an event or object (“Mood” 142–3). People “selectively attend to and learn more about stimulus material that is congruent with their feeling” (Bower, “Affect” 387). Though he disputes the sufficiency of an account based on spreading activation in this case, Mathews discusses “biases in emotional processing” in just these terms as well. Specifically, “The nature of these processing biases strongly suggests the potential for a circular relationship between cognition and mood. Anxious or depressed mood acts so as to give processing priority to the type of information that is most likely to enhance or maintain that mood state” (300); moreover, “Selective processing is specific to material that matches the content of the individual’s current concerns” (303). Though inverted, Bower’s treatment of “subthreshold excitation” of memories is precisely the process we have been describing. Moreover, Bower’s, Forgas’, and Mathews’ work indicates that, once the initial memories and emotions are triggered, the emotions will be to a degree self-sustaining. This is because any given emotion will continue to prime relevant memories and to focus the reader’s attention on textual details that themselves sustain the emotion and prime those memories. Of course, this also means that it may become harder to return a reader to a desired emotion once he/she has gotten “off track.” Hence we find the emphasis of the Sanskrit writers on the consistency of rasadhvani, and on avoiding certain “contradictory” feelings, such as disgust in connection with romantic love. A similar idea has been put forth by the philosopher and aesthetician, No¨el Carroll. Carroll has argued that fictional narratives are “prefocused” so that “[c]ertain features or situations and characters will be made salient through description or depiction.” The important thing about these features, and so on, is that they govern “the identity of [our] emotional states” (202). This becomes self-perpetuating as our emotional states themselves lead us to focus on particular features that are relevant to those states, as the cognitive research attests. Fear provides a commonsensical illustration. The emphasis and consistency of the writer will draw our attention to aspects of an initial situation that make us fear for the protagonist. Then, fearing

Lexical Processes in Literary Experience

for the protagonist, we focus our attention on just those aspects of the developing situation that are relevant to our fear. Carroll’s discussion of works as “prefocused” suggests that there are many ways in which our initial attention may be directed, and thus many ways in which even our initial emotional response may be guided. For example, much of Carroll’s work is on the horror film. It seems clear that, when someone goes to a horror film, he/she looks for certain things right from the beginning. This affects his/her emotional response. More generally, genre often “prefocuses” our attention, understanding, and response – including the priming of memories and correlated generation of rasa. It would take us away from the concerns of the present chapter to discuss this in detail. However, in the following chapters, I shall argue that there are two primary prototypical narrative structures, cross-culturally – romantic and heroic tragi-comedy. These genres are remarkably consistent in their elaboration of narrative goals and in their development of narrative events. As such, they facilitate the consistent development of rasa, both for the author and for the audience or reader. First, the idea of the genre, then its unfolding structure helps to orient the priming of a reader’s memories, and thus helps to produce a coherent emotive effect. Indeed, as we shall see, these genres are based on prototypical emotion scenarios. Research by Conway and Bekerian indicates that just such scenarios operate to prime or activate personal memories along with their associated feelings. Though Conway’s and Bekerian’s research was done outside a literary context, it seems directly applicable to literary study. First, Conway and Bekerian emphasize the importance of “situational determinants” to our understanding of emotions. Though Conway and Bekerian do not conceive of them in these terms, these situational determinants are, in effect, mininarratives. For example, one situational determinant of happiness, derived from this research, is reunion with a loved one (189). This implies an entire story of falling in love and being separated – and simultaneously suggests how “situational determinants” are just the sort of broad, emotion prototypes that guide the construction of literary narrative, as I shall argue in the next chapter. In any case, the general relation between narratives and situations such as this should be clear. Conway and Bekerian tested subjects to discover just how

The Mind and Its Stories

they came to imagine such situations concretely – something that we do routinely in reading literature, as, for example, Roman Ingarden emphasized (see, for example, his discussion of “schematized aspects” in Chapters 8 and 9). They found that, the vast majority of the time, people concretized these situations by reference to their own “personal experiences.” Thus, “specific autobiographical memories were cued” by “rather general . . . situations” (178), a point that should hold equally for literary narratives that are based on precisely the same sorts of situation, such as reunion with a loved one. The point also fits well with research of Shaver, Schwartz, Kirson, and O’Connor who found that personal memories of emotion episodes tended to be much the same as “typical” emotion episodes, a result that may be explained by the idea that “prototypes guide memory . . . of personal experiences” (181). Even more significantly, in the Conway and Bekerian study, the memories had direct emotional effects on the test subjects, just as our analysis (following Abhinavagupta) would predict. As Conway and Bekerian summarize, “Mood state appeared to shift” in keeping with “personal memories of experiences which mapped onto the specified target situation” (179). There is also direct empirical research on literary response that supports this analysis. Specifically, Uffe Seilman, Steen Larsen, Lászlo´ ´ and others have conducted experiments deHalász, János Lászlo, signed to determine the degree to which literary and nonliterary texts spontaneously elicit personal memories from readers (see the ´ and articles by Halász; Seilman and Larsen; and Larsen, Lászlo, 5 Seilman). They have found that literary works trigger memories of “more personally experienced, and therefore self-relevant, events” ´ and Seilman, “Across” 102). These memories appear (Larsen, Lászlo, to govern a sense of “personal resonance to and engagement in” a literary work (102). Oatley and his colleagues report further empirical work that bears out these conclusions as well (see “Emotions” and citations). Oatley cites experiments going back to the 1930s indicating that “concepts” of “the suggestive (dhvani) . . . and sentiments 5

There is also a long history of less experimentally formalized work that suggests similar conclusions. Norman Holland’s writings are a particularly good case of this sort.

Lexical Processes in Literary Experience

(rasas)” should be “of great importance for understanding the impact of narrative.” Among the studies cited by Oatley is research by F. C. Bartlett that examined the reaction of men and women to a story about war. He found that the readers’ anxieties about or memories of separation from their loved ones during war had a significant impact on their response to the story: “Most of the male subjects had been in the war or faced the prospect of going. For the women, losing relatives and friends was an ever-present threat.” In consequence, they responded most strongly to this aspect of the story. In research done sixty years later, Elise Axelrad “found an effect comparable to that of Bartlett. She had people record autobiographical memories that surfaced as they read [James Joyce’s short story] ‘Clay,’ and found that pieces of these memories became part of what they retold when they reproduced the story” (Oatley, “Emotions” 52). In keeping with this, Hal´asz explains that, according to his research, literary texts affect “the emotional memory network of the reader,” activating (or priming) “relevant events and sources which are embedded in an autobiographical context” (“Effect” 83). Hal´asz specifically links this with the development of empathy for the protagonist (83 – recall that rasa may be understood as an empathic version of an emotion). The general idea is also consistent with research in children’s responses to stories, research indicating that children often use stories to deal much more directly with emotional concerns drawn directly from their own lives. Miller, Hoogstra, Mintz, Fung, and Williams point out that “a child might appropriate and use for his or her own purposes someone else’s experience, someone else’s story. Framed in this way, any story has the potential to be a personalized story” (91). These researchers go on to discuss one young child’s rather remarkable use of the Peter Rabbit story. Kurt, a two-year old, hears the Peter Rabbit story and becomes fascinated with it. He listens to the story repeatedly. He then retells the story, such that “real-life events that Kurt had experienced . . . in the company of his mother and grandmother . . . are attributed to Peter Rabbit and his mother” (103). The fact that Kurt integrates his own memories into his retellings of the story – his explicit “personalization” of these stories, as Miller et al., put it – suggests that memories played a part in his enthusiastic response to the story initially.

The Mind and Its Stories

The theory of aesthetic response presented above suggests that aesthetically successful works, while not entirely “personalized,” do have a strong component of personalization. The personalized stories discussed by Miller et al., involve the child self-consciously taking over the story, integrating it with his/her own life. Aesthetic experience involves in a sense the first stage of this process – the priming of congruent or semicongruent experiences of just the sort a child might self-consciously integrate into the story. We could see aesthetic response, then, as a development out of the childhood tendency to appropriate a work entirely to one’s own autobiographical concerns. The point is reenforced by another aspect of Kurt’s relation to the story. As Miller et al., note, Kurt did not respond to the story simply because of its broad thematic relevance to his developmental stage. Rather, he responded to the story because it was linked with his own experiences and memories in direct and detailed ways – for example, “the fact that the story is set in a garden.” Kurt “had spent many enjoyable hours in his grandparents’ gardens” and only a month before hearing the story for the first time “his maternal grandparents had told him about the mother rabbit who had a nest of babies in their garden.” Moreover, some of Kurt’s earliest personal narratives, prior to the story, treated this garden (96). Thus, despite Kurt’s young age, the story related to a complex and emotionally significant set of personal memories. Miller et al., state that these experiences “prime his interest in Peter Rabbit’s adventures” (96). They are speaking rather loosely here. More technically, their analysis indicates that the story primed the memories, leading to the interest, as he listened to the story on the first reading, and subsequently. Again, the obvious difference between adult aesthetic response and Kurt’s retellings is that Kurt accessed the memories and integrated them into his retellings. But perhaps this is not a difference at all. Perhaps Kurt did not do this. Perhaps Kurt’s retellings integrated the autobiographical material simply because, being primed, it impacted his elaboration of details of the story, his “play” with the story. Axelrad’s research indicates that this happens with adult responders – as one might expect from our general account. Indeed, it seems likely that this sort of thing occurs with authors all the time. It seems likely that authors incorporate autobiographical material even in entirely nonautobiographical works (for example, in filling

Lexical Processes in Literary Experience

out characters in historical novels) through just such a process. Certainly, biographical criticism has given us many examples of this sort. In this way, Kurt’s retellings support not only our account of aesthetic response, but a parallel account of artistic creation.

objects, causes, and sources of emotion The mention of Kurt’s revision of the Peter Rabbit story in terms of his own life brings up a possible misunderstanding of the present analysis. One might argue that, in responding to the story, Kurt is actually responding to people, places, and situations in his own experience. Whether or not this is happening with Kurt, it is important to emphasize that this is not, in my view, what happens in adult literary response. That would be akin to what Levinson calls “the shadow object proposal,” that “objects of response” in fiction are not characters, but “real individuals or phenomena from the subject’s life experience, ones resembling the persons or events of the fiction, and of which the fiction puts the subject covertly or indirectly in mind” (23). To say that our memories are primed and thus their affect enters into our response to a fictional work is not to say that the people in those memories (including ourselves) are the objects of our emotional response. In relation to this, we might clarify the present analysis by drawing a simple and common distinction between the object of an emotion, the causes of an emotion, and the source or sources of an emotion. The object of an emotion is always some intentional object. It is some person, place, or thing as we understand or imagine it. Our relation to the world is always a relation to the world as we see it. My response to Smith is not a response to Smith as he really is, but as I imagine him to be. In other words, I have an idea of Smith, an “intentional object,” that may or may not be accurate. I may trust Smith; thus, my “intentional Smith” is trustworthy, even though the actual Smith may be duplicitous. The situation is much the same with respect to fictional characters. I have an intentional Hamlet or Leopold Bloom. Moreover, I build these up in just the same way that I build up intentional versions of real people, imputing motives and broad character traits on the basis of the person’s/character’s actions, statements, and so on. It is considered a great paradox that a reader might care about Hamlet in the way he/she cares about Smith. (Indeed, this is Levinson’s

The Mind and Its Stories

concern.) However, this does not seem to me to be much of an issue. Yes, I know that Hamlet is not real, but the process of constituting an intentional version of Hamlet is automatic or spontaneous. I do not plan it out. It is just part of the way our minds work. Once an intentional person is constituted, then he/she is open to the same sorts of emotive response as anyone else. This is because our emotive responses to persons are not based on a bare idea that they are or are not real. Rather, they are based on the salience, detail, particularity, and other aspects of the intentional object. This is why we may weep over Hamlet and yet remain indifferent to the lives of millions of real people. Our constitutions of those people are bare, while our constitution of Hamlet is highly elaborated. In any case, the important point here is that, when we weep for Hamlet, we do in fact weep for Hamlet (as we understand him, Hamlet in our intentional version). We do not weep for ourselves. This leads us to the issues of the cause and the source of emotional response. The cause of an emotional response is every condition that gives rise to that response. In the most obvious cases, an emotion directly concerns one’s own welfare. Thus, the cause involves some personal concern and some relevant situation. I am in danger, so I feel fear. But what about cases where someone else is in danger and I feel fear? Sometimes, for practical reasons or due to affection, one’s own welfare is bound up with that of the other person. We may refer to these indirectly egocentric cases as “interested.” But there are cases where there appears to be no egocentric involvement. What about these? In such cases, it seems that we assimilate the other person into a situation from our own experience that was in some way congruent. For example, seeing someone who is suffering from a particular illness, I may connect that person with my own experiences of illness, partially transferring the associated feelings. This is, of course, the empathic case, to which we have been referring. This brings us, finally, to the source of an emotion. In egocentric and interested cases, there is no source for the emotion outside the current situation. However, in empathic cases, the source of the emotion is external. The source itself is still egocentric or interested, even though the current situation does not have direct or indirect egocentric significance. In these cases, a preexisting emotion enters into the current situation. Specifically, in the empathic case, the source of

Lexical Processes in Literary Experience

our emotion is our memories of emotional experiences in egocentric or interested cases. For example, our experience of love comes into play when we empathize with a character in love; our experience of sorrow enters when the object of our empathic emotion is someone suffering grief. Note that this is true whether we are speaking of real life situations or stories. Our empathy with a real Romeo or Juliet is much the same as our empathy with the fictional Romeo or Juliet. Indeed, that is part of the value of the Sanskrit theory, as extended through recent work in cognition. It suggests not only how we respond to fiction, but also how we respond to life. (We shall return to this topic in Chapter 4.)

a brief illustration For the most part, my purpose in these analyses is not to produce a theory of interpretation, but to produce a theory of literature and emotion. In other words, my aim is not to present a theory that allows us to generate more interpretations of literary works. Rather, my aim is to isolate some of the recurrent features of literary structure, set out some universals of literary narrative, and articule at least a preliminary explanatory account of these universals in relation to emotion. Nonetheless, the theory just presented does have some implications for the discussion of particular literary works. I shall conclude by illustrating the point briefly. In “Emotions and the Story Worlds of Fiction,” Keith Oatley considers two lines from Hamlet. The lines are spoken by Horatio just after Hamlet dies (5.2.385–6): Now cracks a noble heart. Good night, sweet prince, And flights of angels sing thee to they rest!

Oatley writes, “Commentators have remarked that the second of these lines is the most beautiful in the whole of Shakespeare.” Not everyone would go that far. Nonetheless, the lines are striking, and lend themselves to analysis in terms of the principles presented in the preceding sections of this chapter. A cognitive priming analysis of rasadhvani should lead us to examine emotive/situational associations suggested by the lines, both general or common associations and individual memories. I shall focus on the former and spare the reader my personal memories. However, I believe it will be clear how

The Mind and Its Stories

such memories arise for readers in a structured, nonrandom way and how they bear on individual response. Obviously, there is sadness in the entire scene. Hamlet has just died. He died when still young, and he died unnecessarily. Certainly part of what we feel has to do with the situation itself – a situation that one could easily analyze in terms of suggestion, priming of memories, and so on. But obviously the effect of the lines is more than the effect of the situation. Whether or not one would label them “best in Shakespeare,” they are certainly more moving than obvious alternatives – “Hamlet is dead,” “The state has suffered a great loss today,” “No, no!,” “Don’t leave us, Hamlet!,” and so on. But why are they more moving? One of the most striking things about these lines, and certainly one main reason for their effect, is that they implicitly present Hamlet to us as a child. There is particular pathos in the death of a child. First of all, in a sense, everyone who dies is like a child dying. If I know someone for thirty years and he/she passes away, I see not only the adult, but also the child I first met. Though Hamlet is not a child now, he was a child – as we are reminded in his reminiscences about Yorick (“He hath borne me on his back a thousand times” [V.i.179–80]). His death is the death of that child, especially as it is a premature death. Moreover, the sorrow of Horatio is presented almost as if it were sorrow for a child. Indeed, that sorrow is what makes the lines moving, at least for me; it is what triggers the most significant memories and associated feelings. In this respect, the crucial line is “Good night, sweet prince.” Though Hamlet is, of course, literally a prince, no matter what his age, the term itself suggests youth. It recalls the fact of his youth, of his infancy even, for he was a prince from the moment he was born. Moreover, the term “prince” is something like “seed” or “sapling.” It implies a further condition that is its culmination – kingship. When a “prince” dies, that person has failed to reach the culmination his very title implies. This impression is only increased by the adjective “sweet.” Though clearly used more broadly in the Renaissance than currently, it nonetheless seems a term that should apply most readily to children – or to one’s beloved. “Sweet prince,” used as a form of address to the nobleman in line to inherit the kingship, shows the affection that one could easily show to a child, but which would be more difficult – not

Lexical Processes in Literary Experience

impossible, but more difficult – to show to an adult of that social stature. Conversely, if it was easy for Horatio to show Hamlet that degree of affection, it suggests that Hamlet retained a youthful openness and a sort of friendly warmth characteristic of children. “Good night” furthers the point. Of course, we say “good night” to adults. But saying good night to children is particularly important. The moments before they sleep are the moments when they can be readily frightened in the dark, fearing goblins under the bed and monsters in the closet. Horatio beside the prostrate body of Hamlet is like an adult at the bedside of a boy. He wishes him good night and tells him that there will be no goblins and ghosts, but rather “flights of angels.” He is like a father, reassuring his son. I hardly need to say in this context what personal memories might be primed for me or how very similar personal memories might be primed in others, for the general experience is common, as are the associated feelings. But, there is a problem here, for Hamlet is not a boy going to sleep. He has died and cannot hear Horatio’s assurances. Is Horatio simply denying the fact here, as a bereaved father might deny the death of his son, looking at his corpse and imagining that he only sleeps? The ending of the second line contains both an acceptance of Hamlet’s death and a sort of denial as well. Most obviously, Horatio adopts the common euphemism for death, “rest” – a euphemism that itself both asserts and occludes the death. More significantly, this word was part of the last sentence spoken by Hamlet. Just before the death and this lament, Hamlet spoke his famous dying words, “the rest is silence” (5.2.384; I am grateful to Lalita Pandit for drawing my attention to this line). It is as if Horatio is troubled by that word, that phrase. Hamlet says that his future is an end to speech. The line implies despair. But Horatio changes “the rest,” the nothingness that Hamlet suggests will follow death, into “thy rest,” a peace. More importantly, he transforms the silence of Hamlet’s death into something even greater and more beautiful than speech, the song of angels. It is all both acceptance and denial – something anyone experiences who has lost someone they love. This may seem excessive on Horatio’s part. He was, certainly, Hamlet’s close friend. But would he really have responded to Hamlet’s death in this way, with this degree of denial and anguish? There are two points to make here. First, it does not matter. The

The Mind and Its Stories

preceding analysis concerns the way a passage suggests ideas, scenes, meanings, and associations, that would foster the priming of tender and sorrowful memories, leading to feelings of tenderness and sorrow over Hamlet’s death. It does not matter whether Horatio, the particular vehicle for these feelings at this particular moment, realistically might have all these emotions. In fact, in some ways the phrasing itself is understated (contrast “Oh, Hamlet! Do not leave us!”). It is only through an elaboration of rasa and dhvani that one comes to see the emotion these lines involve and thus the source of their power. The second thing to remark is that this excess itself reminds a critic that the source of these lines is no doubt Shakespeare’s own experiences. It is well known that the character of Hamlet bears in some way on Shakespeare’s son, Hamnet, who died, at the age of eleven, not long before Shakespeare wrote this play (Wright and LaMar xxx). In treating Hamlet’s death, Shakespeare no doubt drew directly on the death of his own son. In other words, his own composition of these lines was animated by the same associations and the same sorts of personal memory that make the passage powerful for a reader. The writing of the lines was, in short, a process of precisely the sort we have been discussing with respect to the reading or hearing of the lines – though, like Kurt’s retellings of Peter Rabbit, done in the opposite direction. In this way, the passage is exceptionally appropriate for our purposes, for it indicates again that the preceding analysis applies not only to the emotional response of readers, but to the emotional expression of authors as well, and in precisely the same way. In following through the development of plot and character, the author too feels emotions, due to the priming of his/her own memories. These memories help to guide the author’s selection of further details, his/her specification of conditions, characters, and events, in just the way that they guide a reader’s concrete imagination of conditions, characters, and events.

conclusion In short, both the production and reception of literature appear to be inseparable from rasadhvani, understood as the patterned, cumulative priming of personal memories. These personal memories are

Lexical Processes in Literary Experience

representationally congruent with the literary situations developed in the course of the work and are sometimes cued initially by genre expectations. The emotional components of these memories serve as the primary source for our empathic emotional response to literary situations. On the other hand, the situations – or the characters in those situations – remain the (intentional) object of our emotional response. Finally, the emotions are to a degree self-sustaining, as they themselves further prime relevant memories and guide our attention and interpretation while reading (or, in the case of the author, while writing and rereading).

3 Four Hypotheses on Emotion and Narrative

deﬁning emotions and narrative Keith Oatley and Philip Johnson-Laird have argued persuasively that emotion is the product of an agent’s evaluations of his/her success or failure in achieving particular goals within what is, in effect, a narrative structure. In connection with this, they argue that there are five “basic” emotions: happiness, sadness, fear, anger, and disgust. In part, Oatley and Johnson-Laird are drawing on the research of Paul Ekman, who has demonstrated that the facial expressions for these emotions are universal, as are the expressions for surprise and “interest.” Ekman has maintained that all seven represent basic emotions. Oatley and Johnson-Laird delete surprise and interest from Ekman’s list on the grounds that they are not genuine emotions. Surprise appears to be a mode of experiencing other emotions (Oatley 60). For example, surprise is what makes fear into fright; in this case, it is a mode of experiencing fear. Anyone who approaches the theories of Ekman, Oatley, and Johnson-Laird with a knowledge of classical Sanskrit dramatic theory is likely to be struck by the remarkable similarity between these modern developments of biological and cognitive science, on the one hand, and the ancient Indic ideas about narrative and emotion, presented in the N¯a.tya´sa¯ stra and elsewhere over two millennia earlier. The N¯a.tya´sa¯ stra distinguishes eight “permanently dominant” emotions or bh¯avas. The list includes mirth, sorrow, fear, anger, and 76

Four Hypotheses on Emotion and Narrative

disgust, the five basic emotions of Oatley/Johnson-Laird. It also includes astonishment or wonder. This is roughly equivalent to “surprise,” from Ekman’s list, but seems less open to Oatley’s objections. Wonder is not simply what one feels when something unexpected happens – something frightening, for example. Rather, it is what one feels when something almost inconceivably great happens. It is, for example, what one experiences in witnessing a miracle, or what one takes to be a miracle. In other words, perhaps the relevant facial expression is ambiguous between surprise, which is indeed modal, and wonder, which is, rather, an emotion, and Ekman merely chose the wrong alternative. I stress the universal facial expressions here in part because they were themselves part of the Sanskrit theory. Bharatamuni, to whom the N¯a.tya´sa¯ stra is attributed, develops the theory of emotion in relation to a set of stereotyped facial expressions. Learning these expressions was an important part of the actor’s training in classical Sanskrit drama, and continues to be a part of an actor’s training in regional theater. Unsurprisingly, the set of plates in Ekman, illustrating the universal facial expressions, bears striking similarity to the expressions learned by actors in Indian dance and theater traditions continuous with the N¯a.tya´sa¯ stra (see, for example, the photographs following page 96 in Singha and Massey). Moreover, though Oatley relies on Aristotle for his understanding of plot, Sanskrit narrative theory would have been far more relevant. Aristotle, despite his well-known references to hamartia (error or flaw), tends to treat plot in terms of a relatively agent-neutral causal structure, discussing “necessary or probable” sequences of events linking beginning, middle, and end. In contrast, the Sanskrit theory is entirely agent-centered. Indeed, Oatley ends up modifying the theory of the Poetics until it bears less similarity to its original than to the theory of the N¯a.tya´sa¯ stra. More exactly, the Sanskrit writers relate their eight emotions to a narrative theory that is explicitly structured around an agent’s pursuit of goals and his/her periodic reevaluation of likely success or failure in achieving those goals. This reevaluation is formalized in Sanskrit narrative theory through the concept of the “pause” or “deliberation,” a critical element of narrative structure, and one that involves “anger or passion or temptation” (Dhanam . jaya 24). The standard translations refer to

The Mind and Its Stories

this as a “juncture,” which is, coincidentally, just what Oatley terms the moments of evaluation that give rise to emotion (see Oatley 25, 36). More exactly, like Aristotle, the Sanskrit theorists distinguish three parts to a narrative. The Sanskrit writers have relatively little to say about the beginning (¯arambha, which implies the commencement of an undertaking [see Monier-Williams]) and conclusion – also termed “culmination” or “attainment of the object.” However, they do analyze the “middle” – the “development,” “expansion,” or “progression” as they call it – in some detail. If we follow Abhinavagupta (Locana 438) and combine elements from two different parts of this analysis, we may divide the middle “progression” into three sections: 1) the hero’s initial effort; 2) his/her “pause” to evaluate the prospects for success; and 3) the renewal of effort, with full concentration. In Sanskrit theory, which is designed for comedy, the third stage also carries the implication that the hero is fully confident that he/she will succeed. Hence, this part is usually translated “certainty of success.” However, the Sanskrit term, niyatapti, has a meaning related to constant or steady concentration involving the control of passion. This is the element I have emphasized in defining the parts of the progression. I have dropped the assumption of success in order to allow application outside of comedy. (I should say that this interpretation is far from uncontroversial. For a different view, see Ingalls 441n.) In any case, the relevance of this to Oatley’s and Johnson-Laird’s views should be clear. Oatley at one point remarks that the sort of theory he is developing “is in a sense, familiar to the European or American mind,” a mind accustomed to the “idea of rational planning” (29). Oatley is probably just being cautious here. Afraid that someone will accuse him of imperialistic universalism, he confines his claim to Europe and America. But in fact his theory is much closer to Indian theories than to Euro-American theories. This congruence serves to remind us that much of potentially great intellectual value is lost when we remain ignorant of other intellectual traditions, or assume that our own ideas are necessarily more in keeping with those of “our own” tradition, and necessarily incommensurable with those of “alien” traditions. Ekman’s theory of basic emotions would have been advanced and clarified by incorporation of the Sanskrit views. Much of Oatley’s preliminary analysis regarding narrative and emotion would have been

Four Hypotheses on Emotion and Narrative

unnecessary, and his work could have begun at a more advanced point.1 More importantly, however, this congruence adds weight to Oatley’s theory, by showing that its application is not confined to Euro-American “rational planning.” Still, it is important to be clear about the precise nature of this support. For example, does the similarity between Oatley and the Sanskrit writers suggest that there are, in fact, basic emotions? This is not clear. The whole issue of basic emotions is highly controversial. It is clear that some emotion terms are basic within certain limited domains. For example, it is clear that “fear” is the basic term by reference to which we define, say, “terror” or “fright.” In a similar way, “cold” and “hot” are basic terms by reference to which we define, say, “scalding” or “frigid.” But it does not seem to make sense to say that coldness itself – not the term, but the property or the feeling – is basic, whereas scaldingness or frigidness is derivative. The same point applies to “fear” and “terror.” Further difficulties arise when Johnson-Laird and Oatley see jealousy and hatred as forms of disgust (114, 116), an analysis that I suspect most readers will find difficult to accept. (For a fuller discussion of problems with theories of basic emotions, see Ortony and Clore and citations. We shall return to the issue – or, rather, the related issue of the biological givens of emotion – in the Afterword.) Oatley and Johnson-Laird have made an argument regarding basic emotions that must be taken seriously, and they have clarified the concept in a way that few other theorists have managed to do. Though certainly worth pursuing as one alternative in a research program, however, the idea of basic emotions remains problematic. At the very least, we cannot simply assume it as a default hypothesis. 1

I should note that, once I pointed out the connection with Sanskrit writing, Oatley was delighted. He set himself the task of reading extensively in the field, and incorporating Sanskrit poetics into his own theoretical and interpretive work. Unfortunately, most western humanists do not appear to be so open minded. Many have developed such a thoroughgoing commitment to cultural difference that they are incapable of recognizing similarity. A good example of this is found in the introduction to Hjort’s and Laver’s fine collection, Emotion and the Arts. There, Hjort and Laver cite the “basic emotions” of rasa theory as “specific to a cultural community” (15) and as evidence of “radically different cultural self-understandings” (15). I should perhaps note that the emotions that do not overlap with Ekman’s universal facial expressions are far from exotic, as we shall discuss below. For example, one is romantic love.

The Mind and Its Stories

But, if the idea of basic emotions is problematic, what is one to make of the universality of some facial expressions and of the close correspondence between the theories of the Sanskrit writers and those of Ekman, Oatley, and Johnson-Laird? The first thing we need to do here is separate the issue of universality from that of basicness. The existence of a limited set of universal facial expressions of emotions certainly suggests something, but what? The fact that these emotions turn up in the Sanskrit theory indicates that they have some sort of special importance for aesthetics, but it seems rather unlikely that all societies developed the same facial expressions so that they could all have theater. What, then, are we left with, beyond the hypothesis of basic emotions? To answer this question, we need to concentrate for a moment not on the emotions, but on the facial expressions. The most obvious implication of the universality of such facial expressions is not that the correlated emotions are more basic than other emotions, but that these emotions are more directly involved with and germane to empathic communication. After all, that is what facial expressions do – communicate emotion. Moreover, this is consistent with other data as well, such as the finding of Shaver et al., that the putatively basic emotions are the most prototypical emotions. It would make sense for prototypicality to be bound up with interpersonal saliency. Finally, this view fits the Sanskrit case as well, for the communication of emotion is obviously central to facial expression in the theater – and to other, nonfacial aspects of literature.2 This leads us to a significant difference between the Sanskrit theory and that of Oatley and Johnson-Laird. The Sanskrit theory does not require that the emotions specified in the theory be basic. It requires only that they have some sort of special place in aesthetic feeling. Moreover, the reason for this place is suggested in the precise formulation of the theory. Specifically, Sanskrit theorists do not focus

Oatley and Johnson-Laird stress the communicative function of emotions. However, they are speaking of emotions themselves, not facial expressions, and their view is that all emotions have communication as their primary function. I have only urged that certain emotions may be more readily open to empathic communication – a very different thing. Indeed, it is not even clear that the use of the word “communication” is the same in the two contexts. For Oatley and Johnson-Laird, “communication” is simply the transmission of information, not only between persons, but within one person (see Oatley 44).

Four Hypotheses on Emotion and Narrative

on emotion per se or “bh¯ava,” as we noted in the preceding chapter. Rather, they focus on “rasa.” Once again, the term refers to an aesthetic feeling. That aesthetic feeling is, roughly, an empathic version of an emotion or bh¯ava. When the hero is faced with a dragon, we do not fear the dragon, experiencing the bh¯ava or emotion itself, and therefore running from the theater. Rather, we fear for the hero. We feel something clearly related to fear, but not identical with fear: We feel empathic fear, which is to say, the “terrible” rasa. In the preceding chapter, I simply used “emotion” to refer to “bh¯ava.” In this context, however, “bh¯ava” refers most importantly to egocentric feelings, and in that way it is narrower than “emotion.” “Rasa,” again, refers to empathic feelings. From this point on, I shall use “emotion” to refer to feelings of both sorts, reserving “ bh¯ava ” and “rasa” for the more specialized meanings. From this, we may conclude that, whatever we decide about basic emotions, the congruence of the Sanskrit theory with the cognitive theories of Oatley and others has to do with the particular empathic salience of certain emotions and the centrality of empathic communication to literary feeling. For the moment, I should like to turn away from the particular list of empathically salient emotions and focus on the more general issue of the empathic communication of emotion. This, in turn, will provide us with a way of reunderstanding the relation between emotion and narrative. At the same time, it will help us to rearticulate, and organize, the list of predominant empathic emotions or rasas.

empathy and emotion prototypes Empathy involves, among other things, a conscious or unconscious inference to what someone else is feeling. (It also involves an openness to putting oneself in the place of the other person – see Davis 16–17; I shall not be considering this aspect of empathy here.) When we identify an emotion, we access an internal lexical entry for the emotion in question – linking, for example, some set of perceptions with the entry for “fear,” “disgust,” or whatever. These entries include some sort of connection to the “phenomenological tone” of an emotion – what it is like to feel fear, disgust, and so on. However, this cannot be all there is to our lexical entry. Most importantly, this does

The Mind and Its Stories

not give us a way of identifying when someone else has an emotion, unless he/she tells us – for we have no access to his/her “raw feels.” Moreover, his/her statement that he/she is experiencing a certain emotion is of no value in this context, because if the only way to identify someone else’s emotions is through verbal communication, we have no way of learning emotion terms to begin with. For this Wittgensteinian reason, Anna Wierzbicka has argued that “emotions can be identified only by a reference to a standard situation” (“Emotions” 60) and thus that the meaning of emotion terms has the general form, “X feels emotion1 = X feels as one does when . . .” For example, “X feels sad = X feels as one does when one thinks that what one has desired to happen has not happened and will not happen” (61). Ronald de Sousa makes a similar point: “We are made familiar with the vocabulary of emotion by association with paradigm scenarios” (142). De Sousa goes on to distinguish two aspects of these scenarios – roughly, causes and effects. Along the same lines, Oatley distinguishes eliciting conditions (what leads one to feel an emotion) from expressions or action consequences (what one does when one feels the emotion). The Sanskrit theorists, once again, anticipated this, distinguishing “vibh¯avas” or determinants and “anubh¯avas” or consequents of bh¯avas. In keeping with these divisions, then, we might say that our lexical entry for any given emotion term includes not only some link to the feeling of the emotion (typically via memories, as discussed in the preceding chapter). It also includes some account of the kinds of situation that give rise to the emotion and some account of the kinds of expression and action that result from an emotion. Wierzbicka characterizes these accounts in fairly abstract terms. She is no doubt correct that we have, or are able to generate, abstract definitions of this sort. However, I doubt that our actual cognitive processes surrounding emotion – for example, our ordinary empathic responses – operate at such an abstract level. Rather, I suspect that, like most of our lexical entries, the entries for our emotion terms are more concrete. Empathic and other forms of inference, imagination, projection of emotion are, in other words, less a matter of subsuming instances under categories, than of comparing instances with more fully specified types. More exactly, a good deal of research suggests that our lexical entries are, most often, structured around prototype cases, not necessary

Four Hypotheses on Emotion and Narrative

and sufficient conditions (see Holland et al., 182ff. and citations). For example, our use of the word “bird” is more a function of comparison with a bird prototype or with exemplary instances of that prototype, such as robins, than of any strict definition – “warm-blooded, oviparous vertebrate with alar forelimbs,” or whatever. In identifying something as a bird, in discussing birds, and so on, we do not begin with an abstract category and judge whether or not instances fit the general conditions specified in that category, whether they fit the definition “warm-blooded vertebrate,” and so on. Rather, we base our inference, identification, or whatever, on similarity to prototypes.

ﬁrst hypothesis: the prototype basis of emotion terms This leads to the ﬁrst hypothesis I should like to put forth here: Emotion terms are prototype-based in both eliciting conditions and expressive/actional consequences. In other words, the “situations” identified by Wierzbicka and “scenarios” discussed by de Sousa should be understood as specific in the way that the prototype for “bird” is specific. Thus, when we judge someone to have a certain emotion, we do so by comparing his/her situation with prototypical situations and his/her response with prototypical responses. For example, our lexical entry for “sad” would not include a Wierzbicka-like abstract equation, but rather something along the following lines: “What you feel when someone you love dies and what you express through weeping.” (Or, rather, if it does include an abstract equation, that equation is subordinate to the prototype in our usual cognitive processes.) Here, it is valuable to return to the view shared by Oatley, JohnsonLaird, and the Sanskritists that emotions are embedded in stories. One way of putting the difference between Wierzbicka’s view and the prototype account just proposed is to say that the latter defines emotion concepts in terms of what are in effect mini-narratives, seeds of stories (to use one of the Sanskrit writers’ favorite metaphors). For example, suppose that the prototype – or one prototype – of the eliciting conditions for sorrow is indeed the death of a loved one. This is, implicitly, the outcome of a story, a story of personal affiliation between the sorrowful person and the person who has died, and the story of that person’s death. As the prototype becomes

The Mind and Its Stories

more specific, the implicit narrative becomes more fully developed as well. For example, suppose the prototype is “the premature death of a loved one.” This further concretization of the prototype entails a further concretization of the implied narrative. Readers familiar with the literature on emotion will know that, in ¨ influential work, George Lakoff and Zolt´an Kovecses have proposed a prototype-based account of emotion concepts as well. However, the ¨ Lakoff/Kovecses view of emotion prototypes is quite different from the view I shall be developing in the following pages. Before going on to discuss my own account in more detail, it is worth indicating what these differences are. ¨ Most obviously, Lakoff and Kovecses treat much broader sequences of events, and they treat these events much more abstractly. ¨ Kovecses explains that, in their view, the “prototypical cultural model of anger” is the following “five-stage” sequence: “cause of anger, existence of anger, attempt at control, loss of control, retribution” (173). “Cause of anger” involves five elements: “Wrongdoer offends self,” “Wrongdoer is at fault,” and so on (143). Though these clearly do not present necessary and sufficient conditions, they are far more abstract than my prototypes – so abstract that I would refer to them as “schemas” or, alternatively, “scripts.” (I shall return to this point in a moment.) For example, in my view, “wrongdoer offends self” is too broad. I am offended when someone cuts in front of me in line. But I would hardly consider this a prototypical eliciting condition for anger. Rather, in my usage, the prototype eliciting condition for anger – or one such prototype, depending on context – would be something more along the lines of “someone unexpectedly strikes me” or perhaps “someone unexpectedly tries to knock me off my feet.” ¨ More importantly, perhaps, the Lakoff/Kovecses five-stage sequence collapses different levels of response and analysis. For example, I do not believe that “Wrongdoer is at fault” is commonly a prototypical eliciting condition for anger. Rather, it is a necessary con¨ dition for the justiﬁcation of anger. Thus, the Lakoff/Kovecses category “Cause of anger” combines direct considerations of anger and meta-level considerations of our response to or evaluation of anger. Their next two categories operate at this meta-level as well. “Attempt at control” and “loss of control” point toward actional/expressive outcomes. However, the former bypasses those outcomes entirely

Four Hypotheses on Emotion and Narrative

for a meta-level response to anger. The latter includes quite general outcomes – such as “aggressive actions” (143) – but also incorporates the meta-level response to the motivating impulse of anger. The final category is somewhat odd as it overlaps with “loss of control,” repeating the actional outcome. In this case the outcome is a “retributive act against wrongdoer,” explained again in very general terms as “usually angry behavior” (144). However, this category does not lack its own meta-level addition, for an evaluative element enters, determining that “The intensity of retribution balances the intensity of offense” (144). It is no doubt the case that our ideas about emotions include not only the emotional sequence itself – eliciting conditions, phenomenological tone, and actional/expressive outcomes – but also (vague) principles of justification, (variable) norms bearing on the self-conscious evaluation and control of the emotional impulses, and so on. Indeed, we shall find many elements of just this sort in the development of literary narratives in relation to emotion prototypes. Nonetheless, I doubt that there are scripts of precisely the sort set out ¨ by Lakoff and Kovecses. Rather, I suspect that their findings are a residue of our separate ideas about the causes of emotions, the justifications of emotions, prudential or ethical responses to emotions (for example, it is normative that one try to control one’s anger), and so on. In any case, it should be clear that sequences of this sort are not “emotion prototypes” in my sense of the phrase. Even if we do have scripts of this kind with respect to different emotions, they are not what I am concerned with here.3

¨ This variance in usage of the term “prototype” is not confined to Lakoff and Kovecses. It indicates a much broader difficulty regarding terminology among writers on cognition. James Russell points out that some “prototype” accounts focus on “remembered individual, concrete experiences”; some treat “generalized schemata”; and some focus on “average” cases (39). In the preceding chapter, I distinguished schemas, prototypes, and exempla. But the quotation from Russell indicates that “schema” (Russell’s “generalized schemata”) and “exemplar” (Russell’s “individual, concrete experiences”) are often used as equivalent to “prototype.” In fact, for many writers, the only important conceptual division seems to be one between necessary and sufficient conditions, on the one hand, and everything else, on the other hand. It is undeniably important to distinguish between accounts based on necessary and sufficient conditions and accounts not based on necessary and sufficient conditions. Indeed, that distinction will figure prominently in the following pages. But that is not the only consequential distinction in this area, hence my more systematic usage.

The Mind and Its Stories

In short, the sort of emotion scenarios described by Lakoff and ¨ Kovecses, as well as those treated by James Russell and a number of other writers – including, for example, Wierzbicka – are schemas, in my sense, not prototypes. Consider, again, sorrow. I have suggested that the prototype for sorrow is, roughly, “what you feel like when someone you love dies and express through weeping.” This illustrates the point nicely. It is not abstract, but concrete. It includes the default values characteristic of schemas (for example, desiring something to happen and it not happening). But it specifies and further concretizes these with “average” properties and distinguishing characteristics. As elsewhere, the “averaging” here occurs over highly salient instances. The displays of sorrow that accompany death, along with the elaboration of mourning in funerary rituals, make deaths highly salient instances of sorrow. (The same point could be made about marriage and joy.) Deaths are also highly distinguishing instances of sorrow. Many things might make us feel sorrow. But death is distinctive in its almost invariable association with sorrow and in the way it intensifies sorrow-relevant properties (for example, irreversibility).

second hypothesis: the basis of prototypical stories in emotion prototypes As I have already emphasized, the concreteness of emotion prototypes is related to the fact that they are bound up with implicit stories. This leads us to our second concern, the nature of narrative. Here, I shall propose a more complex hypothesis. First of all, the research on prototypes suggests that we should understand narrative too in terms of prototypicality. Like birds, there are some narratives we consider to be “more standard” cases than others. For example, the story of young lovers overcoming obstacles to be united is a more “standard,” which is to say, prototypical narrative, than the story of how a furnace operates.4 There are no doubt many properties that tend 4

To get a sense of prototypicality, it is useful to have a nonprototypical example. Take any standard love story – canonical or ephemeral, moving or banal – and contrast the following story from Schank, which is highly nonprototypical: “Diane was trying to figure out why hairdressers won’t take credit cards. She thought that maybe they had a poor clientele but realized it was also true in Westport, a well-to-do area. She never found an answer” (150).

Four Hypotheses on Emotion and Narrative

to characterize prototypical narratives. For our purposes, the most important of these is emotional interest. In other words, one difference between prototypical narratives (for example, stories about lovers) and nonprototypical narratives (for example, stories about the operation of appliances) is that the former engage our feelings, or at least address and appeal to feelings.5 This part of the hypothesis seems fairly uncontroversial. But what does it mean to say that some narratives are prototypical and others are not? It means that our mental lexicons include prototypes for narratives and we judge narratives – including what is or is not a narrative – by reference to these prototypes. (Again, just how we understand the existence of these prototypes is not important for our present purposes.) But the relevance of prototypes to literature does not end there. We employ prototypes not only in identifying narratives, but in understanding and creating them. Moreover, we employ a wide range of prototypes – for genres (for example, epic or novel), characters (for example, romantic heroine or action hero), scenes (for example, bar scene or battle scene), and so on. Prototypes guide our ongoing interpretation of stories, as we hear or read them, and our ongoing generation of stories, as we speak or write them. The point is merely a generalization of the research on prototype-based thought, mentioned previously. Consider genre. When writing an epic or a novel, an author is almost certainly guided far more by prototypes of these genres than by necessary and sufficient conditions. He/she is guided by the standard or typical opening, character structure, and so on, for epic, not by some rigorous definition. Similarly, when readers are reading an epic, their expectations are guided by their prototypes for the genre. What they find surprising or innovative or faulty is not a result of an abstract definition. Put differently, the sorts of reader expectation isolated and examined by phenomenological 5

The research of Brewer and Lichtenstein clearly supports this contention, for they found a close correlation between emotional effect and significance of narratives, on the one hand, and judgments of storiness, on the other. However, Brewer and Lichtenstein were not operating with a prototype model. Thus, they did not formulate their study nor examine their findings in these terms. Moreover, they treated aspects of discourse order – another factor in prototypicality judgments – rather than story structure per se (that is, story structure in the narrow, technical sense). Again, discourse is an extremely important topic for understanding literature and emotion. However, it is a large and complex topic, requiring treatment in a separate volume.

The Mind and Its Stories

and response-oriented critics such as Wolfgang Iser and Hans Robert Jauss are, to a considerable degree, a function of narrative prototypes. Again, a wide range of prototypes enter into our generation of, or response to, narrative. If indeed one standard goal of literary storytelling is emotional appeal, then we would expect one crucial set of prototypes to address this emotive aspect of narrative. This leads to a complication of my second hypothesis. When the purpose of our storytelling is, in part, emotive – thus, in prototypical narratives – that storytelling will be bound up with emotion prototypes. These emotion prototypes will help guide our decisions as to what sort of story is tellable, what is of interest, what is valid, and what is effective and engaging. This is true whether the narrative in question is fictional, biographical, or historical; set in the form of an epic, a drama, or a novel. In each of these cases, due to the emotive purpose of the tale, emotion prototypes will provide central structural principles for the story, partially guiding its overall shape and outcome, its tone, and so on. To say that an emotion prototype guides storytelling is to say, according to the preceding analysis, that both the eliciting conditions and the actional consequences guide that storytelling. However, insofar as prototypical narrative involves a sustained, empathic appeal to readers or listeners, it almost necessarily involves a particular emphasis on eliciting conditions, themselves understood, once again, as the conclusions of implicit narratives. We may, of course, feel badly for a character who is weeping. But our response is given depth and intensity only through an understanding of what has led to this expression of sorrow. Indeed, our knowledge of the actual eliciting conditions may make weeping comic and laughter sorrowful. The point should become clearer in the next section, when we distinguish different types of emotion and their different functions in prototypical narratives. For now, however, it is important to emphasize the particularly consequential role of eliciting conditions in the generation of narratives. My second hypothesis, then, fully formulated, is the following: Prototypical narratives – including literary narratives – are generated largely from prototypes, prominently including the prototype eliciting conditions for emotions. Put differently, our prototypical stories are, in their broad structure, expansions of the micronarratives that define our emotion

Four Hypotheses on Emotion and Narrative

terms. Moreover, we experience them in that way, implicitly linking events in the story with their projected conclusions in prototype eliciting conditions for emotions – as when we experience Aristotelian fear while watching a sequence of events unfold that we expect will lead to prototypical conditions of sorrow for the hero or heroine. If I am correct in this hypothesis, it should immediately be clear that this coheres nicely with the empathic account of literary emotion given above. For the proximity of a literary narrative to prototype cases of eliciting conditions should facilitate the immediate, unselfconscious lexical identification of the emotion, which is a crucial part of empathic identification. More importantly, it is also likely to facilitate the priming of relevant personal memories, with their associated phenomenological tone (as discussed in Chapter 2). But is this to say that narratives are made from the prototype determinants of any and every emotion? Well, in fact, there are narratives for every emotion. For example, we often tell stories simply to communicate anger or disgust. Thus, if narratives are understood in terms of necessary and sufficient conditions, we have to say that no emotion is excluded from this narrative-generating use. However, this is not true for our prototype narratives. Our prototype narratives may include anger or disgust, but they are not defined by reference to, and thus via prototypes for, anger and disgust. In other words, while emotive narratives are more prototypical than nonemotive narratives, some emotive narratives are themselves more prototypical than others. This prototypicality is, in turn, bound up with the typological and structural relations among emotions.

typological and structural relations among emotions Consider again the emotions discussed by Oatley, Johnson-Laird, and the Sanskrit theorists. Whether we consider these to be basic, to be particularly open to empathic communication, or neither, it should be obvious almost immediately that there is a problem with treating these emotions as if they are all of the same type. In terms of the broader narrative structures discussed by all these theorists – in which narrative is goal-oriented action involving periodic evaluation of one’s likely success – we could say that we never evaluate

The Mind and Its Stories

happiness in terms of fear, anger, or disgust. Rather, we evaluate fear, anger, and disgust – or, more exactly, fear-, anger-, and disgustrelevant events – in relation to happiness and sorrow. For example, fear is, precisely, fear of what will lead to sorrow or block happiness. Happiness and sorrow are, in many ways, general markers for positive and negative feeling, with “positive” here meaning something like “the feeling we wish to achieve and sustain” and “negative” meaning something like, “the feeling we wish to end and avoid.” Their phenomenological tone is, roughly, the mental equivalent of what in bodily sensation is termed “pleasure” and “pain.” Moreover, anger, disgust, and fear are what might be called punctual emotions. They are prototypically elicited by temporally thin events – an attack by a wild animal, the smell of something putrescent – and are not (prototypically) enduring. Happiness and sorrow, in contrast, prototypically arise with respect to more temporally thick conditions. Though the death of a loved one is, in and of itself, momentary, it gives rise to an enduring condition, and that condition is what inspires the emotion. Moreover, they are themselves (prototypically) more lasting. In actual fact, the experience of fear or anger may be more lasting than the experience of happiness; but facts often have limited bearing on our prototypes (due to salience – including the salience of fictional instances – norms, and contrast effects, as discussed previously). For example, happiness may, in fact, last only a few moments. But our prototype for happiness is well-expressed in the formulaic ending for a romantic comedy: and they lived happily ever after. In connection with this, and in keeping with Oatley’s notion that emotions occur at moments of periodic evaluation, the moments of narrative “pause” in the Sanskrit scheme, we could distinguish between two types of emotion. On the one hand, there are emotions that have their most important function at specific points within the course of an encompassing narrative – whether the sort of literary narrative treated by the Indian writers, or the sort of life narrative treated by Oatley. These emotions are typically limited in their temporal scope, and prominently include fear, disgust, and anger. To this list, we might add wonder, as well as mirth, which is distinguishable from happiness, understood as an enduring state, and is probably

Four Hypotheses on Emotion and Narrative

closer to what is expressed in the universal facial gesture. This covers the emotions that have universal facial expressions and thus may include the most important or frequent emotions in contexts, such as literature, where empathy is of critical importance. In any case, I shall refer to these, and any other emotions of the same general type, as “junctural” emotions, for they typically define some sort of juncture, some interruption or pause in the encompassing narratives. More exactly, we might define a juncture as any implicit or explicit assessment of an agent’s position in an imagined trajectory of actions and events as they seem likely to play out relative to some goal.6 Note, then, that junctures are not objective features of a causal sequence, but responsive features, features of an agent’s own realization, an author’s reflection, or a reader’s concern and consideration. Of course, some points in life and in narrative lend themselves to this sort of evaluation. When someone is attacked by a pack of wolves, this is pretty likely to count as a junctural moment for him/her, and for anyone reading about it. Moreover, Labov notes that points of particular intensity in a story, such as the beginning of a fight, often involve a suspension of the action and even an expression of the protagonist’s own assessment of the situation (374). Nonetheless, in principle, such reflection can occur at any moment. In contrast with junctural emotions, I shall refer to happiness and sorrow as “outcome” emotions, for they define the enduring feelings we prototypically consider the final evaluation points for junctural emotions. Thus, we find the prototypical endings of comedy and tragedy – the former, in which everyone lives “happily ever after”; the latter, which carries the implication of a sorrow that is irreversible (for example, due to a death). Of course, happiness and sorrow do operate as junctural emotions as well as outcome emotions. It is particularly frequent for sorrow to operate as a junctural emotion in narratives with an outcome emotion of happiness. But their most important and distinctive role in narratives is as outcome emotions. In actual life, one might argue, happiness is never an outcome emotion, but only 6

In connection with the preceding discussion of emotion and memory, it is worth remarking that, according to Schank, remindings occur with particular force and importance at points of expectation failure. Junctural moments are often just such points – hence points when memories, and their associated affects, are likely to be particularly pressing.

The Mind and Its Stories

a junctural emotion. In other words, in life, there is no such thing as “happily ever after.” However, our prototype of happiness is, again, enduring. Moreover, this is not at all confined to fiction; it is part of our beliefs about real life as well. For example, one purpose of such ideas as Heaven is to allow us to imagine enduring happiness as an encompassing goal of real life. This distinction between junctural and outcome emotions allows me to clarify my second hypothesis. Insofar as an emotion operates as an outcome emotion in a narrative, its eliciting conditions are most crucial for the generation and reception of that narrative. It is precisely these eliciting conditions that define the narrative conclusion. Indeed, insofar as narratives recount pursuit of a goal – and prototypical narratives do just this – they are necessarily organized by reference to the eliciting conditions for one outcome emotion, happiness, since happiness is the aim of action. Moreover, we do not pursue undifferentiated happiness. Our goals in pursuing happiness are necessarily far more specific. The specificity of the goals, the specificity of happiness is of course the specificity of eliciting conditions. To desire happiness is to desire a particular set of circ*mstances that one takes to cause happiness. These too are, most often, prototype-based. In other words, when we imagine happiness, we tend to imagine prototypical eliciting conditions for happiness (that imaginative inclination is, after all, part of what makes such eliciting conditions prototypical). And, of course, the goals we pursue (expect characters to pursue, and so on) are just the ones we imagine will produce happiness – thus, again, generally the prototypical ones. Indeed, the account of emotions as defined by prototypes, rather than by necessary and sufficient conditions, is perhaps nowhere more obviously consequential than with respect to “happiness” and “sorrow,” for these emotion terms are almost vacuous when defined by necessary and sufficient conditions. As Wierzbicka puts it, “X feels joyful = X feels as one does when one thinks that something has happened that one desires to happen” and “X feels sad = X feels as one does when one thinks that what one has desired to happen has not happened and will not happen” (“Emotions” 61). These are, again, almost bare positivity and negativity. But clearly happiness and sorrow mean something much more than that in our experience, inference, action, and so on.

Four Hypotheses on Emotion and Narrative

This leads us to the question of what these prototype eliciting conditions might be. In order to think about this, I should like to return once more to Sanskrit aesthetic theory. Consider again the list of bh¯avas. Most of the bh¯avas overlap with the emotion lists of Ekman, Oatley, and Johnson-Laird, and we have already included the shared items – fear, anger, and so on – in one of our two emotion categories. However, there are two bh¯avas on the Sanskrit list that have no correlates in Oatley et al., and which thus far have no place in our anatomy of emotion types. These are “rati” and “uts¯aha.” Rati is sexual love, but is prototypically represented as romantic love. Uts¯aha is resolute perseverance, but is prototypically represented as heroic devotion to a cause in the face of great odds. Clearly, they are not outcome emotions. They are not emotions that arise at the resolution of a series of events. But, at the same time, they are not junctural emotions. They are not momentary responses to temporally thin events. They are, rather, emotions that motivate us to engage in the actions that make a plot. They are feelings that drive characters to proceed through the various junctural emotions to a final outcome. I shall refer to them as “sustaining” emotions. Clearly, actions based on romantic love or heroic perseverance are actions that one hopes will result in happiness. But, again, this is not generic happiness. Rather, it is happiness of a specific sort. It is happiness resulting from the achievement of a particular goal. In keeping with this, the Sanskrit theorists maintain that every story is organized around one of four more particular goals, because these are the four goals around which all human life is organized: k¯ama, artha, moks.a, and dharma. “Moks.a” or spiritual release and “dharma” or ethical duty are “higher” goals. Though they are important to a complete treatment of our topic, I shall leave them aside for the moment. They have less direct bearing on the topic of emotion and narrative. Moreover, they are, to some degree, narratively parasitic on the other goals. Thus, narrative treatments of moks.a are often modifications of narratives based on k¯ama or romantic union; narrative treatments of dharma are often integrated into stories of artha or political ascendancy. (We shall return to these points in subsequent chapters.) In this way, k¯ama and artha are the two fundamental or primary goals in the Sanskrit scheme. “K¯ama” means “pleasure,” especially sexual pleasure, but its prototype is the romantic union of

The Mind and Its Stories

lovers. To take a non-Indic example, Chikamatsu expressed the idea in Love Letter from the Licensed Quarter: “The only happiness in this broad world” is “True love to true love” (234). “Artha” refers to material well-being, and is often translated as “prosperity.” However, an examination of the classical texts on artha – such as the Artha´sa¯ stra and N¯ar¯ayan.a’s Hitopade´sa – indicates that the prototype for artha is political power or practical social authority. As the Malian Epic of Son-Jara puts it, “All people . . . seek to be men of power” (Fa-Digi/I. 1277–8). The two points are combined by Euripid¯es’ Hecuba, speaking to the dead Astyanax: “[I]f you had enjoyed youth and wedlock and the royal power than makes men gods, then you would have been happy” (Trojan Women 200). These two goals are obviously correlated with the two sustaining emotions peculiar to the Sanskrit list – rati and uts¯aha. Rati (romantic love) is the emotion that inspires and accompanies one’s pursuit of k¯ama (romantic union). Uts¯aha (heroic commitment or perseverance) is the emotion that sustains one’s pursuit of artha (social and political power).

third hypothesis: romantic union and power as the prototypes for happiness and the bases for prototypical narratives This suggests a solution to our quandary about the precise prototypes operating in the formation of narratives from outcome emotions. Thus, it leads to my third hypothesis: Romantic union and social or political power (including material prosperity) are the two predominant prototypes for the eliciting conditions of happiness. Thus, they are the prototypical outcomes from which our prototype narratives – including literary narratives – are generated. Put differently, romantic union and social or political power are the goals sought by protagonists in prototypical narratives. The corresponding prototypes for sorrow are the death of the beloved and the complete loss of social or political power, typically through social and political exclusion, either within society (through imprisonment) or outside of society (through exile). The position of romantic love is clearest. Oatley points out that “being in love is, in our society, the very paradigm of happiness” (370; see also 364) – though I think Oatley slightly misphrases the matter, which is more properly “being united with the beloved.” When the

Four Hypotheses on Emotion and Narrative

lovers are separated, being in love is part of the paradigm for sorrow. More significantly, J. L. Freedman “conducted a large questionnaire survey of 100,000 Americans” and found that the one thing “respondents most closely associated with happiness,” which is to say, the one condition they most spontaneously identified as the prototype case of happiness, is “love in marriage” (Oatley 361), which is to say, enduring union with the beloved. As to sorrow, M. A. Conway and D. A. Bekerian did research asking subjects to list “the sorts of situations in which a person might typically experience an emotion” (154) – a listing that is likely to rely on prototypes. They then translated the most frequent attributes into sentence pairs for further research. They do not report the initial data. However, they do provide the sentence pairs (189–91). For the group “Grief, Misery, Sadness,” over half the resulting sentences describe the death of a loved one or some other permanent separation from a loved one, and the majority of these treat the death of a romantic beloved. As to artha, a quarter of Conway’s and Bekerian’s “Grief, Misery, Sadness” sentences concern poverty or professional failure. A further sentence pair concerns permanently leaving one’s home – in effect, exile. Indeed, only one sentence pair does not fit into one or the other category. That pair concerns bad weather. It is no doubt prototypical, and relevant to literary study. But it bears most obviously on universal image patterns, as they derive from our lexical entries for emotions, rather than on narrative structures.7 Only one-sixth of the corresponding positive group treats professional success. However, that is because five-sixths of the positive group concerns or suggests some form of love, most often romantic love.8 7

In fact, it does fit the third prototype narrative structure, discussed in Chapter 6. I have left this and related observations aside in order to preserve the sense of a research program advancing in the course of the book itself. Another relevant aspect of the Conway and Bekerian study concerns the grouping of emotion terms. Conway and Bekerian found that some emotion terms tend to cluster together. For the most part, the clustering is definitional – thus, “terror,” “fear,” “panic,” “anxiety,” and “apprehension” are linked. However, in two important cases, the groupings are not simply definitional. One group includes not only “joy,” “pleasure,” and “happiness,” but also “love.” Another links “hate,” “anger,” and “jealousy” (152). It seems clear that there is an implicit narrative linking the terms in both cases – the prototype narrative of romantic tragi-comedy, as we shall discuss in a moment.

The Mind and Its Stories

Work by Shaver, Schwartz, Kirson, and O’Connor has direct bearing here as well. This work is particularly relevant to the present study, for Shaver et al., treat prototypicality. First, they stress that some emotions (for example, fear) are viewed as more prototypical than others (for example, boredom). Second, and more important for our purposes, they note that particular emotion terms have prototypical features. Though they do not fully develop the implications of the second idea, their research further supports the preceding hypotheses. Specifically, Shaver et al., had subjects record personal or “typical” accounts of particular emotions. A group of “coders” then abstracted “features” from these stories. The researchers note that they “were coding antecedents at a fairly abstract level” (1082). In tabulating their lists for sadness and happiness, then, these writers included a number of general ideas that do not bear on our definition of prototypes. However, they included a number of directly relevant points as well. Their category for sadness includes eight features of eliciting conditions. The remaining features all bear on expressive or actional outcomes. Of the eight eliciting condition features, three are general – though it is worth remarking that they quite clearly involve evaluating the final point of a narrative (for example, the first is “Undesirable outcome” [1074]). In this way, they lend support to the larger, narrative account of emotion. The more concrete or prototype-like eliciting condition features are the following: “Death of a loved one,” “Loss of relationship; separation,” “Rejection, exclusion, disapproval,” and “Discovering one is powerless, helpless” (1074). The first and second are part of the romantic prototype; the fourth is part of the social dominance prototype; the third, though somewhat abstract, suggests both – the romantic prototype through “rejection” and the social dominance prototype through “exclusion” and “disapproval.” (I have left out one final concrete feature, “Empathy with someone who is sad, hurt” [1074]. Its presence is extremely important – indeed, crucial for literature – but the other person’s sadness or hurt is itself presumably imagined in terms of “death of a loved one,” and so on. Thus, its inclusion does not affect our understanding of the relevant prototypes.) The happiness category is also directly supportive of the preceding claims. However, it is somewhat more affected by the coders’

Four Hypotheses on Emotion and Narrative

preference for abstraction. Six of the nine eliciting condition features are too general – though, again, they do indicate the narrative structure of emotion (“Task success, achievement,” “Desirable outcome; getting what was wanted,” “Getting something striven for,” and so on). The more concrete features are the following: “Receiving esteem, respect, praise,” “Being accepted, belonging,” and “Receiving love, liking, affection” (1075). The first is clearly part of the social power prototype. The second is too, for the power sought by the protagonist in narratives focused on artha need not be individual; indeed, both in fiction and in life, it is most often a matter of collective domination, the power of one’s in-group. In this way, in-group definition is crucial to such stories – as indicated by the place of social isolation in the prototype eliciting conditions for sorrow. (We shall return to this point in subsequent chapters.) The third feature is clearly part of the romantic prototype. Interestingly, the link between happiness and power seems more obvious here than that between happiness and love. Evidently, however, this was not the case with the stories. Shaver and his colleagues emphasize the very close relation between love and happiness in their study (see 1078–9), going so far as to assert that “love may be conceptualized as a personalized form of joy” (1079). But why are there two prototypes here, rather than one? In order to answer this, we must note, first, that prototypes are in part contextdependent. Our prototype for “bird” is something along the lines of a robin. But if we are out at sea and I say that there are birds on the prow, you will be surprised if you look and discover robins. Kahneman and Miller make the same general point when they discuss interpretations of the sentence “Jane owns a small dog”: “The same statement will yield different norms of size and different ideas of the size of Jane’s dog if she is known to live in a New York apartment or on a farm in Maine” (140). The different norms here are a matter of different prototypes for “dog” in these different contexts. The two prototypes for happiness and sorrow are, like all prototypes, context-dependent. The romantic prototype bears on personal relations, one’s private or familial life. The power or domination prototype bears on relations with society at large, one’s public or social life.

The Mind and Its Stories

ﬁnal hypothesis: the universality of romantic and heroic tragi-comedy and their derivation from personal and social prototypes for happiness The first and second hypotheses presented previously are, prima facie, good candidates for universals. Indeed, given what we know about the human mind, it would be very odd for those hypotheses to be true, but culturally relative. In other words, they may be false. But, if true, it seems extremely likely that they are true for everybody. For example, it would be very strange were it to turn out that fundamental structural principles of emotion entries in the mental lexicon are culturally variable. However, the third hypothesis seems, on the face of it, less likely to be a universal. Perhaps these particular prototypes apply to modern America or Europe, but – one might argue – surely Africans, Indians, and Chinese think differently; surely, the prototypes apply only to a narrow cultural and historical area. But, in fact, this does not appear to be the case. Rather, it appears that these are cross-culturally dominant prototypes for happiness and sorrow. First, contrary to most people’s intuitions, semantic prototypes are often more likely to be universal than are necessary and sufficient conditions. Color terms provide a striking instance of this. The boundaries of color terms vary considerably across languages, while prototype cases are relatively constant, or vary only within specifiable limits (see Comrie 36–8). Second, the predominance of these prototypes for happiness is particularly clear in literary narratives. Indeed, it seems unlikely that there would have been any extended debate on, for example, the universality of romantic love (see Jankowiak and Fischer) had researchers paid more attention to literary representations of emotion across a range of cultures. In any case, this leads to my final hypothesis: Cross-culturally, there are two prominent structures of literary narrative, romantic and heroic tragicomedy, derived respectively from the personal and social prototypes for happiness. As noted in the introduction, for a number of years, I have read extensively the literatures of the major written traditions and a wide range of oral traditions as well. The evidence for the broad recurrence of these structures seems to me virtually indisputable. Some colleagues have asked me if this means that I have put forth a theory

Four Hypotheses on Emotion and Narrative

that is not falsifiable. No, it does not mean that at all. In general, I doubt that any complex theory is open to definitive refutation. However, this theory is no less falsifiable than others. It makes predictions that can certainly run up against contradictory data. Indeed, I discuss some data of just that sort in Chapters 4 and 6. My point is only that reading in various traditions reveals these structures over and over in such clear forms that I do not see how anyone could read the material and not find these structures. The following sections outline the main features of romantic and heroic tragi-comedy and present some of the evidence from a range of traditions. (As it is somewhat repetitive, I have placed most of this evidence in notes.) In presenting this evidence, I have, of course, cited stories that are either romantic or heroic tragi-comedies. These are, after all, what show the cross-cultural recurrence of my two narrative structures. As I have already stressed, most literary universals do not occur in every work of literature, just as most linguistic universals do not occur in every sentence. For example, any universal linguistic principle bearing on plurals will not apply to sentences with all singular nouns. No one would cite sentences with only singular nouns in discussing a universal bearing on plurals. Similarly, I do not cite examples of stories that are neither romantic nor heroic tragi-comedies. Nonetheless, this mode of presentation may give rise to some misunderstandings. I should deal with those briefly before going on. As I have already noted, my claim here is not that all narratives are of these two forms. Narratives may be about anything. I can tell a story about how the post office has repeatedly failed to forward important bills to my summer address. This is neither romantic nor heroic. Indeed, canonical works of literature are not necessarily romantic or heroic. For example, Waiting for Godot does not seem to be heroic or romantic in structure (though a reading that emphasizes its relation to the French Resistance may make it heroic for a given reader and a reading that emphasizes its hom*oerotic implications may make it romantic for a given reader). It is important to note that in some cases these structures may be present, but not easy to recognize. For example, one might argue that Alain Robbe-Grillet’s postmodern La Jalousie is a variation on the romantic plot. Indeed, I would contend that many postmodern works take up romantic and heroic stories, but

100

The Mind and Its Stories

conceal them through radical manipulations of discourse or presentation (for example, through rearranging scenes, cutting explanatory links, shifting perspectives, and so on.). Nonetheless, there are many works that are neither heroic nor romantic tragi-comedies. Once again, my contention is that heroic and romantic tragicomedy are prominent among highly esteemed narratives – and that they almost exclusively account for what we consider prototypical narratives. Put differently, given the innumerable possible structures that count as stories, the likelihood of any given structure recurring cross-culturally with any degree of prominence is about zero. Heroic and romantic tragi-comedies not only recur across all or nearly all traditions, they constitute a high percentage of canonical stories in those traditions (probably well over fifty percent, though it is not entirely clear how one would come up with a precise figure here, or even if that is a very productive way of approaching the issue). Moreover, they constitute virtually the entire field of what we consider good examples of storiness. Waiting for Godot is a canonical work that is neither heroic nor romantic. But it is also very far from being a prototypical narrative. In connection with this, I should also stress that my claim about these universals is not merely statistical. My contention is that the prominence of these prototype narrative structures is an absolute or at least near absolute universal. Some readers have found this point confusing, so it is worth developing for a moment. Once again, an absolute universal is a universal that recurs across all traditions. If heroic and romantic tragi-comedy are the prominent forms of canonical and popular narrative in all traditions, then their prominence constitutes an absolute universal – even though there are many stories in every tradition that are neither heroic nor romantic tragicomedies. Again, many stories in each tradition do not conform to these structures. But that is simply irrelevant to their status as absolute universals. I hope these points serve to clarify my differences from some other writers who have sought to isolate universal narrative patterns. In addition to basing my conclusions on more systematic reading in unrelated literary traditions, I have put forth more qualified claims. As will become clear in the following chapters, I do believe that heroic and romantic structures are very widespread outside of canonical

Four Hypotheses on Emotion and Narrative

101

literature and that they are deeply consequential for our thought and behavior. However, I do not believe that I have discovered the key to all stories. This is not a monomyth (or a duomyth). Indeed, this is far from the only structure one might isolate even from these particular stories. Like everything else, each story combines many different structures. The romantic and heroic structures have unusual importance – for narrative and for human thought. But they are not everything. Moreover, the importance they do have is primarily the result of the prototypes for happiness. In fact, these prototypes are arguably far more significant than the story structures they generate.

romantic tragi-comedy The most common plot structure across different traditions is almost certainly romantic tragi-comedy, the story of the union, separation, and ultimate reunion of lovers. Note that what has to be explained here is not merely the fact that separation and reunion of lovers is a common narrative theme, but that it is evidently the most universal literary topic. In addition, the precise characterization of the separation in terms of death and other recurrent details have to be explained as well. More exactly, as noted in Chapter 1, the standard structure of romantic tragi-comedy involves two lovers who cannot be united due to some conflict between their love and social structure, typically represented by parental disapproval. This conflict commonly involves a rival as well, a suitor preferred by the interfering parents. The lovers are separated, frequently through exile and imprisonment. This separation often involves death or imagery of death. In the end, they are reunited, sometimes following a direct conflict with and defeat of the rival. It may happen that the reunion of the lovers takes place only in the afterlife. We are dealing here with a highly specific, highly complex, and very widespread literary universal. We need to account for its specificity, complexity, and extent. At the same time, we are dealing with a universal that does not determine all works of literature. Again, there are plenty of stories that do not follow this structure. Thus, we need a theory powerful enough to explain the predominance of this structure, but not so powerful as to predict (falsely) that all literary

102

The Mind and Its Stories

works will be of this sort. Put differently, we need to explain the prototypical nature of this structure, for we are clearly not dealing with necessary and/or sufficient conditions for narrative here. This is precisely what the preceding hypotheses do. They predict that the union of lovers – the prototype eliciting conditions for personal happiness – will define the outcome goal for a predominant narrative structure, with romantic love as the sustaining emotion. But this is not all that can be explained by the preceding hypotheses. Again, the structure of romantic tragi-comedy is quite consistent across cultures and historical periods. It not only involves the ultimate union of lovers, but their prior separation, a separation closely associated with death – sometimes the lovers die and are reunited only in death; sometimes there is a rumor of death or an apparent death; sometimes there is a death and resurrection; sometimes there is a reunion in a divine world that suggests death; sometimes there is a near death; sometimes there is extensive imagistic and ¯ metaphorical reference to death. Instances would include Chariton’s Chaere¯as and Kallirrh˘oe¯, where both lovers are believed dead, though one is abducted and the other is sold into slavery; Shakespeare’s The Winter’s Tale; Bh¯asa’s Vision of V¯asavadatt¯a (in Woolner and Sarup), where the heroine’s death is faked in order to fool her husband, so that he will marry someone else; the Thai folk drama, Manohra (in Brandon Traditional Asian Plays), in which Manohra is to be sacrificed by her father-in-law, though she manages to escape and travel to a heavenlike sanctuary; Chˆeng’s The Soul of Ch’ien-Nu¨ Leaves Her Body, where the heroine nearly dies due to separation from her beloved; the R¯am¯ayan.a, where the heroine contemplates suicide in separation, and so on. We can account for these data if we extend our third hypothesis only slightly. Clearly, the construction of a plot leading to the prototype eliciting conditions for happiness must necessarily develop through a period when those conditions do not obtain. This period will constitute the Aristotelian “middle” or the “progression,” in the Sanskrit terminology. That seems fairly obvious. Here is the addition to our hypothesis: this middle or progression is prototypically assimilated to the correlated prototype eliciting conditions for sorrow. Thus, when lovers are separated temporarily, we tend to assimilate that junctural separation to a prototypical outcome separation; we tend to identify the

Four Hypotheses on Emotion and Narrative

103

junctural sorrow that will end with the outcome sorrow that is the result of death. This is plausible for two reasons outside of the data on romantic tragi-comedy (and parallel data on heroic tragi-comedy, as we shall see). First, we have a general cognitive tendency to choose alternatives from within a lexically defined semantic field. The eliciting conditions for personal happiness and personal sorrow are stored in our lexicons in a way that makes one readily accessible from the other. Thus, when forced to devise a scenario in which the happiness prototype does not apply, we have the sadness prototype ready to hand. Even if we do not take it up literally, its cognitive salience is likely to foster the use of relevant imagery, metaphors, and so on – in this case, images and metaphors of death. A second reason we might expect the “middle” or “progression” to develop in this way is that happiness is intensified by at least some degree of preceding sorrow. As Frijda has noted (Emotions 323), we become accustomed to happiness, so that the joy resulting from joyful conditions tends, in our actual experience, to fade, to become the norm. The contrast between the sorrowful progression and the joyful conclusion serves to prevent this fading and to intensify the final joy. Especially given the salience of the sorrow prototype, it is unsurprising that storytellers would discover this and that the practice would become well established in different traditions. This analysis also allows us to give at least a preliminary account of romantic tragedy. Though not nearly as common as romantic tragicomedy, it is still widespread – ranging from such European dramas as Romeo and Juliet, to the love suicide plays of Japanese Kabuki. In the context of the present theory, romantic tragedies are still organized by reference to the prototype goal of happiness in romantic union, but they stop at the point of prototypical romantic sorrow – most often, the death of one or both lovers. Tragedies are, then, truncated tragi-comedies – unsurprisingly, as the prototype goal toward which the entire narrative aims, the goal pursued by the protagonists, is necessarily the comic goal of romantic union. This is particularly clear in a play such as Romeo and Juliet that, with slight changes (for example, had Juliet awakened from her deathlike slumber only a few moments earlier), would be a paradigm case of romantic tragi-comedy.

104

The Mind and Its Stories

This account – which we shall extend to heroic plots in the next section – gains support from some otherwise surprising facts about tragedy. First, tragedy in general (not merely romantic tragedy) is considerably less common cross-culturally than comedy – a fact that makes much more sense if tragedy is understood as a transformation of comedy than if the two are seen as alternative genres of the same type. Moreover, many prime instances of tragedy are integrated into larger cycles that are, ultimately, comic. Consider, for example, Aeschylus’ Oresteia and Sophocles’ Oedipus plays, or, to take a modern example, the two parts of Goethe’s Faust. Along the same lines, the performance of Japanese No¯ and Bunraku dramas was structured so that tragic plays came in the middle of the day, with comic plays following, so that “all programs end happily” (Gerstle 60). All of this suggests that comedy is, indeed, the primary form, and tragedy operates as a shortened version of comedy. In both romanic tragi-comedy and romantic tragedy, there is another recurrent motif in the narrative middle or progression that fits this analysis as well, if more complexly. In addition to the suggestion or imagery of death, it often happens that one of the lovers is exiled or imprisoned or threatened with exile or imprisonment. This motif is found in works ranging from Romeo and Juliet to the R¯am¯ayan.a, where ˙ and, later, exiled from Ayodhy¯a; Niz.a¯ m¯ı’s S¯ıt¯a is imprisoned in Lanka ¯ where Majnun ¯ is (self-)exiled; Chˆeng Teh-hui’s Layl¯a and Majnun, ¨ The Soul of Ch’ien-Nu Leaves Her Body, where Wang Wˆen-Chu¨ has to ¨ spend three years away in the capital; and Ma Chih-yuan’s Autumn ¨ is sent away from her beloved in Han Palace, where Wang Chao-Chun ¨ Emperor Yuan to the court of Emperor Huhanya. Sometimes, there is imagery of this sort, even when there is no literal exile or imprisonment. Thus, for example, Niz.a¯ m¯ı makes repeated use of prison imagery to characterize the condition of both his hero and heroine. This appears to be a matter of interference from the prototype cases of sorrow for social life – exile and imprisonment. It is unsurprising that prototypes from different life-contexts should retain some salience and have some consequences for perception, expectation, narrative imagination, and so on, outside their usual domains. Finally, as we have already suggested, the ultimate reunion of the lovers is often not a simple repetition of the initial union. There is, in most cases, some sacred or otherwise absolutized aspect to the

Four Hypotheses on Emotion and Narrative

105

reunion. The clearest case of this is when the lovers are joined only after death, in heaven. In other cases, there is imagery of heaven or of some parallel condition. In part, this follows from the introduction of the sorrow prototype in the narrative middle. But it is also the fullest possible development of the prototypically enduring character of happiness – just the sort of contrast effect we would expect, given a prototype-based account.9, 10 9

An excellent example of the general structure may be found in K¯alid¯asa’s ˜ ana´sa¯ kuntalam, generally considered the greatest masterpiece of Sanskrit Abhijn¯ ´ drama. Dus.yanta and Sakuntal¯ a fall in love. They are united, then separated. Their ´ separation is not due to a social disapproval of their union. In fact, Sakuntal¯ a’s father entirely approves of the marriage. However, it is due to another sort of conflict with society – a dereliction of social duties that results from their mutual infatuation. Due ´ to his affection for Sakuntal¯ a, Dus.yanta does not adequately fulfill one of his obligations toward his mother. More significantly, distracted by thoughts of Dus.yanta, ´ Sakuntal¯ a fails to give appropriate honor to a religious visitor. He curses her, and the ´ curse leads to her separation from Dus.yanta. Sakuntal¯ a is in effect exiled to another world, in what is clearly a deathlike scenario, while Dus.yanta suffers misery in his palace – with associated imagery of death (for example, spring flowers do not blossom [K¯alid¯asa, Theater 149]). Eventually, they are reunited in a heavenly hermitage. In fact, it is well known that a romantic structure of this sort is pervasive in Sanskrit ¯ drama. Other clear instances include S´ udraka’s Little Clay Cart, Bh¯asa’s Dream of ¯ V¯asavadatt¯a, Hars.adeva’s Ratn¯aval¯ı, and Bhavabhuti’s Uttarar¯amacarita. Though it is a very complex, and very long work, A Dream of Red Mansions manifests much the same structure. A Dream is generally considered “the greatest novel in the Chinese literary tradition” (Hsia, A Dream 262–3). What may be called its main plot line concerns a young aristocrat named Magic Jade. He deeply loves a cousin named Black Jade. But Black Jade is sickly, and from a less-affluent branch of the family. His family disapproves of the relationship, favoring instead a marriage with Precious Hairpin, another relative, a healthier girl, from one of the most powerful families in the area, the Xues (vol. 1, 111–12). In addition, Black Jade’s father intends to marry her to someone other than Magic Jade. Black Jade appeals to Magic Jade’s family against her father’s decision, but, unsurprisingly, she has no success. When she hears that Magic Jade is going to be married to someone else, she resolves to die. Her health, always weak, slowly fails. Magic Jade begins to show signs of mental derangement. Magic Jade’s health improves only when he mistakenly believes that he is to marry Black Jade. Black Jade, aware of the actual marriage arrangements, dies at “the very moment” Magic Jade unknowingly weds Precious Hairpin (vol. 4, 375). Magic Jade goes through the ceremony with his veiled bride. In effect, Magic Jade believes he is in a romantic tragi-comedy. He believes that, after familial conflict, and near death experiences, the lovers have made it through the ordeal of mental and physical sickness, and will now be reunited. When he discovers that he has married Precious Hairpin, he too begins to waste away, approaching death. Unaware of Black Jade’s fate, he asks to be moved to her room so that they will die together (vol. 4, 370). Here, too, Magic Jade imagines a romantic tragi-comedy, for he imagines that he and his beloved will at last be joined in death – and thus, one may assume, in life after death.

106

The Mind and Its Stories

Footnote no. 9 (continued) Eventually, Magic Jade does recover from his illness. But, once recovered, he is not the same. He joins a wandering monk and renounces the world. Hsia contends that “we should perhaps feel happy that he has finally gained wisdom and leaves this world of suffering. . . . But we cannot help feeling that his spiritual wisdom is gained at the expense of his most endearing trait – his active love and compassion.” In Hsia’s view, Magic Jade is “the most tragic hero in all Chinese literature” (270). However, I do not believe that this is a tragic ending. Indeed, I do not believe that it is tragic even from the perspective of the love story. The ending is a version of the standard, spiritualized reunion of the lovers, even if this is only implicit. Specifically, in his assessment of the novel’s end (in Chapters 119 and 120), Hsia leaves out an important point from Chapter 98. When he learns that Black Jade has died, Magic Jade begins “howling” in sorrow until he passes out (vol. 4, 371). In his unconscious state, he finds himself on the road to the Nether World, though it is not yet time for him to die. A “stranger” on the road tells him that he must return to life, explaining that “if you really want to find [Black Jade] you must cultivate your mind and strengthen your spiritual nature. Then one day you will see her again” (vol. 4, 372). After this, whenever he was tempted by thoughts of suicide, Magic Jade “remembered the words of the stranger” (vol. 4, 374). Magic Jade’s final attainment of spiritual release is, in part, a fulfillment of the stranger’s prophecy, a spiritualized reunion with Black Jade. Of course, this may be a triviality, for perhaps the reunion is merely the same reunion he experiences with all life. But the spiritualized love is not undermined for being more spiritual. At the end of the novel, Zhen Shi-yin says “If the Fairy Flower [that is, Black Jade] regained its true primordial state, then surely the Magic Stone [that is, Magic Jade] should do likewise.” When Magic Jade’s “worldly karma was complete” – when the effects of his past acts no longer bound him to the world, his “substance had returned to the Great Unity” (vol. 5, 371). This is, in effect, an absolutization of the romantic reunion. The structure is quite common in Japan as well. Hanakatami, a No¯ drama by Kan’ami, revised by Zeami, concerns a prince and his beloved who are separated when the prince becomes emperor and leaves for the capital. They are reunited when they meet by chance and the emperor recognizes a flower basket that he gave to the woman upon his departure (see Waley 263–5). The Reed Cutter, by Zeami, concerns two lovers who are separated because of the man’s poverty. The wife travels to the city to seek employment. Having succeeded in gaining a position, she returns to the village to find her husband. After some slight complications, they are reunited and “return to the city” together (Keene, Twenty Plays 162). These plays involve love, separation, exile, and reunion. In both cases, the reason for the exile is unusual, but nonetheless a variation on the standard structure. Specifically, there is no representative of society who forbids the union of the lovers. However, in each case the lovers are separated due to some disruptive social condition. In the case of Hanakatami, the lover is assuming a social role that inhibits his individual freedom. His role as emperor circ*mscribes his personal choice and action. In The Reed Cutter, it is poverty that separates the lovers. Indeed, this play is remarkable not only for presenting poverty as a cruel and unjustifiable impediment to love. It is equally remarkable for making the woman the active figure who goes away to seek success before saving her husband.

Four Hypotheses on Emotion and Narrative

107

Footnote no. 9 (continued) An in some ways, more standard narrative may be found in Zeami’s Lady Han. In this story, a young man and a prostitute fall in love. The man leaves, promising that he will return. The woman’s proprietress, unhappy to find the heroine mooning over one client, dismisses her and sends her away. The lovers are eventually reunited. Clearly, the proprietress is the interfering social figure here. But again there is a difference, for in this case the objection to the lovers is that their love is decreasing economic productivity. The woman is not making money for the proprietress. Here, as elsewhere in Zeami, the conflict between the lovers and society or the representative of society is very practical. Of course, the point holds in other traditions as well, where the parents’ choice for their child’s mate is usually someone of greater wealth or social prestige. In these last two cases from Zeami, however, the systematic injustice of social structure is perhaps more obvious than usual. Of course, romantic plots in Japan are not at all confined to work by Zeami. Consider, for example, the eighteenth-century play, Love Letter from the Licensed Quarter. ¯ This work represents the relationship of Izaemon and Yugiri, a relationship seen as “typical” for its genre of Kabuki drama (Brandon Kabuki 6). Izaemon, a wealthy ¯ young man, and Yugiri, a prostitute, have fallen in love. However, Izaemon has been disinherited by his family. Thus, he must leave the licensed quarter. Here, the parental/social interference is presented in a distanced, but nonetheless effective ¯ manner. The moment of exile is clear as well. So is the imagery of death, for Yugiri falls ill due to the separation (226), and even feels close to death (231). Izaemon re¯ turns to visit Yugiri. However, believing that she has been unfaithful, and has been enjoying the company of a wealthy samurai, he spurns her when they meet. She convinces him that he has no rival. They are reconciled and, by good fortune, his family decides to reinstate him, sending him chests full of money, enough to pay his ¯ debt and to purchase Yugiri’s freedom (236). A tragic version of the structure may be found in Chikamatsu’s Love Suicides in the Women’s Temple. Kumenosuke is to be a priest and Oume is to be married to Sakuemon. However, Kumenosuke and Sakuemon fall in love. The High Priest of the temple denounces Kumenosuke and evicts him when he learns of the affair. Oume’s parents push ahead with the marriage to Sakuemon, even when they learn of the love between Kume and Oume. Oume’s father directly rejects a suggestion that this couple marry (Chikamatsu, Major Plays 146). The lovers escape together and take refuge in a women’s temple where, by coincidence, Kume’s sister has brought the ashes of their recently deceased father. Kume conceals his identity. In the temple, thus in a spiritually elevated place (reminiscent of the final reunion of lovers in full comic versions), Kume and Oume realize that they have no place to go. Calling on the Buddha to redeem them, they commit suicide. However, before this, Oume recalls the love she shared with her mother and Kume honors his father’s ashes. They pray: “We shall be reborn on one lotus with our parents” (159). A reconciliation with parents is often part of the comic conclusion in a romantic tragi-comedy. This prayer for posthumous reconciliation in a tragic play is one of those striking variations on standard structures that makes for a great and distinctive work. Turning to the Persian and Arabic traditions, we have already noted the relevance ¯ In this story, Layl¯a and Majnun ¯ fall in love, but Layl¯a’s father of Layl¯a and Majnun. ¯ refuses to allow them to wed. Rather, he marries Layl¯a to another man. Majnun

108

The Mind and Its Stories

Footnote no. 9 (continued) goes into self-imposed exile in the desert, while Layl¯a remains a sort of prisoner in her home. (Though, in an unusual variation, Niz.a¯ m¯ı presents Layl¯a’s husband as a relatively admirable man who does not force himself on Layl¯a, but suffers the grief of unrequited love, eventually dying from this illness [see 164].) In the end, they both die and are finally united in heaven. An angel explains that “in the fabled garden . . . they suffer grief no more. So it will be until eternity” (176). Needless to say, this is not the only romantic tragi-comedy in the Arabic and Persian tradition; indeed, Niz.a¯ m¯ı’s is not the only version of this particular story, which has been widely rewritten (see Gelpke xi on the literally hundreds of versions ¯ of Layl¯a and Majnun). Arabic instances of the romantic plot would include such stories as the final voyage of Sindb¯ad and the tale of Aladdin, from The Thousand and One Nights. Restless Sindb¯ad sets out on a journey where he marries and falls in love with a beautiful young woman (the marriage precedes the love in this case). He settles in the woman’s country, only to find that all the people there are demons – literally. At one point, he praises Allah and is therefore abandoned “on the top of a high mountain” (Dawood, Tales 160). With the help of God, he manages to make his way back home and his wife is “overjoyed” (161). She explains that the people are all devils and that her father had come there from a foreign land. They flee the satanic community and return to Sindb¯ad’s home. The story follows the romantic structure point for point – romantic union, followed by separation, a sort of exile, and imagery of death (the transportation by demons up into the sky). Moreover, the separation is due to a conflict between the lovers and the larger community. The lovers are reunited, and the reunion has the usual spiritual component. One thing that is remarkable in this story is the treatment of society. The society that opposes romantic love is almost always criticized in romantic plots, but it is rare for that society to be demonized literally. On the other hand, this characterization is a simple extension of the standard spiritualization of the lovers’ final union. If the final union of the lovers is godly, then it would seem that their separation was demonic. “The Last Voyage of Sindbad the Sailor” merely takes up this clear structural implication. In the story of Aladdin, our hero falls in love with the beautiful Princess Badral-Budur. His mother claims that they cannot marry because of the class difference: “Your father was the poorest tailor in this city” (185). Aladdin insists that he will die if they are not united. Through magic, he convinces the Sultan to agree to the marriage. However, the Sultan later marries the princess to another man – thus acting as the forbidding father and introducing the rival. Through magic, again, Aladdin manages to separate the newlyweds, ultimately ending the marriage and wedding the princess. The new couple is blissfully happy, but then another interfering character enters, a second rival, himself a magician. This magician abducts the princess, taking her to Africa. When she is abducted, the Sultan threatens Aladdin with death (thus introducing the standard elements of exile, imprisonment, and death). Aladdin defeats this new rival, and returns home with the princess. I take it that the ubiquity of this structure in European literature does not require demonstration. It is the standard structure of New Comedy and its later progeny ¯ (such as Shakespearean comedy), common in prose romances, such as Chariton’s Chaere¯as and Callirrh˘oe¯, as already mentioned, and a staple of popular cinema. Though it ends tragically, Romeo and Juliet, provides a very clear example. It begins with the

Four Hypotheses on Emotion and Narrative

109

heroic tragi-comedy Heroic tragi-comedy also exhibits a surprisingly specific crosscultural pattern, though there is somewhat greater variation in this structure than in that of romantic tragi-comedy. The fullest version begins when the rightful leader of a society is displaced from rule or prevented from assuming rule, most often by a close relative. He/she

lovers separated due to the parents, who prefer a rival. The separation leads to the exile of Romeo and the virtual imprisonment of Juliet. This is followed by the deathlike state of Juliet. In a standard tragi-comedy, Juliet would have awakened from her deathlike sleep to be reunited with Romeo. However, Shakespeare delays her waking just long enough to make it a tragedy rather than a comedy, though even here the lovers are in effect united in death. Romantic tragi-comedy is not confined to written traditions either. For instance, many of the Native American tales recounted by Lévi-Strauss follow this pattern. Lévi-Strauss’s “key myth” (The Raw and the Cooked 35–7), a Bororo story, has this general structure, if in a peculiar form (a form that may suggest a psychoanalytic provenance for the genre). The hero has sexual relations with his mother. Unsurprisingly, his father disapproves of this union. He tries to kill the son, and eventually succeeds in driving him away so that he must live in exile for some time. The boy does eventually return and kill his father. With only slight changes, we have a very standard romantic plot. The main difference is that the interfering father is fused with the rival – often a figure whom the hero must overcome or even kill. In other words, the two most common blocking characters are combined. This fits the social conflict particularly well, for what could define a greater conflict with society than a violation of the central rule of marriage – the prohibition on incest? It is particularly noteworthy that the story, at least in Lévi-Strauss’s version, develops sympathy with the boy. This is entirely in keeping with the tendency of the genre to support the lovers over society in romantic conflict. However, one might have expected the outcome to be different in a case where the social precept being violated is so central to social structure. Though many of Lévi-Strauss’s myths fit the pattern, a particularly touching instance is from the Arawak. In this case, a Jaguar assumes human form and becomes the wife of a hunter. Though obviously not realistic, the point is that the woman and the man come from groups that are not supposed to intermarry. In general, the social conflict that keeps the lovers apart is the result of some social taboo on marriage across a dividing line – class, for example, or race, caste, region, or nationality. The marriage of a man and a jaguar can serve to represent any taboo marriage of this sort. The couple lives happily, for “She turned out to be an exceedingly good wife” (From Honey to Ashes 257). One day, in keeping with her role as a good wife, she suggests that they visit her in-laws, but on the condition that her husband not reveal her true identity. The husband does tell his mother, who informs the village. Though LéviStrauss does not indicate that the community took any action against the wife, she feels “so ashamed” in front of the group that she flees back into the forest, returning to her life as a jaguar. The story ends during the period of separation and exile: “Her poor husband searched the bush in vain, calling out his wife’s name, but there never, never came any reply” (257).

110

The Mind and Its Stories

is exiled or imprisoned. This exile or imprisonment is linked with death – imagery of death, the threat of death, and so on. While he/she is in exile or imprisoned, the kingdom is threatened by some outside force, typically a (demonized/bestial) invading army or, less often, a demonic beast. The hero defeats the threat to the kingdom. He/she then battles the usurper, and is restored to his/her proper place as leader of his/her society. It should be clear immediately that this pattern is precisely what one would expect from our prototype hypotheses, and the corollaries introduced in connection with romantic tragi-comedy. It is a plot based on achieving the prototype eliciting condition for social happiness – social and political power. Moreover, the relevant prototype eliciting conditions for sorrow define the narrative middle or progression. These prototype conditions are, of course, complete loss of social power, typically through exile or imprisonment, or through the absolutization of such loss in death – primarily the death of the hero, but also the death of the nation (threatened by invasion). The complexity of this structure may appear baffling at first. But it is, in fact, readily explained by the preceding hypotheses. Specifically, the complexity of the heroic plot mirrors the complexity of the social prototype for happiness. This prototype encompasses both individual 10

As I emphasized in the introduction, the relation between the study of universals and the study of culture should be seen as complementary, not contradictory. In keeping with this, writers in culture studies have done very valuable analyses of emotional particularity both in and outside literature. The studies in Lutz and Abu-Lughod are a case in point. Yet, many of these studies appear to assume a universality/particularity dichotomy. Abu-Lughod’s essay on the radical particularity of Bedouin love poetry is especially relevant in the present context. She begins by strongly supporting efforts “to deconstruct the concept of emotion through showing its specific cultural meaning” (25) and urges researchers to undertake “the important task of relativizing” (26). She then focuses on one particular example, presumably an example she takes to demonstrate profound cultural difference: Fathalla and a young woman were in love, but “[T]he young woman’s father decided to refuse to give his daughter to the young man. In despair . . . the man set off for Libya. . . . Some time afterward, the girl’s father arranged to marry his daughter to someone else. When Fathalla heard the news, he composed and recorded these poems and sent the tape to the girl’s brother.” When she heard the poems, “Fifteen days after the wedding,” the young woman “gasped for air, fainted, and then fell over dead.” Abu-Lughod concludes, “This story tells us a great deal about the politics of emotion discourse in Bedouin society” (29–30). It does, as Abu-Lughod’s analysis shows. But it does this in a way that his far from radically particularist. It does so by repeating the most cross-culturally common plot structure in the world.

Four Hypotheses on Emotion and Narrative

111

and in-group domination. The usurpation story takes up the individual component. An individual aspires toward (deserved or rightful) dominance in society. He/she is prevented from achieving that dominance, and is even exiled from the society. However, he/she eventually succeeds. The external threat to the kingdom treats the in-group component. The entire society aspires to (deserved or rightful) dominance over other societies. This dominance is threatened by another group, but ultimately the challenge is defeated. The double structure of the heroic plot is in effect necessitated by the double structure of the social prototype for happiness.11 11

A particularly clear example of this structure may be found in the R¯am¯ayan.a, where R¯ama is exiled from Ayodhy¯a due to a dispute over succession relating to his brother. ˙ While in exile, he enters into conflict with the demonic ruler of Lanka, R¯av¯an.a, who poses a threat to all society. (Indeed, this threat is the reason that the god Vis.n.u became incarnate as R¯ama to begin with.) He defeats R¯avan.a in battle, saving not only Ayodhy¯a, but the entire world. He then returns to Ayodhy¯a in triumph, to assume his rightful rightful place as king. As his brother refused to accept the kingship in R¯ama’s absence, there is no need for a battle between them. The Mah¯abh¯arata, too, involves a structure of this general type, if less perfectly. There, too, a familial conflict leads to the exile of the true rulers, the P¯an.d.avas, who also have to save society in battle with soldiers or supernatural creatures. However, these battles do not serve to reestablish them in their rightful place. That occurs only after a civil war with their usurping cousins. Some Sanskrit drama follows this pattern as well, most ´ ¯ scaryacud ¯ . aman.i, a obviously works drawn from the epics, such as Saktibhadra’s A´ revision of the R¯am¯ayan.a. Indeed, variations on this structure are to be found in the many regional, caste-based, and other versions of the R¯am¯ayan.a found throughout south and southeast Asia (on the variety of R¯am¯ayan.as, see the essays in Richman Many R¯am¯ayan.as). The Tamil epic, Shilappadikaram (Cilappatik¯aram), attributed Ilangoˆ Adigal, is an interesting instance of the patterns we have been discussing. (Tamil is a south Indian language, unrelated to Sanskrit.) The first and second parts are romantic, while the third is heroic. In this part, Shenguttuvan, a great monarch, learns of a new goddess and vows to have a suitable image made for her temple. In order to do this, he has to defeat all kingdoms that oppose his project. The threat/defense sequence is manifest here (even if the threat is rather limited and provoked by the ruler himself). Moreover, the story clearly presents Shenguttuvan’s narrative-defining goal as the social prototype for happiness. Though the overt purpose of the king’s action is honor the goddess, the actual motivation – gaining social domination for both the individual and the group – is not at all concealed. As the chief minister puts it, “No one can stop you [Shenguttuvan] if you choose to impose Tamil rule over the whole sea-encircled world” (160). The Malian Epic of Son-Jara – a widely popular and frequently retold epic, with versions across west Africa – manifests the heroic structure in a robust form. SonJara is prevented from acceding to the throne, and his half-brother becomes monarch instead. While Son-Jara wanders, homeless, seeking refuge and facing the danger of death, a neighboring king, Sumamuru, threatens his homeland. Son-Jara defeats Sumamuru, and becomes king.

112

The Mind and Its Stories

But the heroic structure is not only complex, it is also variable. There are two particularly common types of variation. The first is shared with romantic tragi-comedy – truncation. As with romantic tragi-comedy, heroic tragi-comedy may be shortened into a tragedy, with the deposed ruler killed, or exiled without hope of return. More significantly, perhaps, the heroic structure may be and often is shortened without becoming tragedy, for it is more complex than the romantic structure; there are more elements that may be cut. Some diminished versions do not delete the ending of the full structure, but rather delete one of the two component substructures. Thus, many versions of the heroic tragi-comedy involve the hero’s exile and subsequent defeat of the usurper, but do not include the threat to the

The Nyanga epic of Mwindo varies the structure slightly. Mwindo is prevented from assuming the chieftancy by his father. Indeed, his father makes numerous attempts to kill him. This leads to a period of exile in which Mwindo is given support by his maternal aunt, Lyangura. Ultimately, he defeats his father and becomes chief. It is only after this that a supernatural menace threatens his people – a dragon that he defeats. As these examples suggest, the heroic plot is frequently expressed in epics, especially epics of particular social or political significance. In Japanese literature, the work that comes closest to national epic is The Tale of the Heike. This work develops a broad version of the plot, with the exiled Genji ousting the Taira warriors who had usurped rule. The Taira had usurped rule overtly by forcing Emperor Takakura to abdicate in favor of Emperor Antoku. More importantly, they had usurped rule implicitly by taking over the operation of government. The Taira clan head does “whatever he pleases . . . because the court has lost its authority” (43). Thus “Ranks and offices were not conferred at the discretion of the retired and reigning sovereigns in those days, nor yet by decision of the Regent, but solely as the Heike [that is, Taira] saw fit” (46). It also recapitulates the heroic structure in more specific stories. An interesting case is the rebellion plotted by Narichika. The rebellion petered out before it ever began: Narichika “was holding secret consultations and making preparations of various kinds, but they were mere surface activity; there seemed little chance of their leading to a successful revolt” (62). The Taira clan head discovers the rebellion and the conspirators are killed. Narichika is killed in a way that is particularly “cruel” (84). This is, in a sense, a threat/defense sequence, thus the social part of the heroic plot. However, the threat is to the Taira, who are themselves usurpers, and the threat is not much of a threat either. Moreover, the discovery of the plot almost leads the Taira to attack the Retired Emperor – itself a classic threat scenario. However, this, too, peters out before anything happens. In this way, the Narichika section repeats the threat/defense sequence, but with an interesting ambiguity regarding point of view. What is defense from one perspective is usurpation from another. As we shall see in the next chapter, this is not the only point of ambiguity and ambivalence in the Heike.

Four Hypotheses on Emotion and Narrative

113

kingdom. Others do not include a usurper, but do include a threat to the kingdom. The second common type of variation on this structure concerns not the inclusion or exclusion of narrative sequences, but the precise identity of the hero. I have been speaking as if the hero were invariably the ruler. In many cases, he/she is the ruler. But not always. Sometimes, the hero is a loyal retainer of the ruler. One can see how this sort of shift could result from a number of different factors. Taboos surrounding the treatment of monarchs could easily lead to a focus on retainers. At the same time, ordinary readers might find it easier to identify with a retainer than with a hereditary ruler, easier to imagine themselves in the place of the former than in the place of the latter. Moreover, this substitution is possible in heroic tragi-comedy because the retainer typically does achieve the same prototypical social goal as his/her ruler, if to a lesser degree. The retainer’s power and authority are a function of the power and authority of the ruler. In other cases, the hero changes in the course of the story. An initial hero dies and is replaced by a second hero, usually his/her child or grandchild. For example, a story may begin with a deposed ruler, but conclude with a descendant of the ruler who regains the kingdom. Neither variation significantly affects the overall structural pattern.12, 13 12

Having noted these variations, we can point out some further instances of this structure, giving a clearer picture of its cross-cultural breadth. We have already cited three prominent instances from India and two from Africa, all five being epics. We may augment our single, epiclike example from Japan, The Tale of the Heike, by noting that much Japanese drama fits this structure well, though often with the substitution of a retainer for the king. For example, Chikamatsu Monzaemon’s most popular work (Keene, Major Plays 195), The Battles of Coxinga, concerns an exiled minister who returns with his son to defeat the Tartar invaders and reestablish his monarch. As James Brandon explains, “In a very typical oiemono [history] play the young heir to a feudal house is deposed by a scheming clan retainer or relative and forced to escape in the disguise of a commoner. . . . At the end of the play, the evil retainer is defeated by retainers loyal to the house and the young heir is reinstated” (9). More generally, the broad tendencies of kabuki drama fit our hypothesis well, for in the nineteenth century, it was commonplace among Japanese commentators that the two main centers of play writing were associated with plot structures of the two sorts we have been discussing – Kyoto plays being romantic; Edo plays being heroic (see Brandon, Kabuki 13). There are other prominent African examples too. The Ijo¯ epic of Ozidi (see ¯ Okabou) treats the murder of Ozidi’s father. It recounts Ozidi’s birth in exile, his return to his homeland, and his defeat of his father’s murderers. The Bambara epic about Da Monzon of Segou also involves many episodes of this sort, some treating

114

The Mind and Its Stories

Footnote no. 12 (continued) Da Monzon directly, and others treating the heroic warrior Bakari Dian. The latter are particularly relevant, as he is entirely socially isolated when he goes to defend Segou (see Sissoko III 4). In the Middle East, Ferdowsi’s Persian epic, the Shâhnâme – “considered over the centuries the greatest work of Persian literature and the strongest pillar of Persian identity” (Yarshater, “Foreword” xv) – is also a case in point, manifesting a version of the heroic pattern in many of its sections. (The work covers the entire period from the creation of the world to the last pre-Islamic ruling dynasty. It cannot be considered a single narrative in the relevant sense. It is, rather, a series of narratives.) After some prefatory material, the poem begins with the exemplary reign of Keyumars. His son must battle the threatening Black Demon. The son is killed, but Keyumars’s ¯ grandson, Hushang, defeats the demon and becomes king. Though the throne is not lost in this sequence, the rightful heir is killed, and the injustice of this is undone in the next generation through the defeat of the threatening enemy. This is not only an inaugural narrative in the story, but part of the beginning of civilization. The next generation repeats the threat/defense sequence, when Shah Tahmuras defeats the demons and is named “Demon-binder” for his success (9). These two conflicts prepare for a third and fuller narrative in which the Arab Zah.h.a¯ k, working with lblis (or Satan), manages to take control of Iran. Jamsh¯ıd, the Shah, flees to “the sea of China” – a standard instance of dethroning and exile. In line with the common ¯ inhervariation in which the restoration is deferred to the next generation, Far¯ıdun ¯ whose mother flees with its the divine right to rule. Zah.h.a¯ k searches for Far¯ıdun, him to the Alborz mountains, in another instance where the rightful ruler is denied ¯ eventually defeats Zah.h.a¯ k and restores rightful rule to and sent into exile. Far¯ıdun Iran. A further story introduces the element of familial conflict, completing the pro¯ divides the world into three kingdoms, awarding one totypical structure. Far¯ıdun to each of his three sons. Iraj is given Iran. In consequence, his two brothers become jealous and kill him, after the threat of exile arises (32). The retribution is reserved for Iraj’s grandson, Manuchehr. Years pass until the brothers send their armies against Iran. This threat to the kingdom inspires Manuchehr, who defeats the invaders, eventually assuming the throne. The triumph of Manuchehr introduces the character of S¯am, and thus a sequence of stories related to warrior heroes rather than the monarch per se. In the first of these, S¯am rejects his son, Z¯al, denying him his rightful place, and sends him into exile (with danger of death). They are eventually reconciled, and Z¯al himself has a son, Rostám, probably the single most important hero in the poem. He enters actively into the plot in defending Iran against another invasion and rescuing the ´ abducted Shah Kay Kavus. Here as elsewhere there is no point in going through every story in the book. However, it is worth considering the single most renowned story of the Shâhnâme – the episode of Rostám and Sohráb. Sohráb is Rostám’s son. However, he is living with his mother in Turán. He gathers an army together and invades Iran. There is a conflict between the Shah and Rostám in which the Shah threatens to kill Rostám, thus denying Rostám his rightful position. Rostám leaves briefly, in a sort of mini-exile, swearing that he has cut off his relations with the Shah. The two are

Four Hypotheses on Emotion and Narrative

115

Footnote no. 12 (continued) reconciled as the threat to Iran is imminent. Rost´am returns and, after some difficulty, defeats the invading army. At the climax, he kills Sohr´ab, learning his identity after the fact. Unsurprisingly, the result of all this is not the joyful celebration of triumph one might otherwise expect. As we will see in the following chapter, this sort of ambivalence is not uncommon in heroic plots, for specifiable reasons. In any case, other than that final sorrow, the story is clearly a variation on the structure we have been considering. The ta’ziyeh of Iran comprises a set of dramatic reenactments of the massacre of al-H . usayn and his followers at Karbal¯a and the events surrounding that massacre (see Chelkowski). The basic story of this paradigm narrative is of just the sort we have been discussing. The leadership of the Muslim community has been illegitimately taken from ‘Al¯ı. Moreover, both ‘Al¯ı and his son al-H . asan have been killed. Now al-H . usayn remains, and al-H . asan’s son Qasem. The narrative tells of the battles of Qasem and al-H . usayn to defeat the usurper and restore the rightful line of Caliphs. This story is a tragedy, and thus culminates in the final defeat of al-H . usayn at Karbal¯a – though, of course, Sh¯ı’ah Muslims believe that, in the end, they will triumph. Moreover, as often happens, episodes of the ta’ziyeh frequently incorporate the romantic structure as well. Thus, Qasem is married to al-H . usayn’s daughter, but they are separated on their wedding night as Qasem must go to battle the armies of the usurper. He is defeated and dies after explaining to his beloved that “our marital union shall occur at the Judgment Day” (Humayuni 14). Another relevant Persian work is Gurg¯an¯ı’s eleventh-century poem V¯ıs and R¯am¯ın. The story of two lovers, V¯ıs and R¯am¯ın combines the two genres in an unusual way. The narrative is a complicated one in which two brothers, Moubad and R¯am¯ın, vie for the love of V¯ıs. V¯ıs is awarded to Moubad by her mother, but she falls in love with R¯am¯ın. Moubad discovers their intrigue and exiles V¯ıs. As this standard romantic conflict is proceeding, there is suddenly a threat to the kingdom from Rome. One might expect R¯am¯ın to repulse the invading army. However, Moubad defeats this threat, having imprisoned V¯ıs while he was away at war. The lovers nevertheless manage to meet. After further complications, the lovers are estranged, R¯am¯ın nearly dies, and finally they are again reconciled. At this point, the lovers plot a coup against Moubad. They wrest control of the kingdom. As Morrison explains, “R¯am¯ın is left supreme and is welcomed . . . as a liberator” (xvi). When the lovers die, they are reunited in heaven (351). The romantic part of the story, though complex, is perfectly in keeping with the usual structure. The heroic part, however, is very strange. It includes both the usurpation by a family member (a younger brother, in this case) and the defense of the society against an external threat. In other words, it includes both parts of the full heroic structure. However, the defense against a threat occurs first and is undertaken by Moubad, whereas the usurpation occurs last and is accepted as a just outcome – a particularly striking and unusual, even disconcerting, variation on the structure. Arabic epics often fit here as well. Lichtenstadter summarizes their standard format in a way that shows their conformity to the general pattern – “the outbreak of hostilities, the preparations made by the attacker, the fearful anticipation of the intended victims,” and so on (30). A good example of the full pattern may be found in the S¯ırat ‘Antar or Life of ’Antar. ‘Antar is deprived of his rightful place in

116

The Mind and Its Stories

Footnote no. 12 (continued) society by his father, who treats him as a slave. Conflict with his father leads to his exile. ‘Antar subsequently saves his father from captivity and saves his society from conquest, thereby gaining his freedom. This story is interwoven with a romantic tragi-comedy in which ‘Antar continually seeks union with his beloved ‘Abla. She believes that ‘Antar has died and is self-exiled to the desert before their union. (On the S¯ırat ‘Antar, see Heath.) The central Islamic story of the hijrah, the prophet’s flight from Mecca, is another instance of this structure. The flight was, of course, a sort of exile, precipitated, in the Muslim view, by the society’s rejection of its rightful leader. This rejection was largely led by members of Muhammad’s own tribe, the Quraysh, and most importantly his uncle Abu¯ Lahab (see, for example, Ibn Ish.a¯ q 161). Finally, it culminated in his return from exile and his military victory. The point is not affected by the historicity of the events. For, out of the vast array of details that made up the Prophet’s life, Muslims have focused on this particular sequence of events, selected and shaped, in part, according to the prototypes we have been considering. (The emplotedness of historiography has been recognized at least since Hayden White’s pioneering work. The relation between history and narrative has been explored from the literary side as well, through the consideration of historical fiction – see, for example, Paul Hernadi’s Interpreting Events.) The Turkish Book of Dede Korkut, widely “accepted as the national epic of the Turkish people” (Clinton 1497), is a collection of thirteen short narratives. The great majority of these stories clearly manifest the heroic structure, complete or partial. For example, the first story concerns Boghach Khan and his father Dirse Khan, a Turkish noble. Dirse is convinced by his evil retainers that Boghach is committing dishonorable acts. Dirse therefore agrees to deceive his son on a hunting trip and kill him. Dirse Khan wounds his son on Kazilik Mountain and leaves him for dead. Thus, we have the familial conflict, Boghach’s deprivation of his rightful place in the nobility, his near death and a sort of exile (on Kazilik Mountain). (As to the lastnamed element, Chadwick and Zhirmunsky note that “the ‘expulsion and return’ of the unjustly accused hero” is “typical” in central Asian epics [298].) Boghach’s mother manages to cure the boy, with the help of a wandering saint. However, he remains in hiding – again, a form of isolation from society. At this point, Dirse’s retainers capture Dirse in order to sell him to the “infidel.” This is a variation on the danger to the kingdom. Boghach rescues his father, which is to say, fends off the threat. This leads to the restoration of Boghach to his rightful place. Most of the other stories focus on the second part of the heroic sequence, the threat and defense – often in the form of an abduction and rescue. For example, in the second story, Salur Kazan, one of the great heroes of the book, is away from his ¨ attacks and abducts several of the Turks. camp. The evil “infidel beast,” King Shokli Kazan returns and defeats Shokli before the captives are harmed. It is a great social victory, for in the course of the fight “Twelve thousand unbelievers were put to the sword” (57). There is no need to go through every instance. It is worth noting that only a few stories do not fit the heroic format. Of these, two have romantic goals. Though the epic is notoriously absent from China, military romance and history serve much the same function. C. T. Hsia points out that in Chinese military

Four Hypotheses on Emotion and Narrative

117

Footnote no. 12 (continued) romances the reader is frequently faced with a loyal retainer of the emperor who suffers “imprisonment” or “exile” (“The Military Romance” 362), often pretending to be dead until an “invasion at the frontier makes [his] services indispensable” (363). There is also much work of this sort in historiography. Consider, for example, one of the most important ancient Chinese histories, and one of the first to develop an extended, fluid narrative – the Kuo yu¨ or Conversations from the States. As Burton Watson explains, the longest section of the work concerns, first of all, a conflict between brothers – or, more exactly, between a prince, Shen-Sheng and his step-mother, regarding the prince’s half-brother. This results in the disinheritance of the prince and his death. This is followed by the exile and eventual accession to the throne of another prince, who has, in effect, taken up Shen-sheng’s role and completed the heroic narrative. Though history, it is emplotted history, and its plot is one of heroic tragi-comedy. Three Kingdoms, generally considered one of the five masterpieces of Chinese prose fiction, fits the pattern also. This work is too long and complex to count as treating a single plot, though it has a tighter and more prototypically unified structure than a work such as the Shˆahnˆame. Specifically, Three Kingdoms treats the disintegration of the Han dynasty, the period of its division into three kingdoms, and the return to unity under a new dynasty, the Jin. On the whole, the novel treats the denial of the legitimate emperor, various exiles and deaths of legitimate heirs, and, finally, the reestablishment of legitimate succession in a later generation. While some of the other works we have considered stress the middle part of the heroic structure – the external threat and defeat of that threat – Three Kingdoms treats the opening and conclusion of that structure. It concerns usurpation and the internal strife that goes along with it, rather than invasion and the threat of conquest. The novel begins with a rebellion, defeated by the hero, Liu Xuande, and the man who will ultimately become the villain, Cao Cao. This is a sort of threat to the kingdom. However, it is internal, not external. In any case, it puts us in mind of the theme of usurpation and prepares us for the conflict that follows. Through a complex series of actions, Cao Cao undermines the ruling dynasty, killing or otherwise neutralizing members of the ruling family. After Cao Cao’s machinations, Liu is clearly the only legitimate alternative to the new dynasty, illegitimately proclaimed by Cao Cao’s son, Cao Pi. However, the conflict with Cao Cao has driven Liu away from the capital. He has established himself in the Riverlands, organizing an alternative kingdom from which he can fight against the usurpers. Thus we have all the major elements from the first part of the heroic structure – denial of rightful place to the ruler, death, and exile. In the most straightforward fictional version of such a story, either Liu himself or one of his descendants would ultimately defeat Cao Cao (or one of his descendants), reuniting the empire under its rightful ruler – perhaps after defeating some external threat from, say, the Mongols or another non-Chinese group. Indeed, the novel develops in this direction, with Liu aided by an expert advisor (Kongming), a standard character in such works. However, Three Kingdoms has the task of writing a heroic story within the general confines of history. Historically, China was reunited. In this sense, the usurpation was overcome. However, China was not reunited by a descendant of Liu or any one of the murdered heirs. This difficulty is overcome by the doctrine of the Mandate of Heaven. The proper

118

The Mind and Its Stories

conclusion In sum, I have argued, first, that lexical entries for emotion terms, thus our practical cognition about emotion, are most significantly defined by prototypes. Second, prototypical narratives involve the different types of emotion – junctural, outcome, and sustaining. But they are structured primarily by reference to the prototype eliciting conditions for the outcome emotion of happiness. In prototypical stories, the complex process of narrative construction is guided and organized by the expansion or elaboration of the micronarratives that define the prototype eliciting conditions for happiness, whether or not those

ruler is the person who inherits the Mandate of Heaven, even if that person is not a descendant of the prior ruler. Thus, in the end, the new, unifying dynasty is taken to have inherited this mandate. In this way, the new dynasty serves as the heir of the dethroned monarch and its rule operates as a restoration and an overcoming of the usurpers. I take it that European examples are too well-known to require elaboration. However, we could mention The Song of Roland, Beowulf, King Lear, Macbeth, certain plot sequences from such mixed heroic/romantic poems as Jerusalem Delivered (for example, the story of Rinaldo) and the Niebelungenlied. Even Hamlet or The Tempest could reasonably be understood as a variation on this structure, with magic replacing military battle in the latter case – an unsurprising substitution in that magic enters directly into military battle in many works of this sort. The list could be extended almost indefinitely. For reasons discussed in the first chapter, having to do with methodological requirements for the determination of universals, I have drawn almost all preceding examples from the period before the rise of global European colonialism. Some readers have taken this to mean that the genres I am speaking of have long since passed from this world. The impression has no doubt been made stronger by my references to monarchs and warriors. In fact, however, these genres still continue to dominate literature, and now film also. There are three ways in which this is true. Even if no new works were written in these genres, a massive collection of romantic and heroic tragi-comedies has been canonized in every tradition. These works are remembered, read, acted on stage, filmed, taught in classes – Shakespeare being an obvious example. These works are not merely historical stepping stones by which we arrived at Modernism. They are a continuing part of the body of literature. They are as much a part of the modern period as they were a part of their own periods. Indeed, these are the paradigm works of the various literary traditions. Thus, many of them are more central to our contemporary experience and understanding of literature than they ever were in the past. Indeed, they are more central to our contemporary experience and understanding of literature than almost all contemporary works. Hamlet remains more widely read, seen, and interpreted than any nouveau roman. In keeping with this, such works continue to instantiate our concept of prototype narratives. But, of course, that is not all. Romantic and heroic tragi-comedies do continue to be written today. In fact, to all appearances, they dominate narrative production

Four Hypotheses on Emotion and Narrative

119

Footnote no. 13 (continued) today no less than they did in earlier centuries. Most obviously, these genres dominate popular cinema. As I write, Titanic is the largest grossing film ever made and it is a prototypical romantic tragi-comedy. Its development is not merely simple and straightforward. In fact, its variations on the standard structure are sometimes very sharp and effective. (A good example is the way Jack takes up the female role and “saves” Rose from despair. Rose, then, takes up the male role and saves Jack from physical danger.) However, the general outline is so clearly prototypical that it hardly requires explanation – the lovers face parental opposition, based on social class, with a parentally chosen rival, and so on. The second biggest U. S. box office success ever is from the Star Wars series. The four movies of the series are all in the box office top twenty. (For these figures, see http://www.the-movie-times.com.) The trilogy makes up a straightforward, if rather minimal heroic tragi-comedy. The opening, implicit in the first episode, is usurpation – the usurpation of the Republic by the Empire and, simultaneously, of Obi-Wan Kenobi by Darth Vader. The focus of the trilogy is on the struggle against usurpation. This struggle is waged by Han Solo and, more importantly, by the siblings, Luke Skywalker and Princess Leia, a standard generational displacement, if a somewhat unusual one in that they are the biological children of Darth Vader, even if they are the spiritual children of Obi-Wan Kenobi. This culminates in a final victory against usurpation. There is, in addition, a standard, if simple, love story. We also find the standard elements of exile, and so on. The Phantom Menace has a complete sequence of invasion, usurpation, exile of the monarch, struggle, and final victory. When adjusted for inflation, Gone With the Wind captures first place, by the monetary measure of popularity. In the romantic plot, Ashley and Scarlett are in love. Scarlett’s father opposes her, and in any case Ashley is engaged to another woman. The love story works its way through the four hours of the movie, interwoven with a shortened version of the heroic plot – where the home society is threatened and then defeated by what is in effect a barbarian invasion. (The point that the “barbarians” put an end to slavery is not relevant to the film’s characterization of the national enemy, which follows standard heroic principles in this regard.) It also takes up the standard elements of the death or near death of the beloved (all four lovers almost die), exile (from the home plantations of Twelve Oaks and Tara – the latter imagistically represented as a nation rather than a plantation), return from exile, and the final death and/or separation of the lovers (as it is, like Titanic, a tragedy, if a tragedy with a hopeful end). There are certainly complications here. For example, the love story involves a fairly common variation in which the love triangle is doubled (with Rhett Butler pursuing Scarlett). It includes an unusually sympathetic portrayal of all the rivals. It resolves with a partial realignment of the characters’ feelings in order to accommodate the union of two couples – but then splits all the lovers in its tragic conclusion. Moreover, it presents the heroic plot from the perspective of the women at home, rather than the men in battle. In short, it varies standard structures in interesting ways. Nonetheless, the structures that it varies are the prototype romantic and heroic plots. It is no coincidence that the most popular or successful movies ever (Gone With the Wind and Titanic) have prototypical romantic plots with, in the top case, a heroic subplot, while their nearest competitors (the Star Wars movies) form a prototypical heroic story with a romantic subplot.

120

The Mind and Its Stories

Footnote no. 13 (continued) Finally, the importance of these two genres is not confined to popular narrative, but extends to many works of high Modernism and Post-Modernism. Of course, to a great extent, both Modernists and Post-Modernists set out to break with canonical and popular structures. Thus, it would hardly be surprising if their works, self-consciously set against common practices, had little relation to the prototype narrative genres. Indeed, it would have no bearing on the present argument. If a work is designed to be nonprototypical, it cannot count against a theory of prototypical narratives. However, as it turns out, many of our paradigm Modern and Post-Modern works do indeed treat prototypical stories. Romantic structures in particular are clearly present in major writings by such canonical modernists as D. H. Lawrence and Marcel Proust or even in such a paradigm Post-Modern work as Alain Robbe-Grillet’s masterful nouveau roman, La jalousie. Moreover, the presence of these genres is not confined to such relatively obvious cases. Consider, for example, James Joyce’s Ulysses. Ulysses is probably the paradigm Modernist work. It is certainly not a work that anyone would think of as manifesting heroic or romantic patterns. Indeed, there is not a great deal in the novel that could be considered plot. But there is something. In fact, there are a couple of rough plot sequences. The most straightforward plot sequence centers on Bloom. It is unequivocally a version of the romantic tragi-comedy. Bloom’s wife is beginning an affair with another man that very day. Bloom returns home, considering a divorce. This is a common part of the romantic plot – the entry of a rival and the possible estrangement of the lovers. And, of course, the Blooms’ relationship is one of those “abnormal” unions that are at the center of romantic tragi-comedy. Though not forbidden, they remain a socially suspect couple because of their different ethnic and religious backgrounds. Moreover, the epic model for this part of the novel is itself a version of a small part of the romantic plot – the exilic wanderings of Odysseus and his eventual return to and reunion with his faithful beloved, Penelope. The other, even more minimal plot concerns Stephen. Stephen sees himself as the legitimate poet laureate of Ireland. But he is not recognized as such by anyone else. Rather, Buck Mulligan seems to be receiving precisely the accolades that Stephen (perhaps) deserves. It should be clear already that this is a version of the heroic plot, the plot aimed at social domination. In this case, Joyce has primarily taken up the usurpation part of the story, for Stephen has been set aside from his rightful place of social esteem and authority. Indeed, Joyce explicitly models this part of the novel on two heroic stories of the standard political/military sort, two prototypical stories of usurpation – the Telemachus portions of the Odyssey (recall that Telemachus’s position as ruler of Ithaca is threatened by the suitors) and Hamlet, where rule has been usurped by Claudius. The links with military heroism are stressed at various points. For example, Stephen lives with Buck Mulligan in a tower built for military purposes (Gifford and Seidman 1.542). He thinks of his conflicts with Mulligan as a sword fight: “Parried again” (l. 152), he remarks, referring to one of Mulligan’s replies. He explains that Mulligan “fears the lancet of my art as I fear that of his” (l. 152) – though a lancet is not a sword, the point is that Mulligan, a doctor, fights with a lancet, while Stephen battles with “The cold steel pen” (l. 152–53). At the end of the opening chapter, Stephen aptly names Mulligan “Usurper” (l. 744). Joyce extends the heroic resonances still further, suggesting the element of threat to the nation, by linking Mulligan with Haines, and thus implicitly

Four Hypotheses on Emotion and Narrative

121

conditions are ultimately achieved in the narrative. In connection with this, prototypical stories most often incorporate some version of the prototype conditions for sorrow, placing them in the narrative middle or progression. In some cases, the narrative ends with the permanent establishment of these sorrowful conditions. Rather than being an autonomous, sorrow-based structure, however, stories of this sort are truncated forms of the happiness-based structure. In other words, tragedy is a derivative of comedy. Third, the predominant prototypes for happiness are (a) romantic union with one’s beloved and (b) the achievement of political and social power, both by an individual and by that individual’s in-group (for example, his/her nation). These prototypes are triggered by different contexts – those of personal and social happiness respectively. Finally, these happiness prototypes define romantic and heroic tragi-comedy, which are the most common and most prominent narrative structures crossculturally. Moreover, along with the related prototypes for sorrow (and some well-established cognitive principles, such as priming), they account for the complexity of these universal narrative structures in detail. with the history of English colonialism, “The imperial British state” (l.643), one of the “two masters” (l.638) to whom Stephen has been bound as a “servant” (l.638). This colonial conflict culminates in Stephen’s encounter with the British soldiers late in the novel. In his self-conscious revision of European epic, Joyce takes up the Romantic identification of hero and artist, putting the artist into the prototypical heroic structure – for Buck usurps Stephen’s position as poet, not as king. (In some other modern works – such as the highly popular and award winning play, How to Succeed in Business Without Really Trying – the businessman is put into that structure. The point is not confined to the west, as is clear from the Japanese salaryman who “even today sees himself as the heir of the samurai” [Williams 272n. 13, citing Hiroshi Yoshioka].) Other than that, Joyce’s changes in the prototype story are almost entirely a matter of style and discourse – primarily a matter of radical reduction and ellipsis, combined with a great elaboration of nonstory elements and an often heavy use of irony. This is typical of Modernist and Post-Modernist writers.

4 Writing Beyond the Ending A Problem of Narrative, Empathy, and Ethics

In the preceding chapter, I argued that prototype narratives exhibit remarkable regularity, with detailed patterns of romantic and heroic tragi-comedy recurring prominently across unrelated traditions. Moreover, I argued that this pattern may be accounted for by reference to a cross-cultural constancy in personal and social prototypes for happiness – that is, romantic union and social domination, respectively. In connection with this argument, I noted that the heroic structure is more variable than the romantic structure. For the most part, this variability is not problematic and the various alternatives are easy to accommodate into our general account. However, there is one apparently anomalous variation in heroic tragi-comedies. A surprising number of paradigm heroic tales include a peculiar ending, a sort of negative postscript to the achievement of the heroic goal. Moreover, this is common enough across genetically unrelated traditions to count as a universal. This raises a serious explanatory problem, for it is not at all clear why any story would, so to speak, “go beyond the ending” in this way, continuing the tale after the achievement of its guiding aim. After all, this is extremely uncommon in romantic plots. Indeed, its uncommonness in romantic plots suggests the basic reasonableness of our – or at least my – puzzlement over the heroic “continued” or “excess” ending, as one might call it. It also suggests that this odd ending may have the potential to advance significantly a research program in cognition, emotion, and narrative. 122

A Problem of Narrative, Empathy, and Ethics

123

I shall first lay out the problem, noting some of the primary cases from European, Middle-Eastern, African, Indian, and Japanese literature. I shall then consider what I believe to be the two primary elements in an account of this phenomenon – ethical evaluation and empathy – elaborating on the function of these elements in literary creation and response.

anomalies of heroism: the epilogue of suffering The continued or excess ending of heroic tragi-comedy focuses on pain following the achievement of social dominance. I shall refer to this ending as the “epilogue of suffering.” As with the heroic plot more generally, this is a variable structure. In its least developed form, it simply involves a turn from the triumph of the hero to the sorrow of those who have been defeated. A good example of this is to be found in the Iliad. We are perhaps so accustomed to the ending of this poem that it does not strike us as odd. But it is odd. The antepenultimate book deals with Achill¯es’ triumph over Hector. The poem could have concluded here, for from this point the reader can envision the Greeks’ ultimate triumph, a triumph predicted explicitly by Pr¯ıam (see 257). In this way, it could end like the Aeneid, which concludes with the victory of Aeneas over Turnus, modeled in part on the victory of Achill¯es over Hector.1 But the Iliad does not end this way. In the next book, the Iliad continues to the funeral of Patroclos and the games following this ceremony. This too could be seen as a triumphal ending – though even here, the triumph is marred by 1

Of course, the Aeneid was incomplete (see Knight 11). We do not know how Virgil would have continued his poem. Perhaps he too would have added an epilogue of suffering. Virgil does establish a strict parallel between Turnus and Aeneas, between the Greeks at the outset and the Trojans at the conclusion, as writers such as David Quint have noted (see Quint 66–74). Indeed, Virgil does this to such an extent that the poem seems to lead inevitably to an epilogue of suffering. The point is only re¯ which develops far greater sympathy enforced when one recalls the suicide of D¯ıdo, than the parallels with Cleopatra should have allowed, given the political conditions at Virgil’s time (on D¯ıdo¯ connection with Cleopatra, see Quint 28, 109, and 383n. 14). Having referred to Quint’s Epic and Empire, I should perhaps note what some readers will take to be obvious – that I disagree with a number of Quint’s conclusions about “epics of the victors,” such as the Aeneid. These poems, quite a few of them anyway, are far more equivocal and ambivalent than Quint usually allows (despite occasional disclaimers).

124

The Mind and Its Stories

sorrow. But the ultimate conclusion of the poem is Pr¯ıam’s acquisition of Hector’s body, his return to Troy, and the laments of the women (“grief intolerable came upon every heart” [296]), including Andromach¯e’s prediction of the murder of Astyanax: “The boy is only a baby . . . he I think will never grow up to manhood. . . . [W]ives and little children . . . will soon be carried off in ships, and I with them. . . . [S]ome enemy will catch [Astyanax] by the arm, and throw [him] over the wall to a painful death” (296). In short, the final scene of the poem is one of utter despair, focused in this case on those who have been defeated. This is not some idiosyncracy of Homer’s. The Japanese Tale of the Heike has a similar, pathetic conclusion. The Heike is the major heroic tragi-comedy in the Japanese canon, and as such it has a place in Japanese literary culture comparable to that of the Iliad in the west. It concerns the battle of the Minamoto and the Taira. The Minamoto triumph. But the final chapter turns our attention to an Andromach¯elike figure, the Imperial Lady, Kenreimon’in. Her young son, a boy of eight, was the Taira emperor. He and his grandmother committed suicide to avoid being taken by the Minamoto soldiers who, like the Greeks in Troy, killed all possible Taira heirs. Kenreimon’in retires to a lonely hermitage in the mountains, praying to the compassionate Buddha for the well-being of her dead child and relief of her own suffering: “never, in all the lives to come, shall I forget the Former Emperor’s [that is, her son’s] face. I try to forget, but forgetting is impossible; I try to control my grief, but that is also impossible. Nothing causes such sorrow as parental affection: that is why I pray faithfully for the Former Emperor’s enlightenment, morning and evening. I believe my love for him will guide me to enlightenment, too” (434). The misery is not confined to Kenreimon’in: “there was no house free of disquieting winds, even inside jade blinds; there was no dwelling where dust never rose, even beyond brushwood doors. Husbands and wives who had slept on adjoining pillows were as remote from one another as the sky; nurturing parents and their children were set apart, neither knowing the whereabouts of the other” (437). The final sentences of the story recount Kenreimon’in’s death, followed by a statement regarding the sorrow of her two attendants, their eventual enlightenment, and their rebirth in the Paradise of Buddha.

A Problem of Narrative, Empathy, and Ethics

125

The ending of the Heike introduces some motifs that we find, usually presented from the other side, in the more fully developed or “complex” epilogues of suffering. These include at least some reference to the misery of the defeated enemy (if usually less than in the “simple” versions such as those just discussed), but they focus on the triumphant heroes. In the most completely elaborated version of this epilogue, the hero triumphs, which is to say, attains the goal of social power. But this does not bring happiness. Rather, the hero finds that the losses outweigh the gains, often vastly, and he/she is plunged into remorse and despair. In consequence, he/she sets out on a sort of spiritual journey – or is forced into exile – in which he/she experiences great pain and anguish, explicitly or implicitly as punishment for past acts. This is followed by what is sometimes a sort of spiritual transcendence, but sometimes little more than a glorified suicide. Again, not all these elements are necessarily present in any given case. The most complete case I know of is the Mah¯abh¯arata – unsurprisingly, perhaps, as it is the longest of the relevant texts. This poem concerns the battle between the P¯an.d.avas and the Kauravas. The P¯an.d.avas are the heroes. Of the P¯an.d.avas, Arjuna is the most heroic, in that he is most engaging as a warrior, but Yudhis.t.hira is the most admired as a person, and he is the appropriate ruler as well. When the P¯an.d.avas triumph, Yudhis.t.hira weeps uncontrollably and shouts “I am burning with grief, like a person thrown into a blazing fire” (Vy¯asa vol. 7, “Stree Parva” 43). Subsequently, for other reasons, Arjuna too sinks into despair: “O best of men, it behoveth thee to tell me what is good for me now, for I am now a wanderer with an empty heart, despoiled of my kinsmen” (vol. 12, “Mausala Parva” 271). More importantly, at the end of the poem, though it follows the triumph by a number of years, the P¯an.d.ava brothers and their common wife, Draupad¯ı, abandon all earthly comforts, change their clothes for the bark of trees, and begin to wander in the wilderness “resolved to observe . . . Renunciation” (vol. 12, “Mahaprasthanika Parva” 274). One by one, they drop dead, explicitly due to their own spiritual weaknesses. When Yudhis.t.hira himself enters heaven, he finds the Kauravas there, and the P¯an.d.avas in hell. The situation is quickly changed, and the P¯an.d.avas ascend into heaven, but the

126

The Mind and Its Stories

epilogue of suffering intervened – or, in light of this ultimate upliftment, perhaps I should call it the “epilogue of suffering and transcendence.” The Mwindo Epic of the Nyanga (Congo) is another very good example, though the final conclusion is not so much a spiritual transcendence as a sort of ethical reorientation. Mwindo has been deprived of his rightful place in society. When he finally succeeds in gaining the chieftancy that was his due, he performs an act for which ´ a. The punhe is explicitly punished by the god of lightning, Nkub´ ishment consists in a physical and spiritual journey through the sky, which in this case operates as a sort of hell. In keeping with the wandering motif, Mwindo has “No house!” but “live[s] there in mere nomadism” (137–8). After a year, he is returned to his society, an ethically reformed man. The R¯am¯ayan.a also has an ending of this sort, in this case one illustrating the way in which assertions of transcendence are often a sort of gloss upon something that is much more dismal. The center of the poem is R¯ama’s battle with R¯avan.a to rid the world of demonic terror and to be reunited with his beloved S¯ıt¯a. After his victory, both S¯ıt¯a and R¯ama suffer great misery for complex reasons. S¯ıt¯a eventually calls on the earth to swallow her, which it does. R¯ama plunges into a river. Both are represented as transcendent moments of union with divinity. However, there are clearly suggestions of suicide ¯ in each case. Indeed, this may be one reason for Bhavabhuti’s revision of the ending of the R¯am¯ayan.a in his Uttarar¯amacarita. In that play, ¯ represents S¯ıt¯a as trying unsuccessfully to commit suicide Bhavabhuti by jumping in a river – a combination of the two scenes from V¯alm¯ıki’s original. Gilgamesh concludes with a failed quest for transcendence. Indeed, it does so twice. Gilgamesh and Enkidu succeed in killing Humbaba and the Bull of Heaven. But Enkidu dies, and Gilgamesh is deeply distraught. He undertakes a journey to find Utnapishtim and learn his wisdom. But Gilgamesh fails to fulfill Utnapishtim’s challenge; he loses the plant of rejuvenation; and he returns to his kingdom knowing only that everything is impermanent. In an alternative ending, Enkidu dies and returns to Gilgamesh, describing the existence of a miserable spirit who has no one to remember him in all the living world.

A Problem of Narrative, Empathy, and Ethics

127

There are other cases that suggest problems of this sort as well. The Turkish Book of Dede Korkut ends with a chapter treating civil war and the disintegration of the very society that has been celebrated in the preceding episodes. The Malian epic of Son-Jara points to the great king’s demise – but the poet refuses to go into the details, and even warns, “never try, wretch, to pierce the mystery which Mali hides from you. Do not go and disturb the spirits in their eternal rest. Do not ever go into the dead cities to question the past, for the spirits never forgive. Do not seek to know what is not to be known” (Kouyat´e 84). The warning is general and need not refer to anything specific about Son-Jara’s death. But it is certainly ominous and does not suggest happy events at the end of this story.2 Moreover, this is not all there is to the matter. The epilogue of suffering may appear as a part of the original work, or it may be added in a subsequent revision or “reappropriation” of the work within a literary tradition. For example, one of the most famous revisions of a story from The Tale of the Heike is Zeami’s great No¯ drama, Atsumori. In “one of the most cherished stories in the Heike” (Varley 305), Kumagae, a Minamoto warrior, kills Atsumori, a young man of sixteen or seventeen. Kumagae hesitates before killing him, and deeply regrets his death. In Zeami’s play, Kumagae has left his home and become a monk, wandering and “pray[ing] for the salvation of Atsumori’s soul” (707). He meets the ghost of Atsumori, who runs to kill him with a sword, “‘There is my enemy,’ he cries, and would strike,/But the other is grown gentle/And calling on Buddha’s name/Has obtained salvation for his foe;/So that they shall be re-born together/On one lotus-seat” (712). This is, of course, precisely the structure I have been examining – including the moment of transcendence – though it occurs in a subsequent play, not in the main work. Finally, paradigm dramas sometimes represent an epilogue of suffering without relying on or referring to some commonly familiar 2

It is interesting to note that the oddity of the endings we are discussing is sometimes explicit in the tradition surrounding the work. The epic of Son-Jara seems to suggest that there is or was such an epilogue, only to repudiate it; Johnson notes that Son-Jara’s death is something that “bards . . . keep as a secret” (22). The final chapters of The Tale of the Heike and the R¯am¯ayan.a are disputed (see Varley 306 on the former; on the latter, see Dimock 73 and Narayan 171). The endings appear, then, both appropriate and wrong.

128

The Mind and Its Stories

source work. Sophocl¯es’ Oedipus plays are a clear case. Taking place after Oedipus has achieved social domination, the first play barely sketches the story of Oedipus’ success. Moreover, in that sketch, it focuses on a crime that initiates the epilogue of suffering. While the first play begins the punishment for that crime, the second takes up the exile and concludes with the spiritual elevation, completing the usual structure.3

spiritual transcendence as romance and reconciliation The examples I have been considering suggest that the epilogue of suffering involves ethical and what might be called spiritual or religious issues or dilemmas. These concerns may arise directly, as in the fully elaborated form where the hero evidences remorse or even despair (clearly an ethical feeling in these cases), suffers punishment, and ultimately reaches some transcendent spiritual state. But both ethical and spiritual issues are also suggested in the simpler versions that adopt the victim perspective rather than the victor perspective. In the Iliad, the lament of Andromach¯e implies guilt on the part of the Greeks – a guilt that is made fully manifest in Euripid¯es’ Trojan Women, which takes up the events predicted by her. Moreover, the concluding ceremony is Hector’s funeral, a ritual to benefit his soul. The point is even clearer in The Tale of the Heike. In analyzing the contexts for and types of happiness prototype, I drew on the Sanskrit theory of the goals of human life, referring to the goals of k¯ama and artha, romantic union and social domination. These fit well with aspects of Sanskrit aesthetic theory, and, more importantly, with cross-cultural literary patterns. However, for simplicity, I left aside two other goals specified by the Sanskrit writers: dharma, or ethical duty, and moks.a, or spiritual transcendence. These are, of course, just the concerns that bear on the epilogue of suffering. Before discussing the epilogue of suffering in more detail, and before 3

One could see an element of this in the second and third parts of the Oresteia as well. It seems that the revenge plot is a variation on the heroic structure. Indeed, it appears to be closely related to the epilogue of suffering, as the case of Atsumori suggests. However, the precise relation between these narrative types is complex and requires separate treatment.

A Problem of Narrative, Empathy, and Ethics

129

trying to account for it, it is worth considering the goals of ethics and transcendence in relation to literature more generally. The preceding data and the Sanskrit division of goals indicate that our earlier organization of happiness prototypes into two contexts is incomplete. In addition to the default context of personal happiness and the alternative context of social happiness, there seems to be a third, also alternative context, that of divine or transcendent happiness. These three could be schematized simply as happiness in the home, happiness in the social world, and eternal happiness, happiness beyond mundane life (for example, heaven). Put differently, our lexical entries for outcome emotions seem to involve a distinction along these lines.4 However, the three categories are not equal. Not only is personal, romantic happiness the default case, divine happiness does not appear to have any distinct prototypical eliciting conditions. Like many abstract concepts, we understand transcendent joy only by reference to some more tangible, direct, mundane experience of joy, most often romantic union. In other words, to understand religious happiness, we invoke the default type of happiness – union with the beloved. This is particularly true in literature. Indeed, this accounts for another literary universal. The assimilation of spiritual bliss to requited love not only occurs cross-culturally, it is often elaborated into an allegory of God and humanity as lovers – an allegory found in writings from the Song of Songs to the bhakti literature of Hinduism to the Persian and Arabic ghazal. (We will return briefly to this idea in the next chapter.) Perhaps more importantly for our purposes, the converse holds as well. Not only does romance serve as a model for spiritual relations, spiritual motifs pervade romance. Thus, there are often direct suggestions of spiritual trials and achievements in straightforward romantic tragi-comedies. Most obviously, the reunion of the separated lovers frequently involves imagery of spiritual elevation, as I stressed in Chapter 3. In heroic works, transcendent concerns enter as a literal outcome of the events in the epilogue. Somewhat surprisingly, they are often 4

I suspect that the operation of this division is independent of one’s self-conscious religious beliefs; for example, I am fairly certain that it operates in my own lexicon, despite my self-conscious atheism.

130

The Mind and Its Stories

represented by or associated with some sort of union directly parallel to the conclusion of the romantic plot. Indeed, sometimes there is even a reunion of lover and beloved – as in the R¯am¯ayan.a, where the two virtual suicides bring R¯ama and S¯ıt¯a back together in their divine forms of Vis.n.u and Laks.m¯ı. With greater frequency, the transcendental conclusion involves some other sort of familial reunion – a reunion of brothers in the Mah¯abh¯arata, in Hell, then in Heaven; a (possible) reunion of mother and child in Paradise in The Tale of the Heike; the reunion of Mwindo with his father in the Nyanga epic. Familial reunions in heroic stories are, however, somewhat peculiar as prototypical instances of happiness for they often follow not only separation, but severe antagonism. The reunion of R¯ama and S¯ıt¯a follows their estrangement. The reunion of Mwindo and his father, Sh´eMwindo, follows Sh´e-Mwindo’s attempt to kill Mwindo. The reunion of the P¯an.d.avas too fits here, for Karn.a is included in this heavenly family, while in life he was mistakenly despised by his brothers as low caste. Indeed, at the end, he was killed by one of these brothers. There may even be a hint of this in the case of Kenreimon’in and her son, for, when he died, she neither protected him nor died with him. Other instances take us outside the family to the strange union of enemies as friends – or, indeed, as members of one family. Pr¯ıam and Achill¯es almost become father and son for a few hours when Pr¯ıam visits Achill¯es to retrieve Hector’s body. In keeping with this, he introduces himself by saying, “Remember your own father, most noble prince Achill¯es, an old man like me near the end of his days” (291). Another instance of reunited enemies, in this case with an implication of eventual brotherhood, may be found in Atsumori, when the ghost of Atsumori rushes at Kumagae to kill him, but achieves enlightenment due to Kumagae’s prayers, and the two are united, “re-born together/On one lotus-seat” (712). Gilgamesh presents a reunion of friends, rather than enemies. In keeping with the dismal second and even third endings of this poem, the reunion fails. Enkidu returns from the dead, a mere shade: “The ghost of Enkidu issued from the darkness like a dream./They tried to embrace, to kiss one another./They traded words, groaning at one another” (263). But here too there has been a sort of severing or alienation between the friends. For in this episode (which contradicts earlier events in the

A Problem of Narrative, Empathy, and Ethics

131

poem), Enkidu has died because he ignored the words of Gilgamesh when entering the underworld. In all these cases, the eliciting conditions for transcendent joy are specified, literally or metaphorically, by reference to the eliciting conditions for another, more concrete happiness – the joining, or rejoining, of two or more people. In romantic tragi-comedy, this joining is, of course, romantic. In heroic tragi-comedy, it is often familial. Moreover, even when it is not literally familial, there is often a strong component of familial imagery associated with the reunion. Perhaps more importantly, the reunion at issue in heroic tragi-comedy suggests reparation, forgiveness, the reconciliation of antagonists. Thus, the third context for happiness – spiritual, divine, or religious – is bound up with the romantic and heroic contexts as well. Once again, spiritual joy appears to lack its own prototype eliciting conditions for happiness. In consequence, it borrows them from other contexts. While one prominent cross-cultural prototype for spiritual happiness is romantic union, another is the happiness that results from forgiveness and reconciliation – particularly, a reconciliation that moves from the sharpest conflict to the sharpest union, thus, a familial reconciliation of enemies. This reference to forgiveness and reconciliation returns us to the issue of ethics.

ethical traumas of heroism As noted previously, the Sanskrit writers distinguish four goals of human life. I have just maintained that there are three lexical contexts for happiness, which is to say, three specifications of the goal of happiness. In saying this, obviously I have left aside the fourth goal mentioned previously, dharma or ethical duty. I have done this because ethical goals appear to operate somewhat differently from romantic, social, and spiritual goals. First of all, ethical happiness is a happiness in facts, not in substances, to draw on an idea of Elizabeth Anscombe’s. If I am ethically pleased by something I have done, that is to say that I am pleased about the fact of having done it; whether I enjoyed the act itself is irrelevant. One enjoys the substance of union with one’s beloved or reunion with an estranged family member, but one probably does not enjoy the substance of defending a

132

The Mind and Its Stories

tenure candidate against one’s own professional self-interest, say; nor does one enjoy the substance of the various “negative actions” or restraints that we engage in every day – not deceiving someone for self-advancement, and so on. We enjoy these last two types of action only as facts. We enjoy the fact of having defended someone who deserved defending, despite the prejudices of our colleagues. We enjoy the fact of negative actions, or we avoid the factual pain – the guilt and remorse – that we would have felt had we acted otherwise. In connection with this, it is unlikely that ethics will define primary goals – the crucial eliciting conditions for happiness – in life or in literature. Ethical enjoyment is a sort of second-order enjoyment. We might reasonably expect ethical goals to be second-order – “meta-goals” – as well. Certainly, it is possible to set out solely to be moral. However, more typically, ethical excellence operates, not as a final, motivating aim, but as a sort of norm against which we evaluate our goals and the actions that lead to the achievement of those goals. This is, in any case, what we find in prototypical narratives. It is rare to have a story structured entirely around the goal of achieving ethical purity. However, ethical evaluation is part of most heroic and romantic tragi-comedies. The point is suggested by Aristotle’s theory of hamartia – which adheres closely to his ethical account of the “kinds of injury between man and man” treated in the Nicomachean Ethics (415; 1135b). In this view, ethical evaluation of the protagonist is centrally important to our aesthetic response. For example, in Aristotle’s account of plot, “The change of fortune . . . should come about as the result not of vice, but of some great error or frailty” (47; 1453a). Clearly, we are not dealing with ethical goals here. We are dealing, rather, with a sort of ongoing ethical evaluation of actions guided by other sorts of goal. For example, Macbeth’s tragic hamartia is not a matter of failing in his pursuit of ethical happiness. Rather, it is a matter of acting wrongly in the pursuit of social happiness. V¯alm¯ıki’s R¯am¯ayan.a provides a striking instance, for we are repeatedly faced with moral explanations of R¯ama’s actions, especially those actions that appear suspect. There is a recurring ethical commentary that serves to justify R¯ama’s behavior – not always successfully. Indeed, one thing that makes the R¯am¯ayan.a interesting to nonorthodox readers is the fact that it also includes contradictory

A Problem of Narrative, Empathy, and Ethics

133

commentaries, such as S¯ıt¯a’s famous speech on R¯ama’s tendency toward violence. This speech includes her judgment – so at odds with the lifestyle of a warrior such as R¯ama and with the moral rationalizations surrounding his actions – that the “mind . . . is perverted by . . . the use of weapons” (II.17). Drawing on this account of ethical goals, we can begin to see how the epilogue of suffering might derive from ethical evaluations developed explicitly or implicitly in the course of a heroic plot. But before continuing with this, we need to consider a few more general structural issues relating to narrative. As I have already noted, narrative is produced not simply by the unfolding of emotion prototypes, nor even by their specification. There are other prototypes, as well as schemas and exempla, that contribute, in necessarily complex ways, to the generation of any story. One of the schemas appears to involve the development of conflict in the narrative “middle” or “progression.” This conflict almost always has an ethical component; it is almost invariably involved with moral evaluation. I am not referring here to a conflict between those characters who obey moral precepts and those who do not – the battle between moral heroes and moral villains. Rather, I am referring to a conflict between the hero’s or heroine’s actions, on the one hand, and some putatively overriding moral principle, on the other – including the sort of conflict isolated by Aristotle under the name hamartia. Though not a necessary condition for narrative, it is extremely common in the most prominent narratives across cultures. Indeed, this is fully in keeping with the general narrative principle that the converse of a goal precedes the achievement of the goal (for example, the prototype eliciting conditions for sorrow precede the achievement of the prototype eliciting conditions for happiness). In this case, moral conflict precedes (what should be) morally acceptable success. In romantic plots, the conflict is typically between the lovers, on the one hand, and rules of social hierarchy on the other. Of course, these rules do not operate autonomously. They are asserted by someone representing authority. Again, this authority is often parental, as in the typical New Comedy plot. It may be legal or economic, as in some Kabuki plays where the beloved is legally bound to a brothel and subjected to “the owner’s permission . . . the elders of the Quarter” and “the manager of the Quarter,” as one character explains

134

The Mind and Its Stories

in Chikamatsu’s The Courier for Hell (774). It may be religious, as in the conflict between the dogmatic shaikh and the lovers in the ghazal tradition (see, for example, Ralph Russell 38). In any case, the author of a romantic narrative most often treats the hierarchical authority as invalid. Indeed, the overthrow of this authority in romantic tragicomedy is almost invariably something to celebrate (cf. Frye on New Comedy 163ff.). In keeping with this opposition to hierarchical authority, the lovers in these stories typically come together as equals, at least equals in love.5 In heroic plots, the ethical conflict is in many ways similar. In both cases, there is some sort of opposition between affiliation and personal feeling, on the one hand, and a socially hierarchical ethics on the other. At the same time, however, there is a very significant difference that comes from the fact that the defining ends of the two genres are distinct. While hierarchical principles are, prima facie, incompatible with romantic love, they are of a piece with social domination. Moreover, in heroic plots, it is often the hero him/herself who articulates the hierarchical principle – though it may also be a parental, religious, or related figure (as when the sage N¯arada tells R¯ama that he must kill a low caste person who is engaging in spiritual practices permissible only for upper castes [V¯alm¯ıki III.572]). While the hierarchical principles of authority tend to lose in romantic tragi-comedies, they tend to win in heroic tragi-comedies. In other words, in the romantic plots, affiliation tends to overcome principles of ethical hierarchy; in heroic plots, principles of ethical hierarchy tend to overcome affiliation. This occurs most obviously and most strikingly when the hero is bound by duty to kill a relative or someone like a relative in battle (for example, Sohráb in the Shâhnâme or Atsumori in The Tale of the Heike). There are also less extreme cases, where, for example, the hero must abandon a loved one (for example, S¯ıt¯a in the R¯am¯ayan.a, D¯ıdo¯ in the Aeneid). What is perhaps particularly odd about heroic tales is that the reader’s – and the author’s – sympathy may go either to the hero following the principle, or to the victim of the principled

When this is not the case, when the lovers are not equal in love, the comedy tends to become “problematic,” as in Shakespeare’s All’s Well that Ends Well and Measure for Measure or Terence’s distressing The Mother-in-Law, in which the romantic union is based on a prior rape.

A Problem of Narrative, Empathy, and Ethics

135

action. This sort of affective uncertainty is extremely rare in romantic tales. The importance of these conflicts – especially in heroic works – is not confined to the original stories in which they appear. Subsequent literary reappropriations of paradigm heroic tragi-comedies – such as the Iliad, The Tale of the Heike, the R¯am¯ayan.a, and so on – often focus on just these points of moral tension and distress, what we might call “traumas of heroism.” I have already referred to Atsumori. I could add ¯ Namiki Sosuke’s Chronicle of the Battle of Ichinotani, a very different treatment of the same incident, in which Kumagae sacrifices his own son. As Paula Richman and others have carefully documented, there are numerous revisions of the R¯am¯ayan.a in which hero and villain, victor and vanquished, change places. There are revisions in which ¯ acts of apparent cruelty are justified or softened – as in Bhavabhuti’s Uttarar¯amacharita, where R¯ama’s killing of the unarmed (low-caste) ¯ S´ udra is presented as the latter’s spiritual liberation, for which he ¯ ¯ is deeply grateful. In Bhavabhuti’s version, R¯ama blesses the S´ udra with the following benediction: “May those luminous worlds, full of bliss . . . be thine for ever” (20). In contrast, in V¯alm¯ıki’s R¯am¯ayan.a, ¯ R¯ama kills the S´ udra and the gods congratulate him, declaring “Well ¯ done, well done! . . . thanks to you . . . this S´ udra has been denied heaven” (III.574). Other revisions heighten the conflict, often developing sympathy for the victim – for example, Bh¯asa’s Broken Thighs, which takes up the killing of the Kaurava leader, Duryodhana, at the end of the Mah¯abh¯arata war. On the other hand, these final examples suggest that our formulation of the main issue is not quite right. For instance, R¯ama has no ¯ affective attachment to the S´ udra whom he kills. I have been speaking as if the conflict in these cases is between ethics, on the one hand, and feeling, on the other. But this is not accurate. Indeed, even in cases where there is a conflict between ethics and feeling, that is not necessarily the main problem. Consider, for example, R¯ama’s abandonment of his pregnant wife, S¯ıt¯a. What is deeply disturbing about this is not the fact that he has allowed the hierarchical ethics of kingship to overrule affiliation, which is to say, his personal preferences. Rather, the problem is that he has allowed the hierarchical ethics of kingship to overrule the ethics associated with affiliation. In this case – and, indeed, in most cases – what we find is not a conflict between

136

The Mind and Its Stories

preference and principle, and not even a conflict between different specific principles, but rather a conflict between two different types of ethics. In this way, we could see the ethical conflict in heroic tales as fundamentally different from that in romantic tales. Ethical conflict in the romantic genre concerns the meta-ethical issue of just how far principles of hierarchical authority may ethically be extended to constrain individual preference. Ethical conflict in the heroic genre, in contrast, concerns the meta-ethical issue of which of two moral systems should prevail when they adjure contradictory behaviors. Many ethicians and psychologists have sought to formulate some basic opposition that structures ethical ideas. Most often, these divisions seem inadequate. Part of the reason for this inadequacy, I believe, is that most writers assume our internal ethical principles are rough ethical theories with something like necessary and sufficient conditions. In fact, it seems much more likely that our ethical decisions are, in the first place, prototype based. We judge actions good to the degree that they approximate positive ethical prototypes; bad to the degree that they approximate negative ethical prototypes. There are, no doubt, many different prototypes for ethical and unethical behavior. I should like to distinguish two. One is protecting one’s group – nation, religion, family, and so on. This is an ethics of defiance and bravery. It is an ethics of martyrdom, for one’s country or one’s faith. It is of a piece with a politics of identity. In literature, it is an ethics linked with the sustaining emotion of heroic energy. It is in general imagined as a defiance of those who are threatening – more numerous, hostile, powerful, and so on. The other prototype I should like to isolate is that of comforting and sustaining the miserable. This is an ethics of gentleness and compassion. It is the ethics of the healer or of the Samaritan, in the famous parable. It is of a piece with egalitarian and, so to speak, universal politics (as opposed to identity politics). In literature, it is an ethics associated with the sustaining emotion of love – though parental/filial love more often than romantic love. It is in general imagined as service to those who are weak.6 6

Readers will notice a similarity between the second ethics named here and the “ethics of care” associated with the work of Carol Gilligan. There are several crucial differences, however. First, it is not opposed to an “ethics of justice,” but to an ethics of

A Problem of Narrative, Empathy, and Ethics

137

This general division is closely related to that between ks.atriyadharma and s¯adh¯aran.adharma in Hindu thought. Ks.atriyadharma is the duty of warriors, prominently including the prosecution of war, both defensive and offensive – the “protection of life” and “military occupation,” according to Kaut.ilya’s Artha´sa¯ stra (6). S¯adh¯aran.adharma is the “universal” duty of all persons, primarily the duty of ahim . sa, which is to say, the refusal to cooperate with violence. It is clear that these two types of ethics are sharply opposed. Much early Indic literature develops implicitly – and, at times, even explicitly – out of the conflict between these irreconcilable systems of moral thought. As already noted, I do not believe that these are the only ethical prototypes we have. Rather, I assume that we have a number of ethical prototypes, one or another of which is likely to be triggered in any given context – much as different prototypes for happiness are triggered in different contexts. Moreover, this is only a preliminary formulation of even these prototypes. However, it does seem clear that the conflict between these two prototypes forms the basis for the traumas of heroism. This is unsurprising as the prototype of group defense is the one most obviously relevant in the context of a heroic narrative. Moreover, the prototype of compassion stands in direct conflict with the prototype of group defense – and it too is particularly likely to arise in “heroic” contexts, such as war. All of this leads us back at last to the initial problem, for the conflict between these two ethics provides at least a preliminary account of the anomalous second ending of heroic tragi-comedy. Specifically, the despair and punishment found in the epilogue of suffering are repeatedly focused on or directly related to particular killings or other, usually irreparably harmful acts. These acts were performed in the group protection. Gilligan’s opposition of justice and care is, in my view, a false dichotomy. Second, Gilligan’s division is not prototype-based. Third, my division is not intended to be complete. I believe that there are numerous ethical prototypes that may be triggered by context. Finally, my division is not in any way related to gender divisions, except via stereotypes. One is not a male way of thinking about ethics; one is not a female way of thinking about ethics. I have argued elsewhere that Gilligan’s case for a gender-based ethical division is not plausible even on the basis of the data she cites (“Some Prolegomena” 250–4). Indeed, the politics of identity implicit in Gilligan’s analysis falls into my first category, which Gilligan would presumably label “male.”

138

The Mind and Its Stories

name of an ethics of defense, but they were rendered traumatic by their incompatibility with an ethics of compassion. The epilogue of suffering is a sort of attempted moral rectification, a period of mourning, remorse, and punishment that serves to balance the ethical terrors that preceded. For the hero, the author, and the reader are not unaffected by those conflicts. These traumas can undermine the heroic triumph; they cry out for reparation. In some cases, this relation is straightforward, as when Kumagae becomes a priest as a direct result of his murder of Atsumori. In other cases, the link is less direct. Gilgamesh goes in search of Utnapishtim in response to the death of Enkidu and the consequent realization of his own mortality. Even this is not irrelevant, because the warrior lifestyle chosen by Gilgamesh is ultimately responsible for Enkidu’s death. But there is a more important and direct ethical conflict here. Enkidu dies because he killed Humbaba. In the Hittite version, Humbaba begged to be spared and Enkidu argued against it (Gardner and Maier 11). Indeed, after a touching appeal to Gilgamesh, Humbaba took Gilgamesh “by the hand and led him to his house, so that the heart of Gilgamesh was moved with compassion” (Sandars 24). Enkidu’s death and Gilgamesh’s suffering derive directly from this exemplary moment of ethical conflict between a murderous ethic of defense (“He will bar the mountain road against you . . . this Humbaba must die” [Sandars 24]) and an ethic of “compassion.” Other instances do not involve explicit causal connections, but more indirect links. The epic of Mwindo provides a good example. ´ a, putatively for killing a dragon who Mwindo is punished by Nkub´ ´ a’s friend (137). This does not appear to involve any ethiwas Nkub´ cal element as the dragon was himself killing and eating people. So, either this scene of suffering is nonethical or there is some other reason for Mwindo’s painful sojourn through the sky. Biebuyck argues that the punishment is the result of Mwindo’s own ethical failings, particularly his boastfulness (145). This is to some extent true. However, Mwindo’s hamartia is more importantly a particular instance of the sort of ethical contradiction we have been considering. Earlier in ´ a for aid. Specifically, the poem, Mwindo had himself called on Nkub´ ´ a – that as part of his heroic battle against usurpation, he had Nkub´ is, lightning – descend from the sky and destroy a village, Tubondo,

A Problem of Narrative, Empathy, and Ethics

139

his own natal village. This is in a sense recapitulated after Mwindo kills the dragon. When he returns to the village with the corpse of the dragon, several people say, “he who has killed this one cannot fail to kill one of his relatives” (133) – just what Mwindo did in destroying Tubondo. Mwindo responds by killing these people (134). It ´ a descends to punish Mwindo, this is an seems clear that when Nkub´ ´ a/lightning to destroy indirect retribution for Mwindo’s use of Nkub´ the village, and his subsequent behavior toward those who reminded him of the earlier violence. To say that there is a conflict of prototypes in these cases is not, of course, to say that the conflict is solely, or even primarily, conceptual. It is a conflict that is felt – by the hero, by the author, by the audience. It is a cognitive conflict pervaded by the fear and pride that animate the ethics of group protection and by the compassion that animates the ethics of comforting the miserable. Indeed, it is feeling that makes the conflict traumatic; it is feeling that necessitates the mourning and punishment. But there is a cognitive structure enabling emotional response in these cases. The conflict arises when the antiheroic ethics of aid become salient in the relevant episode, when we feel compassion for the “enemy.” Insofar as the ethics of group protection go unchallenged, even cruelty on the part of the hero will not prove traumatic. ¯ Insofar as we respond to R¯ama’s murder of the unarmed S´ udra boy as a protection of society against the violation of social hierarchy, insofar as the opposed ethical prototype and its associated compassion are not triggered, we will simply accept R¯ama’s act. But what is the difference between this case and the case where the murder does constitute a trauma, the case where compassion and its ethical prototype trouble the fear and pride linked with the ethics of group protection? This brings us to our final topic.

two types of empathy and the morally conﬂicted world In some ways, the answer to the preceding question is easy. An ethics of compassion is triggered by empathy. Concerns of compassion become salient when we can adopt the other person’s point of view, when we can think what it means to be in his/her position. Adopting another person’s point of view is a dangerous thing, at least given

140

The Mind and Its Stories

the combative, usually military goals of heroic narratives. It creates feelings and attachments. It leads us to think of this other person in moral terms, as someone to whom we have obligations, as someone who has rights – in short, as someone like ourselves. As Davis notes, in a number of experimental studies, “instructions to imagine the affective state of a target frequently trigger a process which ends in the offering of help to that target” (145) or, depending upon the experimental circ*mstances, in “inhibit[ing] aggressiveness” (162). The effects are weak, and limited in extent, but still significant. Of course, the mere experience of empathy cannot be all there is to the matter. If I adopt R¯ama’s point of view when he is killing the ¯ S´ udra boy, that empathy will not trigger the prototype of compassion, but reenforce the ethics of group protection. More generally, not all empathy is the same. In every case, empathy is based on some sort of similarity. After all, to empathize with someone is to put oneself in his/her place, and that substitution presupposes something that is shared, something that enables the “fusion of horizons,” as Hans-Georg Gadamer would put it. But there are different types of similarity from which empathy may grow. I shall distinguish two, roughly parallel with the two types of ethics. I shall refer to these as categorial empathy and situational empathy. Each of us has some sort of self-concept. We can conceive of the self-concept as a lexical entry with a hierarchized list of features, including character traits, physical properties, and so on – in short, everything a person thinks is true of him/herself, ordered by importance or centrality to his/her self-understanding. For example, I am bearded, a teacher, living in Connecticut, male, and so on. I see myself much more centrally as male and as a teacher than as bearded or living in Connecticut. Put differently, I can far more readily imagine myself shaving or moving to another state than changing profession or sex. Thus, all these features are included in my lexical self-concept, but “male” and “teacher” are placed higher in the lexical hierarchy than “bearded” and “living in Connecticut.” Each feature in this selfconcept defines some group membership for me. “Man” defines me as a member of the group, “men”; “right-handed” defines me as a member of the group “right-handed people.” Some of these groupdefining features are particularly salient, particularly high in the default hierarchy of the lexical entry, primarily because they have a

A Problem of Narrative, Empathy, and Ethics

141

function in society. Sex is an obvious example – as are race, nationality, ethnicity, and so on. It is well established that we oppose in- and outgroups based on such category-defining traits, and the delimitation of such groups has deeply significant consequences in terms of both evaluative judgment and practical discrimination. When “subjects are asked to allocate rewards (or punishments) between ingroup and outgroup members, they do so in a manner that maximizes the differential between ingroup and outgroup even though this may reduce the absolute benefits to the experimental subjects or even to the ingroup” (Duckitt 68–9). Worse still, when people “are given the opportunity to discriminate” against out-groups, they “show increased self-esteem” (85). It is just this sort of collective self-definition of an in-group, and opposition to an out-group, that provides the basis for the social prototype of happiness as group domination. In this prototype, the in-group is, in effect, a version of oneself – a set of people with whom one shares a definitive self-concept feature (for example, “White,” “Black,” “Hindu,” “Christian,” or whatever). I shall use the phrase “identity category” to refer to a self-concept feature that serves to define such an in-group, which is to say, a group with which one shares “categorial identity.” This sort of group definition undergirds the ethics of group protection, for the group to be protected is, precisely, the in-group (national, religious, or whatever). I shall use the phrase “categorial empathy” to refer to empathy based on this sort of identification. The lexical entries that record our self-concepts include not only a set of group-defining traits, but also an archive of particular memories and their various associated feelings. Indeed, these memories prominently include experiences imbued with unusually strong emotions – positive or negative, happy or sad, wondrous or terrible – for those are the events we find most “memorable.” These memories, too, may trigger identification through a sort of structural mapping. A complex similarity of new events may activate or prime these memories. As discussed in Chapter 2, that activation or priming will, in turn, reproduce the original feeling, in a more or less attenuated form. More exactly, a personal memory involves a structure with some sort of perspectival position – my consciousness, feeling, aims, and so on, at the moment of the memory (my tenth birthday, my wedding, or whatever). Each current situation recalling such a memory also

142

The Mind and Its Stories

involves a perspectival structure. The simplest activation or priming of a memory results from a current, “ego-focused” experience, an experience in which the person in question maintains the same perspectival position in the memory and in the current situation. Doe, who was once run over by a car, suddenly hears a car screech to a halt very close by. The screech of the car serves as perceptual probe to activate elements of his/her procedural memory system – most obviously, response schemas (“jump out of the way”) – and to prime or activate relevant autobiographical memories, along with their associated affects. (Procedural memory records knowledge of how to do things, such as ride a bicycle [see Schacter 17].) In this case, Doe is at the perspectival center of both the memories and current experience. In the cases we wish to understand, this sort of priming or activation results from nonego-focused experiences. Specifically, a triggering of autobiographical memories may involve a shift in the structuring perspective that leads us to put ourselves in the place of someone else and thus results in empathy, associating our own feelings with those of the other person. Suppose Doe hears a car screeching to a halt, but it is at a distance. He/she looks up and sees someone else in the path of the car. The screech is still likely to serve as a probe. It still activates relevant memories, including the memory of being run over, along with its associated feelings. But, in this case, the perspective will be shifted from Doe to the person in danger of being hit by the car – or, as Gadamer would put it, their perspectives are “fused” – leading to empathic fear. Doe tacitly maps his/her memory onto the current situation, but his/her perspectival place in the memory is now taken by the other person, who is in the path of the car. This sort of mapping is no doubt facilitated by the fact that we reconstruct our memories from partial and fragmentary information, which is always to some extent integrated with aspects of current experience, as Schacter has shown (see 8, 40, and 104–13). Indeed, in this way, Gadamer’s term, “fusion,” may be more accurate than “mapping.” In any case, I shall refer to this fusion or mapping with a perspectival shift as “situational” empathy. It is situational empathy that animates the ethics of compassion, and thus it is situational empathy that is crucial to understanding traumas of heroism and the epilogue of suffering.

A Problem of Narrative, Empathy, and Ethics

143

Readers familiar with Turiel’s work on the development of moral judgment in children will immediately see a connection here. Specifically, situational empathy is closely related to Turiel’s view that “the child will judge the presence of a victim as undesirable if he has connected his own experience of pain with that of the victim” and that “experience of [one’s] own pain” makes “the observed experience of the victim aversive” (Blair 7).7 Moreover, this is precisely the sort of empathic identification discussed by Halász with respect to literature. Specifically, Halász explains “contact between the reader and the protagonist’s emotions,” which is to say, a “reader’s empathy,” by reference to “the emotional memory network” and the activation (more precisely, priming) of “relevant events and sources which are embedded in an autobiographical context” (“Effect” 83). The point also fits the analysis of “personal resonance to and engagement in” ´ and Seilman (“Across” 102). As noted in literature by Larsen, Lászlo, Chapter 2, these researchers found that such resonance and engagement were in part a function of “remindings,” specifically remindings of “personally experienced, and therefore self-relevant, events” (102). But there is a problem here. In the case of Doe and the screeching automobile, the generation of empathy seems clear enough. But what about cases of empathy that do not derive from a shared particular experience – as when Jones fears for the person in the path of the screeching automobile, just as Doe does, even though Jones has never been in an automobile accident. Or, stranger still, what about the many literary cases where the work leads to empathic identification, though the reader has never been, say, a deposed monarch or an exiled knight? The first thing to say here is that no two people have ever had precisely the same experiences. Every situation is singular. My experiences may be more or less similar to yours, but they cannot be one with yours. Empathy, then, is not a transfer of feeling across identical experiences. All empathy must be to some degree analogical. It must be a matter of experiences that are similar – in structure, consequence, intensity, and so on. Jones may empathize with Smith because they have both had surgery for colon cancer or because they have both had 7

I am grateful to Tamar Szabo Gendler for drawing my attention to the work of Turiel and Blair.

144

The Mind and Its Stories

surgery for cancer of different sorts or because they both have had surgery of some sort or because they both have been in the hospital or because they both have experienced severe physical pain from illness or have had reason to fear a premature death. How this analogical mapping works is obviously a difficult issue. In most cases, it probably involves multiple memories, interacting with one another and with probes from the current situation in complex ways (all no doubt underwritten by universal and innate propensities, perhaps of the sort discussed by Blair). Clearly, a treatment of these intricacies is beyond the scope of the present chapter. The crucial point here is that this form of empathy is not based on categorial identification. Rather, it is based on some tacit comparison of experience, a structure of memories that is spontaneously mapped onto another person’s experiences in such a way as to identify one’s own perspective in the memory with the other person’s current perspective – or (following Schacter’s research) even a structure of memories that is actually reconstituted from fragments in such a way as to incorporate that other person’s perspective as one tacitly imagines it. But this is still not all there is to the matter, for not all comparable situations trigger mapping and empathy. Sometimes mapping is blocked. Most obviously, it is blocked whenever I find myself so different from the other person – in general, or in a particular case – that a Gadamerian “fusion of horizons” appears impossible or repugnant. Categorial opposition between an in-group and an out-group often functions in this way, constraining empathy across race, religion, nation, and so on, even for people who are not self-consciously bigoted. Moral evaluation can serve this blocking function as well. We tend to have compassion for those who are “more sinned against than sinning,” but not for those who “receive their just deserts.” Indeed, part of the meaning of these phrases concerns empathy; the former marks its referents as deserving of empathy, the latter as undeserving. The point was indicated by Aristotle, in the notion of hamartia, for he insisted that the hero cannot be wholly evil because the downfall of such a character does not inspire pity (45; 1453a). More generally, no trauma of heroism will occur insofar as we judge the villain so distant from ourselves – in categorial identity, in moral character, or whatever – that any situational identification is blocked, presumably because any even tacit imagination of his/her perspective is blocked.

A Problem of Narrative, Empathy, and Ethics

145

Saying this, however, makes our initial dilemma appear only worse; heroic tragi-comedy now seems still more anomalous. The epilogue of suffering does appear to be a sort of reparation for the traumas of heroism. Traumas of heroism, in turn, appear to result from the conflict between the ethics of group protection and the ethics of aiding the miserable, a conflict that arises in the course of the heroic narrative along with the development of compassion. Finally, compassion is triggered by situational empathy for suffering. But why is situational empathy there in the first place, given that it almost invariably concerns the enemy? Our empathy with such figures should be blocked both by categorial identity and by their moral depravity as villains. Unsurprisingly, things do not work out that simply. The sorts of self/other opposition that block empathy – including categorial and moral irreconcilability – are typically based on a reification of the other person. To the extent that this reification is undermined, the empathy-blocking opposition may be undermined as well. More exactly, the empathic adoption of someone else’s perspective involves, among other things, a shift from what one might call “objective mind” to “subjective mind.” What I mean here is that we may tacitly conceive of the human mind as a sort of ghostly object or thing, defined, in effect, by a list of specifiable properties, mechanisms, and so on. In doing cognitive science or other sorts of psychology, we often necessarily treat the generic human mind in this way. At times, in daily life, we also treat individual minds this way – typically, the minds of people we do not know well. The alternative to this reified view of mind is, roughly, phenomenological. We may understand the mind as pure subjectivity, ongoing experience, idea, feeling, and memory. If objective mind is quasi-spatial, subjective mind is purely temporal. If objective mind manifests a limited set of traits and procedures, subjective mind is either almost traitless – a continual, flexible negotiation with changing circ*mstances (see Holland et al., 222–4) – or it has a limitless variety of traits, all mixed-up and contradictory, good and bad, cheerful and melancholy, smart and stupid (see Davis 97–8). For ourselves, we are always subjective mind. Others become subjective minds for us as we come to know them, as we come to see them not as mechanically enacting fixed character traits, but as engaging mutable conditions in variable and

146

The Mind and Its Stories

open ways, or as combining an array of irreconcilably contradictory traits. Like all character predicates, ethical predicates, especially negative ethical predicates – “bad,” “disloyal,” “selfish,” and so on – apply most readily to objective mind (see Davis 97). The more subjective our understanding of someone, the more difficult it is to detach ethical predicates, especially negative predicates, from his/her particular acts and apply them to his/her “character.” In literary terms, the more we come to know someone, the more we understand and respond to him/her as subjective rather than objective mind, the harder it is for us to view him or her as a villain. A pure villain – or, for that matter, a pure hero – is almost necessarily a simple list of objective traits. Familiarity pushes toward the middle, setting limits on both idealization and vilification, elevating the bad guys, and diminishing the good guys. A parallel point holds for group identities. The more we come to know someone, the more difficult it is for us to reduce him/her to an instance of some collective category – racial, national, and so on. As Holland et al., explain, “social-category . . . values are typically suppressed in the face of individuating evidence” (221). Literature, to create its effects in the first place, must develop empathy with the hero. But, in developing the subjectivity of the hero, a work not only creates the possibility of empathy – it may simultaneously limit the possibilities for heroism. More significantly, when an author elaborates on the action or condition of an antagonist, a villain, or an enemy, he/she almost necessarily begins to develop a subjectivity for that antagonist, thus dulling the contrast with the hero. In this way, the sorts of opposition that would ordinarily block empathy tend to be undermined through ordinary literary exposition. Villainy in particular is difficult to sustain. Each new discussion of the villain renders him or her more subjective, and thus less villainous. Georg Luk´acs’s well-known distinction between reifying reportage and “dialectical” portrayal fits here. An author may report objective facts about a character – hero or villain. But an author might equally develop the subjectivity of a character, “portray” him/her, painting “a picture with which we can empathize” (57). The point here is that, at least in paradigm heroic plots, what begins as reportage regarding the enemy often develops into an empathy-triggering portrayal, perhaps even against the author’s intent. We might refer to this as

A Problem of Narrative, Empathy, and Ethics

147

the spontaneous tendency to subjectivize, and thus undermine, villainy. Many heroic tragi-comedies provide examples. To take one, H. Paul Varley remarks that, “In the second half of the Heike the Taira, heretofore despised, increasingly elicit our sympathy, as they suffer defeat after defeat at the hands of the Minamoto” (301). But our response in these cases is not solely a matter of greater fullness and detail in characterization of the enemy. Varley stresses the suffering of the Taira in explaining the reader’s change in attitude, and he is right to do so. Given the preceding account of ethical prototypes, it is to be expected that the portrayal of suffering in particular should tend to activate an ethics of compassion. The ethics of compassion apply most naturally to situations in which one is faced with a suffering and debilitated person – or, more generally, someone who is weaker, lesser in strength, weaponry, health, and means. When the Other is represented as a threat, that representation tends to activate our prototype ethics of defense. But when the Other is represented as weak and miserable, that tends to activate our prototype ethics of compassion. Thus, it is precisely in those stories where the hero triumphs that we are likely to feel the greatest conflict – for it is in those stories that the “villain” experiences the greatest and most acute suffering. This is almost a necessary result of the genre itself, of the elaboration of the villain’s acts and condition, and of his/her pain. The point is illustrated with near allegorical precision in a work I have not discussed in this context, a work that is not even primarily ¯ This poem includes heroic, but romantic: Niz.a¯ m¯ı’s Layl¯a and Majnun. ¯ recruits an army to battle Layl¯a’s one heroic section in which Majnun tribe, so that he may be united with his beloved. But as soon as the ¯ regrets it deeply. Niz.a¯ m¯ı describes Majnun’s ¯ battle begins, Majnun condition in words that (at least in the English translation) could apply equally to many writers: “While each warrior thought of nothing but to kill the enemy and to defend himself, the poet was sharing the sufferings of both sides” (58). The conflict between ethical systems is obvious, and unmitigated. It need not be that the poet is a better person or even necessarily more empathic in general. But, in imagining, detailing, elaborating the conflicts of heroic tragi-comedy, in portraying the antagonists and their condition, he/she almost inevitably comes to feel the humanity of all those involved – including

148

The Mind and Its Stories

those enemies who suffer irremediable pain at the hands of the hero. But this too is not all. In addition to this unintended development of subjectivization and compassion, many authors of heroic tragi-comedies, including authors of paradigm canonical texts, seem to have set out to cultivate situational empathy with the enemy – empathy that runs counter to the categorial empathy an original reader might have been expected to share with the hero. A work such as Euripid¯es’ Trojan Women is a particularly obvious case. In this play, Euripid¯es systematically cultivates situational empathy with the Trojans against what he could assume to be his audience’s spontaneous inclination to identify categorially with the Greeks. Indeed, Euripid¯es did this for particular political purposes – to condemn the brutal treatment of M¯elos in the then-ongoing Peloponnesian War, and to protest the Sicilian expedition and the jingoistic enthusiasm of his fellow Athenians (see Hadas 173), an enthusiasm based, of course, on categorial identification and profoundly devoid of compassion. But Euripid¯es is far from the only case. Here I should like to return to one final oddity found in the epilogue of suffering – the “transcendental” conclusion with its motif of familial reunion. As I noted, this reunion responds to a prior severing and opposition which is part of the trauma of heroism. Why is this opposition and reunion so consistently familial? This pattern results I believe, at least in part, from authors’ attempts to cultivate empathy. I do not mean that authors self-consciously decide to create empathy, make a list of possible techniques, and so on. Rather, I take it that the authors set out to achieve a certain sort of effect, an effect which includes empathy, whether they explicitly recognize this or not. As they try different techniques, some seem to work better than others. One frequently successful technique is the identification of hero and enemy as members of one family, literally or metaphorically, for this “familializing” of enemies runs strongly contrary to any strict categorial division that might block empathy. Striking instances occur when the trauma involves the hero killing a close relative. To protect the kingdom, Rost´am kills his son Sohr´ab. Mwindo calls down lightning on his natal village, killing everyone, presumably including his own siblings (90–2). The Book of Dede

A Problem of Narrative, Empathy, and Ethics

149

Korkut ends with Kazan’s killing of his uncle. The Mah¯abh¯arata is a great war between cousins, and the horror of this familial slaughter is a continual theme in the poem, as when the P¯an.d.avas discover that, in killing Karn.a, they have killed their eldest brother (see Vy¯asa, vol. 7, “Stree Parva” 45–6). This structure, opposing close relatives across lines of battle, tends to limit the possibilities for dichotomizing hero and enemy, and tends to enhance the subjectivization of that enemy by stressing his/her similarity with the hero. It also enhances our empathy with the guilt and remorse of the hero in the epilogue of suffering, for we all know the particular pain of hurting those to whom we are closely related – brothers, sisters, parents. Finally, in keeping with the preceding point, this familial structure sharpens the moral conflict that generates the epilogue of suffering, for we generally believe that people have stronger moral obligations to their family than to strangers (as when we believe that we have a greater obligation to feed our own children than to feed other people’s children [see Sommers]). The empathic function is even more obvious when the connection is metaphorical or analogical, rather than literal and direct, for in these cases the empathic aim is often explicit. Consider, for example, Pr¯ıam’s appeal to Achill¯es when they first meet after Hector’s death: “Remember your own father, most noble prince Achilles, an old man like me near the end of his days” (291). This is straightforwardly an appeal for empathy. It is an appeal from one character to another, but it operates also on the reader. Indeed, when I read this passage recently, I thought of my own father. The situational similarity invoked by Pr¯ıam served to subjectivize Pr¯ıam, not only for Achill¯es, but for me. Or consider the encounter of Gilgamesh and Humbaba. Humbaba explains “I have never known a mother, no, nor a father who reared me” and asks Gilgamesh to be his “lord.” Then he takes Gilgamesh “by the hand” and leads him “to his house” – what should be a familial place. Gilgamesh is thus assimilated to Humbaba’s father. As a result, Gilgamesh is “moved with compassion” (Sandars 24). Perhaps an even more striking case may be found in The Tale of the Heike. When Kumagae sees Atsumori, he thinks of his own son, and transfers the feeling to Atsumori’s parents, a transferal that makes the situational empathy explicit: “When I think of how I grieved when Kojiro¯ suffered a minor wound, it is easy to imagine the sorrow of

150

The Mind and Its Stories

this young lord’s father if he were to hear that the boy had been slain” (317).

conclusion In sum, heroic tragi-comedies often include an “epilogue of suffering” in which the story continues beyond its expected conclusion. This epilogue is focused either on the misery of those who are vanquished or on the anguish of the victorious hero, who surrenders the domination he/she has won, suffering remorse or undergoing some punishment. This anomalous “second ending,” following the heroic triumph, involves ethical concerns. Specifically, heroic plots regularly manifest a conflict between two ethical prototypes, one based on group protection, one on individual compassion. Due to the martial circ*mstances of the story, the hero most often acts in accordance with the former, violating the latter. The epilogue of suffering is, in effect, a reparation for that choice. This reparation is not necessitated simply by abstract ethical conflict, however. Rather, the conflict involves us empathically. Here too a distinction is in order, in this case between “categorial empathy,” which is based on shared group identities (for example, nation or race), and “situational empathy,” which is based on shared experiences, especially experiences of suffering. Ethical and empathic conflicts are unusually likely to arise in heroic tragi-comedy because the violent group antagonism is apt to trigger categorial and defense-oriented responses, while the violent resolution of that antagonism, including the defeat of the enemy, may trigger situational and compassionate responses. This is particularly true if the author develops a fuller portrayal of the enemy, for that tends to humanize the enemy and thereby to draw our empathy when that enemy suffers. Thus, the anomalies outlined at the beginning of this chapter seem quite comprehensible, given a fuller understanding of empathy, ethics, and the development of literary plot and character. There is, however, one further implication of this analysis that it is important to draw out. The preceding discussion indicates that there is a contradiction, or potential contradiction, at the center of heroic narrative itself, because there is a contradiction implicit in the prototype of social happiness. Horror, it seems, is an inevitable concomitant of

A Problem of Narrative, Empathy, and Ethics

151

individual or group domination. The traumas of heroism are akin to killing one’s own son or father, for they are a matter of killing someone’s son or father. But they are not the evil product of a particular reified character. They result, rather, from a general condition. Specifically, the traumas of heroism arise in a social world that allows, indeed drives, inexorably toward conflicts between our two sorts of ethics: a social world created whenever powerful men and women pursue a prototype of happiness as individual or group domination. In “Moral Dilemmas and Consistency,” Ruth Marcus argues that one of our most important ethical duties is “to conduct our lives and arrange our institutions so as to minimize predicaments of moral conflict” (121). The traumas of heroism and the epilogue of suffering show over and over again that the pursuit of individual and group domination violates that duty. In short, the apparent anomalies of heroic tragi-comedy are not anomalous at all, but the direct result of that genre’s origin in a faulty prototype for social happiness, a prototype that creates a morally conflicted world.

5 Extending the Theory Emotion Prototypes, Narrative Junctures, and Lyric Poetry

The literary importance of the emotive and plot structures and principles discussed in the preceding chapters would appear to be confined to explicitly narrative literature – stories, plays, novels, epics. This is hardly a narrow scope. Narrative literature is vast and there is no reason to expect that an account of explicit narratives could be extended to another area of literature. On the other hand, such an extension would give greater force to the initial account and would further its research program in valuable ways. In the following pages, I shall argue that we can in fact extend these structures and principles to lyric poetry. Indeed, in the conclusion to this chapter, I shall argue that the preceding theory can be extended even beyond literature to aspects of our ordinary lives, such as religious belief, and that the narrative study of lyric poetry helps us to see this.

lyric poems as implicit plots: the narrative hypothesis More exactly, we tend to think of narrative and nonnarrative verbal art as sharply distinct. However, the structure of plots and the universal features of lyric poetry indicate that this is a misconception. In particular, lyric poems are commonly set off against narrative, opposed to it. For example, one standard handbook defines a lyric poem as “A poem, brief and discontinuous, emphasizing sound and pictoral imagery rather than narrative” (Frye, Baker, and Perkins 268). 152

Emotion, Narrative Junctures, Lyric Poetry

153

However, I shall argue that lyric poems are not separate from narrative. Rather, lyric poems are elaborations of junctural moments in narratives and are governed by the same general cognitive principles as stories. These narratives are, of course, implicit and necessarily to a degree vague. However, they are crucial to the sense and consequence of lyric poems. Put differently, a lyric poem prototypically involves a focus on a particular emotion (see, for example, Padgett 111), paired with a particular event or brief sequence of events. But a lyric poem is not thereby monadically encapsulated in a moment. Rather, the event in a lyric poem draws its meaning and affective force from its tacit relation to a prototype for an outcome emotion and its location within a narrative structure implied by that prototype. More exactly, prototypical lyric poems are tacitly located at junctural moments of heroic or romantic tragi-comedies and imply the emotion prototypes for those genres.1 I shall refer to this as the “narrative hypothesis.” My claim is that this relation of lyric poems to prototypical narratives and emotion prototypes is an absolute universal. As with 1

The general relation between lyric poetry and narrative has not gone unremarked. A few critics have commented explicitly, if very broadly, on the emplottedness of poems. For example, Voigt notes that “Stephen Dobyns has said that every lyric poem implies a narrative.” However, the point has hardly been developed, or it has been developed differently. Thus Voigt explains Dobyns’ idea in the following terms: “What he means is a sequence of past events, left out of the poem, that brought the speaker to the present intensified moment in the poem. Assuming the speaker is the poet, Dobyns also means autobiography” (725). The connection here is obviously a very general one, little more than a matter of causal precedence, which does not necessarily merit the title “narrative” and which is, in any case, not confined to prototype structures. Some other writers have suggested a link of this sort through more historically and textually particularized studies. For example, Sylvia Huot has examined the complexities involved in the use of motifs from Troubadour lyric poetry in Old French narratives. Though this “transposition of lyric into narrative” (263) is different from what we are considering, it is not unrelated. The fact that such transpositions were common suggests again that there was an implicit link between lyric poetry and narrative to begin with, a link that enabled authors to undertake the transpositions. Finally, a few writers have drawn broader connections that could be taken to imply a lyric poetry/narrative link. For example, Sarbin has emphasized the embeddedness of all imaginings in implicit narratives, and lyric poems are certainly imaginings. On the other hand, the precise nature of Sarbin’s claim is somewhat different from mine. In particular, Sarbin’s use of “narrative” is both broader (“a script with a beginning, a middle, and an end,” 21) and narrower (for example, in stressing a person’s “self-narrative” as, say, a “survivor”).

154

The Mind and Its Stories

explicit paradigm narratives, this is not to say that every individual work is implicitly located within a heroic or romantic plot. Rather, it is to say that, in every literary tradition, lyric poems are predominantly of this sort.2 Once stated, the general point about the relation between lyric poems and implicit narratives is almost obvious, at least in many cases. Consider a poem by the eighth-century Chinese poet, Wang Wei, “Written Crossing the Yellow River to Ch’ing-ho” (Mack et al. 1,304). The poem begins with an enigmatic reference, “The boat set sail.” The presumption of familiarity here (as if we know just which boat is in question) signals precisely an implicit narrative context. The speaker goes on to describe how the water extends to the horizon. Then, “suddenly,” he sees “a district capital.” He sees the market and remarks on what goods he can discern. Then he turns “back toward my homeland.” He can see nothing but water extending to the horizon. The poem is complex. It is highly allusive, as is standard in Chinese poetry, and involves subtle Buddhist motifs. However, even without understanding any of this, it is clear that the poem is a single, junctural moment of beginning exile in an implicit narrative of union, separation, and perhaps reunion. The speaker sees clearly the homes of the capital, thousands of them – but none of them is his own home. Looking back in the direction of his home, he can see nothing but endless water. The poem thus recounts an isolated moment of alienation that has its effect by integration into an unstated narrative. Indeed, the poem foregrounds one common aspect of lyric poetry that can give it a unique impact and force, distinct from most elaborated narratives: irresolution. The speaker of the poem is, in effect, trapped in an isolated moment with no clear future. The narrative structure itself implies that there are two and only two possible outcomes: success or failure in returning home. But we have no indication of which will be achieved. 2

My suspicion is that nearly all paradigm lyric poems are implicitly narrative. If forced to give a rough estimate, I would say that two-thirds to three-quarters of these poems imply one of the three prototype genres (including sacrificial tragicomedy, to be discussed in Chapter 6). That would be my rough estimate for the percentage of explicit paradigm narratives in these genres as well.

Emotion, Narrative Junctures, Lyric Poetry

155

Of course, not all lyric poems involve irresolution. Indeed, many poems focus precisely on prototype outcomes – a point to which we shall return. Others imply a resolution that has not yet been achieved. Obvious cases of the last sort would include the psalms of lament in the Old Testament. These regularly involve an implication that “the Lord has heard the prayer” (Murphy 628) and thus that the narrative conclusion will be positive. In other words, they are located in an unspoken comic narrative – if a comic narrative that has not yet been realized and thus has not yet reached resolution. Indeed, the psalms fit the narrative hypothesis very well in other ways also. The standard categorization of the psalms into psalms of lament, thanksgiving, and praise, or their more recent categorization into psalms of “orientation,” “disorientation,” and “reorientation,” indicates that the narrative aura surrounding these poems has been implicitly recognized by biblical scholars. These categories are, clearly, junctural moments in narratives. Moreover, the relation of the psalms to prototype narratives is often straightforward. Consider, for example, one of the best known psalms, number 137, “Beside the streams of Babylon/we sat and wept” ( Jerusalem 920). Like the poem by Wang Wei, this is an elaboration upon the beginning of exile. Here, however, it is clear that the exile is the result of a collective military defeat. It is the loss of group domination, and it points toward the reestablishment of that domination at a future time. Two further features of this poem are particularly significant in this context. The first is that it clearly assimilates the relation between the exile and his home or nation to the relation between a lover and his beloved – for the exile takes a lover’s vows with respect to Jerusalem, swearing that he will never forget her and will always count her as his greatest joy. At a literal level, then, the poem is embedded in an implicit heroic narrative. But it specifies the heroic juncture by reference to the other (romantic) prototype narrative. Second, the ultimate group triumph toward which the poem points is a triumph that explicitly includes the sort of violence that entails an epilogue of suffering. Specifically, the poem ends with the distressing image of the massacre of innocents: “a blessing on him who takes and dashes/your babies against the rock!” One could interpret the poet as simply heartless. But one could also interpret the poem as part of a tacit story in which the heroic triumph is, as usual, tarnished by cruelty. One could,

156

The Mind and Its Stories

in other words, see the poem as much more ambivalent, or potentially much more ambivalent – much more in keeping with a narrative that includes a scene of sorrow after the final triumph.

lyric poems in narratives The idea that lyric poems are embedded in narratives has been implicitly recognized in some traditions through the ways lyric poems have been joined and organized. The development of sonnet sequences in European literature is a case in point, for such sequences often structure lyric poems into loosely plotted series of events – most often romantic, as with Sidney’s Astrophel and Stella, Spenser’s Amoretti (combined with the marital poem, Epithalamion), or Browning’s Sonnets from the Portuguese. Some Sanskrit mah¯ak¯avyas are of this sort as well. Though “mah¯ak¯avya” is often translated as “epic,” these works are more akin to sequences of lyric poems with narrative bridges, as many writers have pointed out. For example, Dimock, Gerow, Naim, Ramanujan, Roadarmel, and van Buitenen discuss the mah¯ak¯avya in the context of lyric poetry, not epic, and emphasize that it is really a sequence of lyric poems (see 152–5; Dimock et al. somewhat overstate their case, primarily because they fail to see that lyric poetry and narrative are not opposed or mutually exclusive). The renowned G¯ıtagovinda of Jayadeva is a case in point, for it is largely a series of lyric poems treating moments in a story of separated lovers. Sequences of this sort are not the only places where we find lyric poems incorporated explicitly into larger narratives. Such incorporation is perhaps particularly obvious in drama. Elite instances may be found in Sanskrit drama, which regularly involves the utterance of verses at key junctural moments in the plot – such as partings, re´ unions, and so on. When Sakuntal¯ a leaves home to join her husband, her adoptive father speaks the following lines, which constitute a perfectly good free-standing lyric poem: “When I see the grains of rice/sprout from offerings you made/at the door of your hut,/how shall I calm my sorrow!” (K¯alid¯asa, Theater 132). Before sending his ¯ brother to take S¯ıt¯a away and abandon her, Bhavabhuti’s R¯ama recites verses to the sleeping S¯ıt¯a, telling her that she should abandon him for he has been made an untouchable by his “atrocious deeds”; he is

Emotion, Narrative Junctures, Lyric Poetry

157

like a poisonous tree and she is like a vine that has used it for support, though it will only prove deadly for her (14). The poem clearly elaborates on the sorrow and guilt felt by R¯ama at this moment of evaluation, a moment that immediately precedes his separation from his beloved and her exile due to a conflict with society. The first act of Bh¯asa’s Vision of V¯asavadatt¯a establishes the heroine’s separation from her husband in conjunction with her brief entrance into a hermitage. The stop at the hermitage marks a clear point for V¯asavadatt¯a’s and our reflection on her loss of romantic union, her exile. The act ends with a poem about the hermitage, explaining how “The birds have returned to their nests” – precisely the sort of return home that is now forbidden to V¯asavadatt¯a – while “The hermits have plunged in the stream,” and the smoke of the sacred fires “is spreading in the penance-grove” (16). The junctural placement of these verses is, of course, crucial. Again, by the narrative hypothesis lyric poems, appearing alone, mark and elaborate junctural moments in implicit plots, predominantly romantic and heroic plots. In the cases just cited, the lyric verses perform the same function, but the encompassing story is explicit. Japanese No¯ dramas also incorporate lyric poetry at such crucial moments. Zeami’s plays, though short, often include many juncturally embedded lyric poems. In Lady Han, the main character’s distress at being separated from her beloved is regularly expressed through the quotation of poems from the Kokinshu¯ (905 c.e.) and other important early Japanese anthologies. The Reed Cutter is a particularly striking case, for Zeami actually developed the play out of the ambiguous narrative implicit in two poems from a tenth-century an¯ u¯ (see Keene, Twenty Plays 148). thology, the Shuish Of course, the practice is not confined to such elite theaters. It is equally obvious in Broadway and Hollywood musicals, where the songs are lyric poems that punctuate the larger narrative. The same point holds for the even more numerous musicals of Indian cinema. Nor is this use of lyric poems confined to drama. A particularly striking case of the incorporation of lyric poems into a nondramatic narrative is found in Gurg¯an¯ı’s V¯ıs and R¯am¯ın. This important Persian work is itself in verse, but it incorporates “Songs of R¯am¯ın,” which are “easily distinguishable” due to their distinctive use of rhyme (Morrison xii). A good example of this is when R¯am¯ın is lying beside

158

The Mind and Its Stories

his forbidden beloved, V¯ıs. He “dreads the break of day and sings of the night” (xiii), a clear junctural moment. Similarly, a lyric poem enters when R¯am¯ın weeps and bewails his separation from V¯ıs (Gurg¯an¯ı 121); when he has just set off on a journey to meet her and is filled with hope – proclaiming, for example, that “The way to union with you will be short for me” (164); and when he finally sees her face again (171). More salient instances occur in prose works. As Lewis points out, the stories of The Book of Dede Korkut “are in prose interspersed with rhythmic, alliterative, and assonant or rhyming passages” (14), which is to say, lyric poems – or, rather, partial lyric poems, for not all of them could be separated successfully from their context. A good example occurs on the second page of the first story. A husband calls to his wife, saying, “Come here, luck of my head, throne of my house,/Like a cypress when you go out walking” (28). In the following five lines, the speaker employs common conventions of love poetry in describing his beloved. It is a standard romantic poem, integrated into a narrative – just at the junctural point when the speaker is considering the fact that he has no children. Another good example occurs a few pages later in the praise poem concerning a character who has just proven his strength and prowess (31). Subsequent stories include further types of lyric poetry, such as religious poems of worship (137–8). The practice is not absent from European literature. For example, as Maureen Boulton has discussed, “lyric insertion” – the incorporation of lyric poems into other works – is “a technique present in a substantial corpus of medieval French texts” (xiii). These poems produce a sort of “lyric ‘pause’” – again, a juncture – in the encompassing “narrative movement” (xiv). Though not common in all European Medieval narratives, this technique is also present in many Spanish works (1n.1). Chinese tradition makes extensive use of lyric poems in prose narratives, often via the convention of having characters compose poems at key moments (a more mimetically plausible version of the common practice of having characters break into song or simply speak poems). Alternatively, Chinese historical fiction incorporates poems written or putatively written by contemporary poets involved in the events or by later poets commemorating the relevant historical occurrences. For example, early in Three Kingdoms, an uprising is marked

Emotion, Narrative Junctures, Lyric Poetry

159

by a “seditious song” – a four-line lyric poem. Subsequently, two of the main heroes are celebrated by a “poet of later times” with a short verse (Luo 10), as is the main hero, Liu Xuande (11). Not all lyric poems are explained in these ways. Some are just inserted by the author. For example, at the end of the first chapter, an interpolated lyric poem addresses the reader’s evaluation of narrative movement. It tells us that Zhang Fei will “pay out every ingrate what he’s due” just before asking the question, “Did Zhang Fei kill the Imperial Corps commander? Read on” (14; see also Brewitt-Taylor, vol. I, 10). Here we find an explicitly marked case of a narrative juncture. The Story of the Stone includes a number of authorial insertions of this sort. For example, in one case, we encounter one of the two main lovers in the novel. She is weeping and the author inserts a poem that serves to elaborate on this junctural moment. He introduces the verses simply by writing, “As the poet says” (Cao and Gao, vol. 4, 166). Of course, in keeping with standard Chinese literary practices, characters also compose poems at such critical points. A good example may be found later in the same volume. After his dear friend, Skybright, passes away, Bao-yu (Magic Jade), feeling mournful, composes a poem addressed to her. We read the poem, then he burns it (201–2), the smoke from the paper rising to the heavens. More importantly, this novel makes explicit the connection between lyric poetry and prose narrative. Early in the first chapter, one character complains about the stereotyped romances that evidently plagued the literary market in eighteenth-century China. The situation is reminiscent of North America today, with one difference – the prominence of lyric poems in these narratives: “The trouble with this last kind of romance is that it only gets written in the first place because the author requires a framework in which to show off his love-poems. He goes about constructing this framework quite mechanically, beginning with the names of his pair of young lovers and invariably adding a third character . . . to make mischief between them” (50). Obviously this did not lead Cao or Gao to repudiate the integration of lyric poems into prose narratives. Quite the contrary. In fact, this novel, like Three Kingdoms, concludes with a lyric poem (Cao and Gao, vol. 5, 376; Brewitt-Taylor, vol. II, 622–3). Note that by our definition, the ending too is a juncture, for it is almost necessarily a point of reflection and evaluation.

160

The Mind and Its Stories

Many Japanese monogatari or “tales,” such as the Tale of Genji, also integrate lyric poems directly into the narrative. Here too, characters frequently compose poems at key junctures of particular emotional intensity. For example, in the ninth-century Tale of Ise, we find Narihira going out to inspect his estates and seeing two beautiful sisters. The passion inspired by the sight of sexual beauty is a standard junctural moment in a romantic narrative. Narihira immediately writes a poem and sends it to the sisters (Keene, Anthology 64). We find other examples of this sort in the Tosa Diary from the same period, the Mirror of Increase, an historical romance from the fourteenth cen¯ tury, and so on. Some striking cases are found in the works of Basho, the seventeenth-century master of the haiku. The Narrow Road of Oku, for example, begins with Basho¯ setting out on a journey, a sort of voluntary exile “that was perhaps to separate us forever.” To mark this, he composes a poem in which birds and fish weep at the end of spring. “I set out after composing this poem” (Keene, Anthology 348), he explains. Finally, it is not only authors who integrate lyric poems into larger stories. At least in Japan, editors sometimes took on this task as well. The love poems in the Kokinshu¯ are integrated “into a longer narrative sequence.” Specifically, the compilers “fashion[ed] a single narrative out of the poems by many different authors.” In consequence, “the poems chronicle the progress of a love affair” (Danly 2,076).

heroism and romance It is crucial, of course, that the Kokinshu¯ poems are integrated, not into any old narrative, but specifically into a narrative of love. My hypothesis is not simply that paradigm lyric poems are implicitly narrative – though that is significant in itself – but that they are, most often, tacitly a part of prototypical narratives. Thus, if the narrative hypothesis is correct, then, cross-culturally, the largest number of lyric poems – including, of course, the poems we consider most prototypical – should fall into the romantic and heroic categories, especially the former as that is, again, more basic and pervasive. This does indeed appear to be the case, from the time of the earliest poems written for emotive (rather than, say, magical) purposes.

Emotion, Narrative Junctures, Lyric Poetry

161

The European tradition of lyric poetry extends back to two great canonical figures who in effect embody this division: Pindar, the au¯ the thor of odes that are explicitly heroic in orientation, and Sappho, author of renowned romantic poems. The prominence of romantic poetry in the West hardly needs the be stressed. Interestingly, the heroic genre was, if anything more prominent in ancient Greek writing, at least in what has been preserved. Our earliest extant poet, Archilochos, was a soldier and proudly proclaims himself such in the first fragment. Kall¯ınos wrote that “honor and glory bedeck the man/who fights for his land” (Fowler 67). Tyrtaios tells us, “To die is a fine thing when a noble man falls/fighting in the forefront of battle on behalf/of his fatherland” (Fowler 77). This generic op¯ best known position itself is marked nicely by some of Sappho’s lines, from fragment 16. She points out that some people find a “host of horsem*n,” “soldiery,” or “a fleet of ships” to be the most beautiful thing on earth. But, for her, “it’s whomever one loves” (Fowler 131). Even more strikingly, classical Tamil poetry of South India, dating from the first three centuries of the common era, is explicitly divided into two sorts: first, “akam, or interior, poems that view life from inside the family and concern the love between man and woman”; second, “puram, or exterior, poems that view life from outside the family and concern such topics as kings, heroism in battle,” and so on (Hart 3). Just north of the Tamil writers, Sanskrit lyric poets focused their poetic energies upon two junctural moments of the romantic tragi-comedy, writing poems of sexual longing and union. Merwin’s and Masson’s The Peaco*ck’s Egg collects Sanskrit poems from 500 b.c.e to 1,000 c.e. treating these junctures. The tradition extends well beyond that, from ˙ such works as Bilhana’s well known Fantasies of a Love Thief (eleventhcentury c.e.), to later works in Hindi and other Indic languages, such as Bih¯ar¯ı’s Satasa¯ı (seventeenth century). In pre-Islamic Arabic literature, there was a considerable amount of poetry that fit implicitly within the heroic plot. Lichtenstadter explains that “Much of the surviving poetry has praise for the heroes or scorn for the enemy as its main topic” (14). Indeed, the Arabic qas.¯ıdah includes a particular section called “praise” that presents a “picture of the ideal Arab hero,” “Virile, enduring, a valiant fighter” (23). The marthiya is a distinct genre devoted to “mourning a fallen

162

The Mind and Its Stories

hero” (23). Clearly, such celebrations of the hero and denunciations of the enemy are an important feature of heroic tragi-comedies. One need only think of the opening of Gilgamesh (“the man to whom all things were known. . . . He was wise, he saw mysteries and knew secret things”; he had “a perfect body,” and so on [13]) or of the R¯am¯ayan.a (“a man . . . who has both virtue and power, who knows the Law and has a lively sense of gratitude . . . who lives a blameless life and has the good of all beings at heart” [V¯alm¯ıki 1]). Indeed, in some cases, segments of heroic poems are used directly as praise poems. Speaking of the Epic of Son-Jara, Irele notes that “The many passages of praise poetry with which the epic is interspersed stand in a close formal relationship to the formulaic plan of the narrative. . . . As generally with praise poems in oral literature, they are composed of strings of epithets, often hyperbolic, emphasizing the uncommon attributes – the heroic essence as it were – of their subject.” Moreover, they “are often performed as autonomous pieces” (2337). In short, the praise and blame poetry of the Arabs and other peoples fall squarely within an implicit heroic plot. There is also a great deal of romantic poetry in Arabic, much of it in the nas¯ıb which served as preface to the qas.¯ıdah. Arberry explains that, while early Arabic poems emphasized “the martial virtues,” the “lyric of love” came into prominence later (12). He goes on to signal the narrative element, noting that Arab “poets developed a kind of drama out of the pleasures and pains of love-making” (17). Persian writers reworked the nas¯ıb into the ghazal, which focuses almost entirely on love. Levy explains that the term “ghazal” itself “derived from an Arabic original meaning ‘lovers’ exchanges.’” Indeed, as Ralph Russell points out, the ghazals typically involve an implicit narrative in which two lovers are kept apart by the censorious legalism of a Muslim shaikh, an “elder” with putative “moral authority” who is the “implacable foe” of the “true lover” – a straightforward instance of the standard romantic plot of conflict between the lovers and an interfering social representative. Levy notes that Persian poetry “in general . . . concerns itself either with flattery or love” (32–3), the flattery in question being largely political and, as such, part of the heroic genre. In keeping with this general pattern, ancient Egyptian lyric poetry too prominently includes poems of praise – largely religious – and

Emotion, Narrative Junctures, Lyric Poetry

163

love lyrics. The love poems treat moments of seduction, brief encounters with the beloved, and other junctures of the standard romantic plot. For example, in one poem, “I think I’ll go home and lie very still” (Mack et al. 59), the lover pretends that he is about to die, so that everyone will come to visit, including his beloved – though she will know it is all a ploy. This takes up the common motif of lovers separated by death or near death, but twists it into a comic pretense, understood and shared by lover and beloved. Returning to China, we find poems of love scattered throughout, from the Book of Songs to the writings of China’s renowned later poets, such as Li Po (Li Bo). A uniquely well-known example among English speakers is Li Po’s “Ballad of Ch’ang-Kan,” familiar in Ezra Pound’s rendition, “The River-Merchant’s Wife: A Letter.” Though one could name many relevant writers, a particularly apt case is Li Ch’ing-chao (twelfth century), generally regarded as China’s greatest woman poet (Ling 83). She has treated many junctures of the romantic plot, with poems of falling in love, poems of union, poems of separation, and poems on the death of her beloved – all collectively recounting a romantic tragedy. The heroic plot figures centrally in Chinese lyric poetry as well. Indeed, Li Ch’ing-chao herself wrote poems in this mode, addressing moments in the rightful emperor’s defeat and exile. In this context, ¨ Chieh’s ‘Ode to one might consider, for example, her “Poems on Yuen the Restoration of T’ang’ to Rhyme with Chang Wen-ch’ien’s Poem,” one of which begins by referring to the reign of Ming Huang and “the amazing story/Of . . . downfall and restoration” (57). The heroic plot was taken up by the man widely viewed as China’s greatest male poet as well – Tu Fu (Du Fu). Indeed, his most famous poem (Cooper 168) is a treatment of the aftermath of war, a lyric version of the epilogue of suffering. In keeping with the structure of heroic tragicomedy, it begins with apparent enthusiasm, evidently celebrating the noise of the advancing soldiers and their weapons. But it ends quite differently, with “white bones” and “bitter” ghosts who “cry aloud” (168). Moreover, these are not isolated instances, as broader surveys of Chinese literature show. James Liu distinguishes a range of important motifs in Chinese poetry: the sorrow of parting, the horror of war, nature, time, history, leisure, nostalgia, love, and wine

164

The Mind and Its Stories

(Art 48–60). Love, parting, and war obviously fit our argument well. The others do also. Liu’s example of time is, precisely, the passage of time as experienced by separated lovers, while his example of nostalgia focuses on exile. The historical motif is, unsurprisingly, heroic. In addition, Liu’s categories mark out just the sorts of junctural moment we have been considering. Elsewhere, in a discussion of tz’u poetry, Liu stresses “worlds in lyric poetry . . . dominated by romantic love” (“Literary” 135). He also emphasizes “lyrics which embody heroic sentiments” (144). Waley organizes the Book of Songs into seventeen categories. Like the items on Liu’s list, these are largely junctural. The first three are “Courtship,” “Marriage,” and “Warriors and Battles.” Two further sections address “Dynastic Songs” and “Dynastic Legends.” These are often straightforwardly concerned with heroism – “Terrible in his power was King Wu; None so mighty in glory” (230); “Oh, gloriously did the king lead”; “He brought peace to myriad lands. . . ./Bold was King Wu, Guarded and aided by his knights/He held his lands on every side” (237). These are by far the longest sections and, collectively, cover roughly two-thirds of the volume. Bownas presents a catalogue of early Japanese poetic concerns, showing the same tendencies: “Nature, love, partings, and time” ¯ the largest cate(lvii). In the important early anthology, the Kokinshu, gory of poems focus on love, followed by poems on nature. (We will consider the category of nature poems in the next section.) Together, these comprise roughly two-thirds of the collection. The remainder includes work devoted to such relevant topics – or junctures – as “Felicitations,” “Parting,” and “Laments” (see Berggren 410). In his ¯ generally considered the inaugural essay preface to the Kokinshu, of Japanese literary theory, Ki no Tsurayuki comments that “Poetry makes sweet the ties between men and women and comforts the heart of the brave warrior” (quoted in Keene, Major Plays 201n.). The statement is clearly designed to evoke the breadth of lyric poetry. It tells us that lyric poems range precisely from the romantic to the heroic. Brower and Miner quote many relevant poems. One romantic instance, again marking the moment of separation, is Hitomaro’s “On Parting from His Wife as He Set Out from Iwami for the Capital,” a forty-nine line poem that expands this junctural moment, dwelling on its pathos (116).

Emotion, Narrative Junctures, Lyric Poetry

165

These patterns are not confined to Europe, Asia, and the Middle East. George Cronyn begins American Indian Poetry with a sequence of lyric poems from Abanaki, “The Parted Lovers.” The first line reads, “My parents think they can separate me from the girl I love” (3). Subsequently, he has love songs translated from Chippewa. The first marks the separation of the lovers (21), while the last represents reunion (24). Trask’s The Unwritten Song includes two poems from the Trobriand Islands. The first is a “Song of War,” covering “a fight, a death, and an exile” (203), while the other is a “Love-Magic Spell” (205), thus a specification of the moment when the lover seeks to attract the beloved. He includes two songs from the Kurelu (New Guinea). One is a “Dirge for a Son Killed in Battle.” The other is a rather general lament of men for girls they have lost. Of the next ten songs from New Guinea, one concerns a courtship ceremony, another is the “Lament of a Widow for Her Dead Husband,” and another six are straightforward songs of longing for a beloved. Firth and McLean divide songs by the Polynesian Tikopia into a small number of categories – songs of “travel,” “eulogies and farewells,” “protest,” “erotic arousal,” “laments and funeral dirges,” and “songs on historical and mythic themes.” A number of the poems in various categories emphasize one or another element in romantic and/or heroic stories (for example, exile – see 129–30, or 182). The songs of “erotic arousal” clearly take up the former. The most directly heroic poems are in the “historical” section, which includes songs on “the midst of a battle” (257), “Evading arrows” (261), “An attack in Vanikoro” (262), “Struggle with invading Tongans” (264), and so on. The Yirrkalla poems (in Berndt) treat sexual union almost exclusively, concentrating on what is obviously a crucial juncture in the romantic plot, if a juncture often passed over without explicit mention in more priggish traditions. Trask’s section on Africa begins with “Song After Defeat,” from the Bambara. It includes poems of men and women who have just fallen in love, poems of men and women who have been separated from their beloved through exile, elegies and funeral songs, songs of exhortation for battle, praise songs for rulers, and so on. Indeed, many examples could be cited from Africa. One particularly interesting – and challenging – body of African work comes from the Dinka. Deng’s

166

The Mind and Its Stories

collection of Dinka lyric poetry includes a range of heroic poems. War songs are one of his main topical categories. Moreover, many of the nonwar poems treat social domination. Thus, many poems that do not strictly fit the heroic tragi-comic structure, nonetheless involve an implicit or partially explicit narrative based in the social prototype for happiness. Even more interesting than the heroic songs of the Dinka are their love songs, for these operate not only in a literary narrative, but in life stories that are at least imagined to fit the standard structure of romantic tragi-comedy. Deng explains that poems “are often a basis for winning affection between young men and women, and for imploring elders to respond to the demands of the intending partners. They are also used by the partners themselves to influence each other in order to strengthen their hands against their opposing elders. Many are the cases in which marriages result from a song addressed to elders, the intended bride, or the groom” (80).3

eulogy and nature: some complications of the narrative hypothesis Without actually admitting it, I have slightly modified the narrative hypothesis from its original statement in order to admit praise and blame poetry. Poems of this sort clearly fit into the heroic genre. However, they do so via character traits or physical properties rather than narrative junctures per se. This addition is unsurprising, indeed predictable, for such elaboration of a character is an important part of heroic plots, as we have noted, and it occurs at moments that are almost necessarily junctural – for example, just before or just after a battle. A second very important type of nonjunctural poetry is that concerning nature. (Thus poetry focusing on the third component of narrative, scene, rather than event or character.) This too is in keeping with the spirit of the narrative hypothesis, though not its 3

A large number of Dinka poems concern cattle and thus do not clearly fit into either the romantic or the heroic structure. This is not in itself a problem. Again, we are not dealing with necessary and sufficient conditions here. However, we shall see in the next chapter that there is a third prototype narrative, concerning hunger and food. Many of the Dinka ox songs fit directly into that structure.

Emotion, Narrative Junctures, Lyric Poetry

167

initial formulation, for patterns of nature imagery are a crucial part of romantic tragi-comedy. A great deal of nature poetry is, in effect, an elaboration of this imagery. I mentioned above that Sanskrit drama often integrates lyric poems into narrative sequences. Again, some excellent cases of this ˜ ana´sa¯ kuntalam. Throughout the are to be found in K¯alid¯asa’s Abhijn¯ play, key junctures are marked and expanded by verses. Some of these make no direct reference to the story, focusing only on nature. Thus, if they were printed separately, they would seem to have no narrative element. However, in context, they are clearly bound up with the encompassing story. For example, when the king is first ´ overcome with passion for Sakuntal¯ a, there is a short poem about a stampeding elephant. After a short separation, the lovers are reunited, and the king recites a couplet: “Parched with thirst, the bird has only to crave for water/and a shower from a fresh rain-cloud falls ´ into its mouth” (210). When Sakuntal¯ a is leaving her parental home, one of her friends utters another poem that consists only of three images – a doe that does not eat, motionless peaco*cks, and faded leaves (224). Nature imagery is used in precisely the same way in much lyric poetry. Li Po’s “Ballad of Ch’ang-Kan,” spoken by a young woman separated from her beloved, makes reference to the thick mosses by the door, suggesting that no one has walked through the entrance. She notes that butterflies are flying in pairs – the contrast with herself walking alone is obvious. Without explicitly mentioning the husband’s travel to distant places, such natural images operate to suggest that absence, thus reenforcing the junctural element of separation. Of course, nature poetry is not always explicitly linked to narrative. Nonetheless, these explicit connections suggest that there may be implicit connections in other cases. My contention is that a great deal of apparently “pure” nature poetry derives from and implies an unspoken romantic narrative. Consider, for example, a poem by ¯ perhaps his most famous (Yuasa 149n.2). First, the Japanese Basho, (quoted in Shirane 322): furu ike ya kawazu tobikomu mizu no oto

168

The Mind and Its Stories

Then a translation (Aston 295, altered): An ancient pond. A frog plunges in, The sound from the water.

The implications of the poem are incomprehensible outside its literary context – the complex, well-defined patterns of topic and imagery that governed the genre. Shirane explains that the frog or kawazu was “almost exclusively associated with the blossoms of the yamabuki (kerria), the bright yellow mountain rose, and with limpid mountain streams.” Moreover, the “poetic essence” of the kawazu was understood as its “beautiful voice.” The typical poem on the kawazu would treat “the plaintive voice of the frog singing in the rapids or calling ¯ poem, implied out for his lover” (323). Thus, the context of Basho’s by the conventions of the literary tradition, is one of exile from one’s beloved. Note also, however, that the conventions themselves suggest an eventual resolution or reunion. The voice of the kawazu is beautiful. He is surrounded by blossoms. The moving water of the limpid stream indicates advancement, at least movement toward some sort ¯ poem is almost despondent. In place of future. In contrast, Basho’s of the moving water, and the future it suggests, there is a motionless pond. In and of itself, this need not be a despairing image. For ponds can be beautiful and peaceful. But this pond is not filled with lotuses. The beautiful blossoms, expected due to convention, are nowhere to be seen. Neither is the pond animated with the new and vigorous life. Rather, it is old. Perhaps the frog is old too, for he does not sing to his beloved. In context, it seems that the separation has become hopeless. There is no longer any future. There is no murmur of a moving stream, but only the isolated sound of the frog entering the still water. This is a good example, for initially it appears to be a straightforward, uncomplicated nature poem, representing a pure natural experience, entirely separated from any narrative – a moment sealed off from any story. But upon examination, it turns out to be thoroughly ¯ poems are like this. embedded in a romantic plot. Many of Basho’s Indeed, it is worth noting in this context that Basho¯ formulated an explicit poetics based on the concept of sabi or loneliness. In a sense, ¯ poetry is oriented toward the lover’s the entire aesthetic of Basho’s period of isolation in romantic tragi-comedy.

Emotion, Narrative Junctures, Lyric Poetry

169

Given this addition of much nature poetry, and the previous inclusion of much eulogy and satire, we need to reformulate the narrative hypothesis. Prototypical lyric poetry is not, as one might imagine, opposed to or even independent of narrative. In fact, it is regularly bound up with narrative in its production, meaning, and impact. Individual poems typically imply encompassing narratives. More exactly, though lyric poems may in principle treat any number of topics, prototypical lyric poems are most often elaborations of emotively distinctive constituents from romantic or heroic tragi-comedy. The constituents in question are primarily narrative junctures, with their particular junctural emotions. However, these constituents may also be emotionally consequential features of character or natural images associated with such junctures (as when imagery of spring is linked with the reunion of lovers).

religious lyric poetry and the narrative structure of spirituality I should like to conclude by suggesting that this analysis of lyric poetry has consequences that extend beyond our understanding of literature and genre. Specifically, I should like briefly to consider a particular aspect of religious belief, by way of a look at devotional lyric poetry, an important type of poetry that we have not yet discussed. In keeping with the preceding analyses, there appear to be two standard types of relation to God in lyric poetry. The first is a relation of awe, expressed in poems of praise. The second is a relation of longing, expressed in poems of separation and mystical union. Lyric poems of the first sort are so widespread that they hardly require isolation. This type ranges from ancient Egyptian poems (“God . . . majestic, awesome, bedazzling, exalted. . . . Revealed like a king in glory” [Mack et al. 45, 47]), to some of the Hebrew psalms of praise, to some passages in Chinese poetry (“Mighty is God on high,/Ruler of His people . . . Swift and terrible,” as the Book of Songs puts it [Mack et al. 785]). Poems of this sort clearly relate to the heroic genre. More significantly, they manifest a view of divinity based largely on the prototypical features of the heroic genre. This view includes the assimilation of God to a temporal monarch, a related stress on God’s combative, protective, and punitive aspects, and an

170

The Mind and Its Stories

emphasis on the rightful domination of society by the religious ingroup. Correlatively, this view of divinity as absolutized hero often carries with it an enemy of absolutized evil, though this figure does not often inspire lyric poetry. In any case, instances from Christianity and Islam are too obvious to require mention. An instance from Hinduism may be found, outside of lyric poetry, in the case of R¯avan.a – who, of course, makes his appearance in the heroic tragi-comedy of the R¯am¯ayan.a. Returning to lyric poetry, we find the devotional alternative almost equally well represented. Just as praise poems are part of the heroic structure, devotional poems are most often part of the romantic structure. Hindu bhakti poems and the ghazals of Muslim Sufis illustrate the point strikingly, for they are explicitly based on romantic tragi-comedy. They directly assimilate the relation between a devotee and God to the relation between lover and beloved – for example, characterizing life on earth as exile from the beloved. Poems of this sort manifest a view of divinity based on prototypical features of the romantic genre, emphasizing mutual love, mutual devotion, the despair of isolation and exile, and the ultimate union, even unity, of God and devotee. The Song of Songs, as commonly interpreted in the Jewish and Christian traditions, draws on this structure as well, as do the poems of such Christian mystics as St. John of the Cross (see Brenan for examples). Finally, the Nine Songs of the fourth- or thirdcentury b.c.e. are among the earliest extant lyric poems of the South Chinese Ch’u culture. As Owen points out, “If the north Chinese writing was largely concerned with statecraft and sober social virtues, the Nine Songs, by contrast, celebrate the senses and a passionate, often erotic, relation between gods and mortals” (828). A good instance of this romantic devotionalism is discussed by David Hawkes. This particular poem begins “The goddess comes not, she holds back shyly” and goes on to treat the devotee’s relation to the goddess in terms of frustrated romantic love – as in the lines, “The lady is sad, and sighs for me;/And my tears run down over cheek and chin” (45). The connection of all this with actual religious belief indicates that the embedment of lyric poetry in prototypical narrative is not cognitively special or isolated. Rather, our conception of and response to divinity – more generally, our religious attitudes and thus much of what goes along with those attitudes – are, to a great extent, shaped

Emotion, Narrative Junctures, Lyric Poetry

171

by implicit prototype narratives, and not only when we write lyric poems. Moreover, it seems unlikely that this relation between plot and religion is itself unique or anomalous. Rather, it seems likely that broad areas of nonliterary cognition are bound up with standard narrative structures and emotion prototypes in much the same way.

conclusion The “narrative hypothesis” asserts that prototypical lyric poems most often treat emotionally distinctive elements (primarily junctural moments) in implicit heroic or romantic narratives. In this chapter, I hope to have shown that the narrative hypothesis gives a highly plausible account of some otherwise very surprising cross-cultural features of lyric poetry. The final reflections on devotional lyrics and religion suggest in addition that some form of the narrative hypothesis may be extended, with equal plausibility, to a range of nonliterary phenomena.

6 Testing, Revision, and the Program of Research in Narrative Universals Ainu Epic and the Plot of Sacriﬁce

testing and research programs in literary narrative As I have already stressed in the Introduction, the work of Imre Lakatos suggests that the greatest value of a theory lies in the research program that it enables. Specifically, Lakatos distinguishes between “progressive” and “degenerating” research programs. Both encounter anomalous data. Both uncover phenomena that contradict their theoretical predictions. The difference is that a degenerating research program deals with recalcitrant data through ad hoc formulations. In other words, a degenerating research program formulates principles that account for the recalcitrant data, but do not explain anything beyond those data. In contrast, a progressive research program is continually reformulating its constituent theories in such a way as to expand the explanatory scope of those theories beyond any recalcitrant data. Indeed, a progressive research program actually seeks out recalcitrant data as a way of increasing explanatory reach and precision. The most obvious way of generating new data – recalcitrant or not – is laboratory experimentation. When engaging in experimentation, we artificially create circ*mstances in which our hypotheses might prove problematic. In other words, we create situations in which our theory may not work, situations in which events may develop in a way that contradicts predictions made by our theory. This sort of 172

Testing, Revision, and the Program of Research

173

experimentation is possible for certain areas of literary research, including research in narrative. It is important and should be pursued much more vigorously than has been done in the past. Unfortunately, in many areas of literary theorization – including most of what we have discussed in the preceding chapters – this option is not readily available. For example, a colleague trained in laboratory research suggested an experiment in which independent observers, unfamiliar ˜ ana´sa¯ kuntalam has the strucwith the preceding theory, judge if Abhijn¯ ture of a romantic tragi-comedy, The Epic of Son-Jara has the structure of an heroic tragi-comedy, and so on. But I suspect that readers familiar with these works will not find this to be something that requires ˜ ana´sa¯ kuntalam is controlled study. It seems uncontroversial that Abhijn¯ about two people who meet, fall in love, are separated due to a conflict with social authority, and so on. It seems uncontroversial that Son-Jara is the rightful ruler, exiled from his kingdom, as a relative takes the throne, that there is a threat to the kingdom, that Son-Jara defeats that threat, and so on. In short, an experiment of this sort seems unlikely to produce anything that would advance a research program in narrative universals and emotion. Here, it is worth returning to the example of linguistics that is, again, the one field in which the study of universals has made significant progress. In some respects, the situation in literary study is not greatly different from that in linguistics. Many areas of linguistic theorization are open to laboratory research. Many, however, are not. Of the latter, the isolation of linguistic universals is a case in point. There is no doubt laboratory work that bears on the study of linguistic universals. However, the isolation of the universals themselves, their full description and elaboration, is dependent upon the study of natural languages. Fortunately, in this area of linguistic study, there is a parallel to experimental research – the examination of a previously unstudied language. We may follow Lakatosian prescriptions and test hypotheses regarding linguistic universals by examining languages that were not used in the initial formulation of those universals. The same point holds for bodies of literature. We may to some degree test hypotheses regarding literary universals – in this case, narrative universals – by studying a body of narratives that did not contribute to the initial formulation of those hypotheses.

174

The Mind and Its Stories

Broadly speaking, a new language will be most valuable to evaluating linguistic hypotheses if it is genetically and areally distinct from languages previously studied and if it is widely attested. As to the last, if we are considering a living language, this is not a problem, for we can simply recruit a native speaker. If it is a dead language, however, it is valuable to have a large body of literature – not simply a few lists of merchandise or the like. These general criteria apply directly to literature as well. We discussed genetic and areal distinctiveness in Chapter 1. As regards attestation, almost all literary traditions are, in effect, in the position of dead languages. Outside oral traditions, it does not really make sense to refer to a native informant. For example, it is not clear what he/she would inform us about. Thus, we need a tradition that has an adequate body of extant work. These are not the only criteria for choosing a linguistic or literary test case. In addition, one should seek typological diversity. In other words, a good test case would be a language that is of a different type from the majority of languages studied thus far – with “type” here being simply a matter of some potentially theoretically relevant variable. (Indeed, all these criteria are ultimately a matter of controlling variables, in direct parallel with laboratory experimentation.) If a linguistic theory is well supported by research in synthetic languages, then a researcher may be well-advised to look into analytic languages. The same point holds for universals of verbal art. The most obvious uncontrolled variable in the present theory is authorial sex. In order to limit areal contamination, I have focused on relatively early writings in each major written tradition. Unfortunately, one result of this historical focus is that most of the works I have considered were composed by men. Certainly, romantic tragi-comedy figures prominently in the novels of Jane Austen or in the devotional lyrics of the women bhakti poets such as Mah¯ad¯eviyakka and M¯ır¯ab¯a¯ı. But the preceding pages include few references of this sort. Moreover, some writers might object even to these references, arguing that I have not considered women’s own narrative tendencies, but only women’s writing that has thoroughly absorbed the preceding, largely male traditions, has been canonized and preserved by men, and so on. While it could be argued that, for example, the female bhakti poets were not entirely determined by a literary tradition – or

Testing, Revision, and the Program of Research

175

that, more generally, women authors are not so entirely subservient to male forebears – it would nonetheless be beneficial if we were able to uncover a body of narrative dominated by women writers or oral tellers. Indeed, oral tellers would be better, since I have relied disproportionately on written literature (that is, literature proper) as well. Though I have drawn on orature, it is relatively less well represented. In part, this is necessitated by the material available. Most orature has necessarily passed forever from the world, or been incorporated into written literatures. However, though a degree of such underrepresentation is unavoidable, further research in orature remains a desideratum.

ainu epic Ainu epic poetry constitutes an almost ideal body of work for our purposes. First, Ainu is one of those anomalous languages that bear no clear genetic relation to other languages. As Philippi points out, Ainu “cannot be genetically connected with any other language groups in the world” (7; for discussion, see Shibatani 5–8). This suggests that Ainu traditions of narrative are genetically distinct as well, since story telling appears to be coeval with full human language use. Moreover, on the island of Hokkaido, the Ainu were relatively isolated. Philippi argues that, for the most part, “the Ainu were living a way of life rooted in the remote past, and Hokkaido remained untouched by the main currents of Asian history” (6). There was interaction with the Japanese. However, Japanese influence did not become strong until the end of the seventeenth century (see Philippi 6). Philippi has assembled and translated a set of Ainu narrative poems that reflect “the social conditions and the cultural milieu” of the pre-Japanese period (9). Specifically, “The ideas which are expressed in the epics and which, in fact, formed the basis for the traditional Ainu way of life, are extremely archaic and share many common features with the ideology which was prevalent among the Paleoasiatic peoples of northeast Asia until recent times. . . . The Ainu epic tradition, with its extremely archaic mental patterns and modes of diction, is one of the purest and most beautiful surviving examples of the oral literatures of the hunting and fishing peoples of northeast Asia” (50). Philippi’s selection itself provides the relatively broad attestation we

176

The Mind and Its Stories

require as well. Rather than one or two canonized epics, Philippi gives us thirty-three narrative poems of different lengths, types, purposes, and topics, poems recited by different bards in different regions. Thus, the Ainu epics satisfy the general criteria we set out initially – genetic distinctness, areal isolation, and wide attestation. They also satisfy the criteria that are more specific to our hypotheses regarding narrative. First, as the preceding discussion suggests, the poems in this collection were recited by oral bards working in an oral tradition. Indeed, the tradition is unusually nonliterate; Philippi points out that “The Ainu never developed any system of writing” (6). Perhaps the most striking feature of the Ainu narrative poems, however, is that “nearly all” of them “were recited by women” (Philippi 50). Thus, we have a large selection of works of verbal art, evidently preserving highly archaic elements, that derive from a tradition that is to a great extent female. Indeed, this is particularly significant as “Ainu women had a highly developed culture of their own which differed from that of men and contained many elements which were kept strictly secret from them” (50). Moreover, among the Ainu, poetry was closely related to shamanistic traditions and “Among the Hokkaido Ainu, almost all the shamans are women” (3). In short, the Ainu epics provide a nearly perfect test case. What, then, do we find when we examine these epics? How do they affect the ideas put forth in the preceding chapters? Beginning with the variable of gender, we do find some gender related differences from the other works we have considered. But these differences are superficial and manifest underlying indexical universals, such as a desire for autonomy, a particular concern with one’s own in-group, and so on. For example, in the Ainu narratives, women characters are more active, more independent, more likely to be the focus of the events or to provide the point of view. Such differences are consequential and worth further research, as well as cross-cultural comparison. But they are not in any way contrary to the preceding hypotheses. Moreover, the preceding account of heroic tragi-comedy – a genre one might expect to be affected by gender – fits the Ainu material without significant alteration (though the Ainu poems do highlight the variability of heroic narratives and the derivation of

Testing, Revision, and the Program of Research

177

political revenge plots from the usurpation sequence of the heroic structure).1 On the other hand, this is not to say that the Ainu epics leave us exactly where we started. In fact, these poems do lead to two 1

The final long poem, “The Epic of Kotan Utunnai,” provides a good example. (Here and elsewhere, citations of Ainu epics refer to Philippi.) It concerns the great Ainu hero Poiyaunpe. He is raised by an “enemy” woman in the enemy land. He learns that his father “held sway over the upper/and the lower regions/of Shinutapka” (368), but was killed by the enemy when abroad. On hearing this, Poiyaunpe sets off on a series of battles. After defeating various enemies, he returns in triumph to Shinutapka. Though the discourse or telling of the epic slightly alters the order, the story or plot is straightforwardly that of heroic tragi-comedy, with the common variation that the usurpation of the rightful ruler is overcome by his son. Specifically, the rightful ruler is killed; in consequence, the heir lives in exile. When he reaches maturity, the heir battles various dangers to his society, then returns to his society and assumes his rightful place. The enemies faced by Poiyaunpe are particularly interesting here. Most are of the general sort we considered previously. For example, at one point he battles “Etu-rachichi,” a renowned warrior of the repunkur people, the traditional enemy of the Ainu. At other points, he battles foreign rulers or armies from strange lands. However, some of his antagonists are less traditionally realistic. For example, the last battle fought by Poiyaunpe is against the “bad weather demon,” an obvious threat to his people, but a different sort of threat than an enemy army or warrior. In fact, many Ainu epics focus on threats from nonhuman antagonists. Consider, for example, the third “Song of Aeoina-kanui.” In this story, the country is threatened by an attack from Big Demon. The hero departs to defend his society from the danger. After a long battle, both the hero and the Big Demon die. However, the hero is revived by the gods and returns home. Here we have a clear case of the central sequence from the heroic plot. A threat to the society is repulsed by the hero. We also have death in the narrative middle. However, we are not faced with a traditional, military/national enemy. Moreover, this is not an unusual case. As Philippi points out, Ainu “epics in which the culture hero” defeats “evil deities” are “very numerous.” Other Ainu stories make animals into the threat. In “Song of an Evil Bear,” for example, Okikurmi deceives and kills a bear who poses a threat to human society. In the first “Song of Aeoina-kanui,” the enemy is a huge char. In all these cases, the greatest theoretical interest of the variation is probably in the way it highlights the common metaphorization of enemies as animals. In the heroic tales considered earlier, the enemy is most likely to be another national group, not a demon or an animal per se. However, in those stories, it is common to characterize the national enemy in demonic or bestial terms. (This is clearly the alternative or contrary to the empathic humanization that occurs in the epilogue of suffering.) That dehumanizing of the enemy is rendered more salient through the Ainu data. Interestingly, almost all these cases manifest only the threat/defense sequence. In part this is an instance of a universal pattern. Cross-culturally, the threat/defense sequence appears more likely to occur on its own than the usurpation/exile sequence. In other words, the former is more autonomous than the latter. To some

178

The Mind and Its Stories

alterations in the preceding hypotheses, one small, one quite big. The first is a relatively minor change in the precise delimitation of the romantic plot. The second involves the addition of an entirely new prototypical narrative structure. As it turns out, these alterations are

extent, this is true for formal reasons. If there is a threat to the society, the hero must defeat it, wherever he/she might be. His/her action does not in any way require a prior displacement and exile. However, the displacement/exile sequence does to some degree require the threat/defense sequence, for the hero proves the injustice of his/her displacement and exile and regains his/her place in society precisely by defending the society against threat. In the Ainu case, however, there is more to it. The relative scarcity of usurpation plots in this case results also from the social and political structure of Ainu society. Historically, the Ainu lived in minimal communities: “Generally speaking, Ainu settlements were small, consisting of perhaps one to ten households” (Philippi 6). Moreover, “The Ainu never developed . . . any concept of a political state” (6). The very small local groups of five to ten households did have headmen (Watanabe 79, Ohnuki-Tierney 73–8). However, the larger collectives of local groups operated “without any chief” (Watanabe 79; see also Chapter II of Watanabe). On the other hand, the usurpation sequence is not absent. We have seen one poem in which the usurpation/exile sequence occurs – “The Epic of Kotan Utunnai.” There is one other poem of this sort as well, “Song of the Young God Okikurmi.” This song recounts a dream dreamt by Okikurmi’s son. He has been raised by a cannibal woman, who brings home great quantities of human flesh to eat. Clearly, she is a threat to human society. In addition, she is raising Young Okikurmi because she killed his father. Thus, we have the standard sequence of the hero being killed, while the son is exiled (thus denied his rightful place). In this case, there is the additional twist that the older hero’s murderer is the threat to human society and the young hero’s exile is in the land of that murderer. The cannibal takes young Okikurmi out to the place where she killed his father. However, young Okikurmi kills the woman. He does not return to his home in triumph, for he wakes and realizes that it was a dream. Nonetheless, the presence of the standard structure is clear. These two direct instances – and an indirect instance in the romantic “Woman of Shinutapka” (to which we shall turn below) – indicate clearly that the usurpation sequence is part of the heroic structure among the Ainu. Given the formal and social factors working against the manifestation of this structure, its appearance in three poems (which, due to their length, comprise over a sixth of the collection) suggests the psychological persistence of this structure and of the prototypes it follows. In keeping with this, “The Epic of Kotan Utunnai” and the “Song of the Young God Okikurmi,” the two poems directly containing the usurpation sequence, are also among the few Ainu poems that point toward the positive social goal of domination, rather than the negative social goal of avoiding destruction. The “Song of the Young God Okikurmi” shows us something else as well. In this poem, the pursuit of domination takes a form that is consequential for our concerns. That form is revenge. Of course, this is not unusual. Indeed, it is noteworthy precisely because the motif of revenge is widespread cross-culturally. Usurpation generally results from the actions of some heroic rival. Moreover, the final return of the hero, ruler, or heir is regularly accompanied by the punishment ofthat rival. The political

Testing, Revision, and the Program of Research

179

not ad hoc, but apply to a wide range of literary works in the traditions considered earlier, revealing patterns that had been left aside in the earlier formulations. In other words, these reformulations do just what Lakatos says they should do. They extend the explanatory scope and precision of the theory. The first reformulation is really just a matter of emphasis. In setting out the structure of romantic tragi-comedy, I stressed that the separation of the lovers was primarily a matter of conflict with society, most often conflict with parents. I noted in this context that there is often a rival as well, frequently a rival chosen by the parents as a more suitable spouse for the lover or beloved. It also happens, however, that the rival acts more independently – even to the point where a love triangle occurs. In other words, the rival may be developed more fully and autonomously, and the interfering parent may be deleted, though this rival character does often retain the same function of standing for social norms. For example, in opposition to the lover, the rival may be of the right religion or social class and thus he/she may appeal to the beloved for precisely the reasons he/she would appeal to the beloved’s parents. Outside the Ainu material, an example of this general sort may be found in Jane Austen’s Emma, where Harriet rejects a loving farmer, Robert Martin, setting her sights on the more socially elevated Mr. Elton – just the sort of thing a dominating parent would demand in the more complete version of the romantic structure. Of course, as often happens, this is not a pure case of a love triangle, for revenge plot merely separates out this elementof the usurpation sequence – a point well illustrated by, for example, Hamlet. It is perhaps worth noting in conclusion that clich´es about gender difference would lead us to expect that female bards would craft their poems in such a way as to undermine violence and warfare. Thus, given standard beliefs, we would expect relatively little emphasis on battle, and much greater emphasis on remorse and penance. In fact, we find almost the exact opposite. One striking feature of these poems is that the women regularly join in battle, killing the enemy with great strength and vigor. At the same time, we find less attention to the horrors of war and the remorse that follows battle. The greater presence of female warriors is presumably due to the fact that the bards are women. Following a general human tendency, they place themselves or people like themselves in more active and heroic roles. Indeed, this may be one reason for the relative lack of remorse, and the virtual absence of an epilogue of suffering. (We find it only in one poem, “Song of the Daughter of the Mountain God.”) Insofar as the poems allow women to experience a sort of heroism that is not ordinarily available to them, an emphasis on remorse and penitence would undermine one important function of the work.

180

The Mind and Its Stories

Emma herself has in effect taken up the role of interfering parent, if in a diminished form, by opposing the marriage to Martin and urging Elton as an appropriate alternative. Obviously, the love triangle is not confined to Emma and Ainu epics. Despite its very noticeable recurrence, however, I had understated its importance in romantic plots. The repetition of this general structure in the Ainu poems drew my attention to it elsewhere as well. (The Ainu poems also suggest that love triangle plots often incorporate elements of the heroic structure to an unusual degree.)2 2

A number of Ainu epics fit the romantic structure quite well. Consider, for example, the “Song of a Human Woman.” In this tragic story, a human woman marries a god. The older brothers of the god find out about the marriage and force the husband to leave the earth and return to heaven. The wife continues to long for her husband until she dies. The story represents the separation of lovers due to a conflict with family and, implicitly, with society (here, the society of the gods). This conflict leads to separation and to the exile of the lover. The ascent of the husband into the heavens carries suggestions of death as well, and, again, the wife literally dies. The story ends tragically – but perhaps there is a hint of a possible spiritualized reunion after the wife’s death. A more complex case may be found in the “Song of the Younger Sister of the Owl God.” In this story, the Owl God has followed the urging of the other gods and arranged for his younger sister to wed the dislikeable god of Poroshir. The great hero, Ainurakkur, visits the Owl God’s home because he is “so exceedingly/lonely” (215). Poroshir mistakenly believes that the Owl God intends to give the younger sister to Ainurakkur. So he gathers his armies to “launch/a war of annihilation” (218). Ainurakkur has to leave the land of the Owl God and fight the armies of Poroshir. However, it seems that Poroshir’s apprehensions were not entirely misplaced, for Ainurakkur takes the younger sister along with him, and she joins him in battle against Poroshir. They eventually defeat Poroshir and are married. The story ends with the younger sister explaining that “We lead/a magnificent married life” (228). The story begins with an implicit conflict between the lovers and both a parental figure – the Owl God, who raised the “younger sister” – and society more generally, as represented by the other gods. There is a period of exile and imagery of death (Ainurakkur actually describes what will happen if he dies in battle). The hero – and the beloved – struggle against the rival in a section that borrows from the heroic structure. Finally, the lovers are united. This story presents a good instance of the gender differences noted previously. Here as elsewhere, the women are far more likely to join directly in battle or other struggles with the men. They are far less likely to be passive victims, merely saved or protected by men. However, their activity does not change the basic structure. This story also illustrates the importance of the rival suitor. I had earlier treated the rival as a relatively passive character. Clearly, that is not the case in this story. Indeed, here the main conflict is directly with the rival. He is the one who engages in battle with the lover and the beloved. The theme is, of course, extremely common cross-culturally. Obvious instances would include the R¯am¯ayan.a, in which the romantic part of the plot concerns the abduction of S¯ıt¯a and her rescue by R¯am¯a; the

Testing, Revision, and the Program of Research

181

The other alteration is a much deeper and more consequential theoretical revision. The presence of several Ainu stories treating famine, sacrifice, and ritual indicate that we should add a third prototype narrative, sacrificial tragi-comedy, which concerns a physical, rather

tragic story of Deirdre and Naoise in Irish epic tradition, where Naoise is killed by Conchubar, who takes Deirdre for himself; the broader story of the lliad, involving Helen and Paris, and so on. My previous formulation of the structure did note that a complete romantic plot often involves the defeat of a rival. However, the Ainu poems indicate the degree to which this part of the romantic plot may be split off and emphasized. Moreover, there seems to be more to it than that. As these instances suggest, particular stress on the love triangle often appears to involve the importation of elements from the heroic plot, even the joining of the two plots, with the rival assimilated to the threatening enemy. Indeed, abduction can be the act of a rival in a romantic plot or the act of a threatening enemy in a heroic plot. A particularly striking instance of the love triangle version of the romantic plot is found in the “Song of the Fire Goddess.” In this story, the Goddness of the Waters abducts the husband of the Fire Goddess. The latter pursues her rival, only to find the husband apparently accepting the situation quite willingly. In keeping with what we have just noted about the incorporation of heroic structures, the Fire Goddess then defeats the Goddess of the Waters in battle. However, she does not take her husband back with her. Rather, the husband collects “ashimpe,” a compensation for damages. He finally returns home with “many precious treasures,” apologizing abjectly. “After that,/he remained [at home]” (74). Here too we have the familial conflict, the abduction, and battle. We also have a strange epilogue, directly parallel to the epilogue of suffering in heroic plots. After the reunion has been made possible, the husband goes off on a exilic journey of repentance, ultimately returning home in the sort of somber and contrite mood that characterizes the conclusion of the heroic epilogue. This case, then, shows a particularly extensive incorporation of heroic elements into the romantic structure. There is no point going through all the instances of the romantic plot structure in the collection. However, there are two other instances worth particular mention. These are two of the three longest poems in the collection. The first presents an interesting variation on the abduction structure, and on the structure of familial conflict. This is the “Woman’s Epic.” The speaker was abducted by Repunnot-un-kur when she was still a child. Repunnot-un-kur raises her as a father, with the intention of marrying her when she reaches maturity. The girl is warned by a bear cub, who rescues her from Repunnot-un-kur, returning her to her home. When they arrive at the home, the bear cub is ritually killed by the girl’s brothers. He returns to life as a human being and marries her, explaining that his father disapproved of the marriage. They live happily. This version reshapes the standard structure remarkably. However, it remains a clear variation of that structure. The conflict with the family is given an oedipal twist when the father and the abducting suitor are, in effect, identified. The hero first saves the beloved from the abduction, and is then killed (by the beloved’s family), leading to separation. Finally, when they are married, we learn about his father’s disapproval of the union. All the parts here are from the standard romantic story – but with strange repetitions, combinations, and specifications. The result, though

182

The Mind and Its Stories

than a personal, social, or transcendental goal – prototypically, food, plenty of the primary means for maintaining life. Since the revision of the romantic structure is minor, I shall concentrate on the new prototype structure. Before going on to this, however, I should return to the issue of just what stories are covered by these three structures. In presenting these ideas, I am sometimes asked if I believe these three structures are “exhaustive.” The problem with this question is that it is ambiguous. Most obviously, the question may refer to universal, prototype clearly within the general structure of romantic tragi-comedy, is a strikingly innovative story. Certainly the most outstanding poem in the collection is “The Woman of Poi-Soya,” a technical tour de force that uses different voices and perspectives in an extremely powerful way. The story content is no less fascinating than the form. The male hero is Otasam-un-kur. His parents wish for him to marry the Woman of Poi-Soya, a militaristic woman, with whom Otasam-un-kur engages in battle. Disgusted with Poi-Soya-un-mat, Otasam-un-kur pursues another woman only to find that she has been unfaithful – though this is not her fault as she has been deceived by a rival pretending to be Otasam-un-kur. He kills this woman, her lover, and himself. The lover’s sister, Kunnepet-un-mat, revives Otasam-un-kur and the two of them are married. This, too, is a complex revision of the standard structure. First, there is the parental conflict in which Otasam-un-kur rejects his parents’ choice of a spouse. Then there is the variation on the love triangle and abduction when the beloved is deceived and seduced. Here, Otasam-un-kur follows the standard sequence in defeating the rival. Finally, we have the motif of death and rebirth, followed by union. The big difference, of course, is that it is not union with the beloved, for she has been killed by the “hero” and is not reborn. Rather, in an incorporation of an element from the usurpation plot, her place is taken by her sister. What is perhaps most striking about this poem is that it is so cynical about romantic love – an attitude that is not absent from some of the other narratives also. After Otasam-un-kur’s initial account of the evils of Poi-Soya-un-mat, we hear about many of the same events from Poi-Soya-un-mat herself. She is humanized in the course of this section, and Otasam-un-kur’s flaws begin to appear. The following sections only make his character less and less sympathetic, until we can only pity kind and generous shamaness, Kunnepet-un-mat, for marrying him. Indeed, Otasam-un-kur virtually curses Kunnepet-un-mat when he thinks about their marriage, explaining that it is only because she saved his life that he “allowed her to cook for me” (365). Here, we seem to have the most extreme case of gender difference – a use of the romantic structure to criticize male cruelty and bias. (Of course, this is once more a superficial difference that manifests widespread commonalities – in the desire for autonomy, respect, and so on.) In fact, the aim of this story appears to be the “deromanticization” of romantic tragi-comedy. But, once again, to do this, the poet remains within the structure of that genre. The story repudiates the sexual politics of male domination. But it preserves the story form in which romantic union is imagined, however falsely, as a prototype for happiness and is sought by characters as such a prototype. It varies, but nonetheless retains the universal romantic structure.

Testing, Revision, and the Program of Research

183

narrative structures, or it may refer to stories of any sort. In the latter case, the answer is an emphatic “No, these genres are not exhaustive!” Here I must emphasize once more that we can tell stories about anything. There are infinitely many stories with infinitely many topics, aims, and combinations of emotion. Again, that is what makes the recurrence and prominence of these three types so noteworthy. But these infinitely many stories are not prototypical stories – or they have a significantly lower degree of prototypicality. Nor do they recur universally and prominently. More importantly, I am inclined to believe that these structures do exhaust the range of universal, prototype narrative structures. My inclination here is both empirical and theoretical. Empirically, these are the structures that appear to recur consistently and prominently in the cross-cultural data. Theoretically, it is difficult to imagine what other common context there may be for happiness beyond physical, personal, and social (given the preceding analysis of the spiritual context as lacking a distinct prototype). Of course, I can hardly assert this with complete certainty. After all, prior to reading the Ainu material, I had thought there were only two universal, prototype narrative structures. Who is to say that I or someone else will not discover a fourth pattern and context in the future. Still, it seems likely that the three structures are exhaustive at this level. Here, a third way of interpreting exhaustiveness arises. Do the three structures exhaust not only universal prototype structures for happiness, but even culturally particular prototypes? Put differently, are there culturally particular (thus nonuniversal) prototypes for happiness? One’s first inclination is to say, “Of course there are.” This seems true experientially. There is also theoretical reason to expect this. As Barsalou points out, we generate new, ad hoc categories all the time. Thus, we might speak about “the sort of thing you would sell at a garage sale” or “the kind of gift someone might buy for Mother’s Day.” Barsalou has shown that these categories exhibit the usual prototype patterns. For example, perfume might be a more prototypical Mother’s Day gift than, say, a set of coffee mugs. Barsalou also explains that these ad hoc categories might become well entrenched, thus firmly established lexically. In some cases, these ad hoc categories are, in effect, contextually specified prototypes (for example, “Mother’s Day” provides a context for specifying the gift prototype).

184

The Mind and Its Stories

The point clearly applies to happiness. Prototypes for happiness can and do come to be contextually specified for individuals, for small groups, for larger cultures, in at least apparently nonuniversal ways. But here one might ask whether many seemingly idiosyncratic prototypes are not just more fully specified instances of the universal prototypes. For example, my own evidently idiosyncratic prototype for happiness involves a prestigious university position in a particular region of the Midwest. This seems to be irremediably culturally particular. After all, historically, most cultures have not even had universities. Moreover, the geographical location is highly specified. But in fact this is not, ultimately, culturally idiosyncratic at all. The prestigious university position is clearly a specification of achieving social dominance, if of a rather moderate sort, within a group (in this case, a professional group). The other element is proximity to the city where my wife teaches. In this way, my apparently idiosyncratic and culturally relative prototype turns out not to be idiosyncratic and culturally relative at all. Indeed, it turns out not to be a distinct prototype at all. It is, rather, an instance of the social and personal prototypes.3 3

Of course, it may not be a very good instance, especially for the social prototype. Exampla, too, exhibit “prototype effects,” such as grading. In other words, some exempla are better cases of a prototype category than others. Indeed, despite attempts by some linguists and psychologists to anchor prototypicality in a “basic level” of human cognition, it seems clear that “schema,” “prototype,” and “exemplum” are relative to topic. For example, the basic level is commonly defined as the highest level that can be encompassed by a “single mental image,” the level at which “motor programs” function, and so on (for a summary of criteria, see Lakoff 47). The idea is that we can encompass “chair” in an image, but not “furniture.” However, this is not true. One can easily argue that imagining “furniture” simply involves imagining a variety of different pieces simultaneously – as in a living room or a furniture store. Conversely, one can argue that we don’t actually imagine “chair” at all, but rather a dinner chair or a comfy chair (itself understood as a variety of different parts, such as legs, imagined simultaneously). This would seem to make both “furniture” and “comfy chair” basic, but not “chair.” The reason for this ambiguity is that, again, “basicness” varies with topic. While chairs are a highly prototypical sort of furniture, “chair” is also a schema of which particular subtypes of chair (for example, dinner table chairs) are more prototypical than others (for example, dentist chairs) – a point that may be repeated at every level. Of course, there may be a “basic level” in the sense of a default level, a level at which we tend to operate spontaneously. In fact, that seems pretty well established. But that does not fix these levels in any theoretically rigorous way. It merely says that we will tend to think about the world at a certain level, and thus, that our default use of “prototype” may refer to that level. But we can and do shift levels, and our three sorts of categorization follow such shifts.

Testing, Revision, and the Program of Research

185

On the other hand, whether or not all happiness prototypes are reanalyzable in this way, it is clear that these three prototypes do not exhaust all specific happiness-related goals. Moreover, it is clear that all traditions include highly esteemed works that do not fit any of the prototype structures. My own intuitive sense (for whatever it is worth) is that the three genres cover perhaps two-thirds to three-quarters of canonical and popular narrative, with the remaining works not fitting into any particular cross-cultural pattern. The proportion is toward the high end in popular narrative, perhaps even surpassing this range; it may be closer to the two-thirds mark in canonical narrative. Within these percentages, the proportions of each genre seem to shift with the circ*mstances of the poets and their societies. Hierarchical, militarized societies in which the poet is patronized by a ruling elite tend to produce a higher percentage of heroic plots. Societies in which narratives are produced for broader, popular consumption particularly emphasize romantic plots. Societies in which famine is a constant threat have a higher percentage of sacrificial plots. Of course, at this point, comments on exact proportions are conjectural. Moreover, it is very difficult to say just how one would determine such proportions rigorously. Canons do not have strict boundaries. No one can give an exact count of all the members. Finally, I could be mistaken about the minority of nonuniversal stories. Perhaps they fit some universal pattern which I have simply failed to isolate. In any case, the crucial point is that the three genres may be exhaustive for universal, prototype narratives – as I am hypothesizing – without in any way exhausting all the stories that can be told, all highly esteemed narratives, or all happiness goals. In short, stories come in countless varieties. However, these three structures are, it seems, uniquely important. They form the core of what we mean by prototypical narrative. They constitute the majority of the stories that inspire our collective admiration. Finally, they help guide not only story telling, but our reception of stories, and our The points raise interesting issues for the present analysis, such as the degree to which there are patterns to prototypicality at lower levels than that of story. In other words, story prototypes appear to be of just three sorts – romantic, heroic, and sacrificial tragi-comedy. Do subprototypes for, say, romantic tragi-comedy fall into any significant patterns? Issues such as this could obviously be pursued as part of a research program in this field.

186

The Mind and Its Stories

thought about other matters (such as religion) as well. Once again, I take their prominence to be an absolute or at least near absolute universal. But they simply do not define all stories. Indeed, that is in many ways the point of speaking about prototypes rather than necessary and sufficient conditions. Only necessary and sufficient conditions would define all stories. With these qualifications, we may now turn to our third prototypical structure.

sacriﬁcial tragi-comedy As I just noted, a number of Ainu epics treat the genesis and overcoming of famine. Some of these incorporate concerns about famine into heroic plots such that the agent of famine is some enemy who must be defeated by the hero. In itself, this could be viewed as a variation on the heroic structure. However, there are several stories that treat famine separately. This suggests that, in the heroic cases just mentioned, famine has been incorporated into the heroic structure in the way romance is often incorporated into an heroic plot and heroism is often incorporated into a romantic plot. In other words, it may be that the famine motif is part of another recurrent structure, parallel with romantic and heroic tragi-comedy. This makes sense within our general explanatory framework. Just as there are personal, social, and transcendental varieties of happiness, there should be a form of bodily or physical happiness as well. For most of humankind, throughout most of history, it is very likely that the physical prototype for happiness has been plenty while the correlated prototype for sorrow (which would define the middle or progression in relevant narratives) has been famine. In connection with this, the sustaining emotion of a narrative generated from this prototype would be hunger. (It may seem inappropriate to call hunger an emotion. I shall return to this issue in the afterword.) This may be less intuitively plausible than the romantic and heroic cases. However, our intuition here is likely to be the result of habituation. While hunger is still very common today, most people who can concern themselves with cognitive science and literary theory simply expect plenty as an ordinary condition of their lives. Thus it would be surprising if the force of plenty as a prototype for happiness had

Testing, Revision, and the Program of Research

187

not become dulled. The point finds empirical support in research by Harris, Olthof, Terwogt, and Hardman. These researchers studied children’s views about “the situations that provoke emotion,” which is to say, the eliciting conditions for emotion. They investigated a set of European children and a set of Nepalese children from a remote area. They found that the children in general responded similarly. However, the Nepalese children were considerably more likely to bring agricultural “anxieties” into their accounts and to associate “pleasure” with “foods” (338). For example, they report only two eliciting conditions for happiness from the Nepalese children. The first one, translated from a seven year old child, is “With good things to eat, people feel happy” (334). Needless to say, this research is not the only empirical source of support for such a hypothesis. Literature provides many cases, cross-culturally. Returning to the Ainu tales, we find that, in the nonheroic famine stories, famine is ended by the benevolent action of a deity, who awards food to the hungry people after petition and/or some sort of sacrifice. In each case, the award of food is followed by a ritual. More exactly, the human community has committed some ethical or spiritual violation that offends the gods – typically a violation in the treatment of food. This has given rise to the famine. In consequence, they have to make some sort of offering to the gods and/or send a messenger, petitioning for divine intercession. The gods grant humans their food, introducing some ritual that will preserve their well being. For example, in “Song of the Goddess of the Waters,” the goddess receives a goblet of wine as an offering from the humans. Through the wine, Okikurmi speaks to the goddess, explaining that there is famine and that he has used his “last grains” to make the wine for this offering in the hope that she will help. The goddess invites a number of deities to a feast and speaks to them about the famine. She learns that the humans have treated various gods with disrespect through their actions toward deer and fish. She restores food to the humans, but simultaneously explains that they must practice specific rituals, and they must make a propitiatory sacrifice of wine and ritual wands to the offended gods. “Song of the Owl God” more or less follows the same structure. It begins with a strange sequence of events in which the Owl God asks a crow and a jay to deliver a message to heaven.

188

The Mind and Its Stories

Each one falls asleep when listening to the message. The Owl God reacts by killing them. Then the “Dipper Boy” agrees to take the message. He is fully attentive, learns the message, and leaves for the land of the gods. The message concerns a famine that is occurring among the humans. The gods explain that the famine has resulted from the insulting manner of the humans in their treatment of game and fish. The Owl God teaches humans the proper rituals and the famine ends. Other Ainu poems include at least parts of this sequence as well. Of course, it does not follow from these few cases that there is a third cross-culturally prototypical narrative structure. One might simply count this as an oddity of Ainu orature. After all, a certain percentage of stories in any collection will not fit the romantic and heroic prototypes. There is no reason why some of the nonprototypical tales might not form themselves into a culture-specific pattern. However, in reading these stories, I was struck by the fact that this particular structure recurs in culturally significant narratives elsewhere. For example, one of the most important stories of the Yoruba people of west Africa concerns a famine brought on by the greed of “Land” with respect to a small bit of game. Land had been hunting with Heaven and they killed a mouse. Land refused to give the mouse to Heaven. In consequence, Heaven withheld rain. Due to the terrible drought that ensued, people had to sacrifice to heaven. In Amos Tutuola’s retelling of this story, they used a human “carrier,” a man who could transport the offerings up to Heaven, asking that Heaven allow rain to fall once again (301–2). The sacrifice is highly ritualized and of a piece with actual Yoruba practices. Heaven accepts the sacrifice and rain falls to earth, ending the drought. Though the story speaks of the error as made by “Land,” it is clear that “Land” here represents the human community. The greed of the human community, and its disrespect for “Heaven,” led to the famine. Only ritual sacrifice, which allows the carrier to convey prayers and offerings from humans to the gods, can restore food. (Tutuola’s novels actually include numerous sacrificial stories, drawn from Yoruba sources. For an analysis of The Palm-Wine Drinkard in these terms, see Chapter 3 of my Empire.) The general structure of these “sacrificial” narratives appears to be something along the following lines. There is a violation of divine prescriptions, hence a divine/human conflict. The violation may be

Testing, Revision, and the Program of Research

189

individual, but its results are collective. It commonly concerns food and results from overstepping one’s proper (human) bounds through arrogance and/or greed. This violation leads to famine, the loss of food. (We will see later that this can be represented as a sort of exile, in keeping with the structure of romantic and heroic tragi-comedy.) Death and imagery of death follow. The community must engage in a sacrifice that serves to reconnect the human and divine worlds. This is frequently associated with the institution of a ritual bearing on food. The result of this sacrifice and ritual is the restoration of plenty by a now benevolent deity. The sacrifice may be a “blood sacrifice” or not; if a blood sacrifice, it may be the sacrifice of an animal, but, in literature, it is often the sacrifice of a human.4 4

A structure bearing some of these features, especially that of sacrifice, has been isolated by a number of writers, in very different theoretical contexts. No doubt the most famous discussion of this sort is by Ren´e Girard. Readers interested in pursuing the topic of sacrifice would certainly be well advised to consult Girard’s seminal Violence and the Sacred. However, Girard’s fundamental propitiatory view of sacrifice is entirely different from mine. Specifically, Girard takes the propitiatory purposes of sacrifice to be a sort of surface appearance only. The real social function of sacrifice is to release violent feelings that, if left uncatharted, would tear apart the whole society. Early in his book, he maintains that the “theological basis of the sacrifice” is a functional “misunderstanding,” for “The celebrants do not and must not comprehend the true role of the sacrificial act” (7). Specifically, “The sacrifice serves to protect the entire community from its own violence” (8). I disagree with Girard on two points. First, it seems clear that all peoples seek some ethical explanation for disaster. It is a human tendency to posit moral blame for famine, plague, and so on, beyond simple physical causality. We explain normal events by normal causal sequences. We explain extraordinary events by abnormal casuality, which almost always has an ethical component. (Indeed, one source of moral feeling probably is just this sense of bafflement when faced with abnormal events, especially very painful events.) It is a human tendency to see extraordinary suffering – for example, that produced by famine – as punishment for immorality. Sacrifice arises almost predictably in this context as a reparation aimed at ending that punishment. Thus, I do not see the “theological basis” as a mere manifest content or ideology. Certainly any social act has many motives. For example, there may be an enjoyment of cruetly that contributes to sacrifice. In actual human sacrifices, there are probably political motives as well, individual jealousies, and so on. However, the theological explanation is part of a genuine belief, a belief that is bound up with important feelings and goals. My second disagreement with Girard concerns catharisis. Girard’s assumption is that one act of violence reduces our tendency to commit further acts of violence. But this seems to be empirically mistaken. If Girard were correct about sacrifice, capital punishment should have much the same effect. It should not only frighten potential murderers. It should in general reduce the violent tendencies of the population. But, in fact, just the opposite seems to be the case. In general, capital punishment shows

190

The Mind and Its Stories

cross-cultural evidence for sacriﬁcial narratives But is this new category merely an ad hoc addition? It does cover some instances from Yoruba as well as the Ainu material, but that is hardly compelling. In fact, once one notices the sacrificial structure, it is clear that there are literary instances in a wide range of cultures. In other words, its addition is not ad hoc, but extends the explanatory reach of the theory beyond that of the initial data. Oedipus the King is a particularly prominent example of this structure (though, in combination with its sequel, it is also part of an epilogue of suffering, as noted previously). This play, “universally recognized as . . . the most brilliant example of theatrical plot” (Knox 131), begins with a famine in Thebes: “Thebes is dying. A blight on the fresh crops/and the rich pastures, cattle sicken and die” (Sophocl¯es 160). The fault is that someone killed King La¨ıus and the land will not be fertile again until the murderer is punished – in this case by exile. Individual expulsion from the community is a common variant, substituting for actual death in some instances of the sacrificial structure. (This is distinct from, but perhaps related to the communal exile that is sometimes associated with famine, as mentioned previously.) Specifically, the scapegoat is often seen as bearing something desirable to the offended deity as a sort of appeasem*nt. However, he/she may also be understood as purging the society by transporting the sin to another place. In the former case, the victim is more likely to be killed, so that his/her spirit may ascend to the deity; in the latter case, he/she is likely to be exiled. As the oracle says, “Drive the corruption from the land/don’t harbor it . . . don’t nurse it in your soil – root it out!” (164). Euripid¯es offers instances as well. Acts of human sacrifice often appear in Euripid¯es as part of heroic tragi-comedy. However, some instances of sacrifice in his plays are more clearly part of the sacrificial structure (even if they continue to draw on the heroic plot as well). no deterrent effect. Indeed, it seems to encourage violence. As Amnesty International points out, “The death penalty never has shown to benefit a society. In fact, there are strong indications that it increases people’s tolerance of and tendency toward violence.” Thus “The Bowers-Pierce study, analyzing executions between 1907 and 1963, concluded that an average of two additional homicides were committed in the month after an execution took place.” In short, this basic premise of Girard’s analysis of sacrifice seems mistaken.

Testing, Revision, and the Program of Research

191

One good case is the Phoenissae. In this play, T¯ıresias explains that ¯ es demands Cadmus “had killed the dragon born from the earth.” Ar¯ “the blood of a descendant” of the men who sprang up from the dragon’s teeth (O’Connor-Visser 76). For complex reasons, the only descendant who will satisfy this demand is Menoiceus. If Menoiceus is not sacrificed, Thebes will be overrun by the invading army. The explicit reason for the sacrifice, a reason borrowed from the heroic structure, is not hunger. However, the link with the preceding, prototypical structure, is not difficult to discern. The original violation is a crime against the earth, and Menoiceus “had to be sacrificed to the earth” (84). Indeed, J. P. Gu´epin has argued that the play involves a detailed agricultural allegory (207–8). Moving outside Europe, we find numerous instances in a wide range of cultures. Consider, for example, The Injustice Done to Tou Ngo, a drama by the most renowned Chinese dramatist, Kuan Han-ch’ing. In this work, the death of Tou Ngo is a violation that leads to drought and requires reparation before abundance is restored. A Japanese example may be found in the Kabuki play, Narukami. In this play, a recluse has been snubbed by the emperor. As an act of vengeance, he captures the God of Rain and imprisons him. Halford and Halford summarize the results: “No rain falls. The country is on the verge of drought and starvation. . . . As a last resort, Taema-hime, a princess of the Imperial house, offers to sacrifice herself” (231). Thus, we have an initial violation, followed by drought and famine, rectified by a sacrifice – or possible sacrifice. In the end, Taema-hime does not sacrifice herself, but overcomes Narukami’s power through her seductive charms (for a fairly full version of the play, see Saint Narukami and the God Fud¯o in Brandon); nonetheless, the basic structure is present.5 The structure appears to be particularly common in Native American narratives, at least those that have come to be seen as paradigm works. John Bierhorst picks out the Mayan Cuceb and the Nahuatl/Aztec Quetzalcoatl as, each, “the leading example,” which is to say the central paradigm work, “of the cultures that produced 5

Interestingly, the sacrificial structure is integrated into a heroic structure. Narukami is in league with the emperor’s usurping brother and his actions are designed to force the emperor to resign. Moreover, it is joined with a classic romantic plot as well. Taema-hime is able to overcome the objections of her rival’s father and marry her beloved precisely because she manages to restore rain to the kingdom.

192

The Mind and Its Stories

them” (xi). Cuceb, a “ritual epic” (187), is a dense and complex piece. But the general structure is clear – and it is a sacrificial narrative, partially combined with a heroic narrative. The heroic section concern’s the “people’s deliverance from an alien ruling class, the so-called Itz´a” (187). But, what is important here, is that this is combined with a sacrificial narrative in which the rule of the Itz´a is a period of “drought” and famine which is “brought about by the sexual sins of the people” (224). The drought and famine are stressed repeatedly:“It will be then that the water sources are dry”; “Their bread/will be snatched away, the water snatched out of/their mouths”; “The earth/will burn, the sky will burn” (196; the poem is cast as a prophecy, hence the future tense). Subsequently, we read about “Days of hunger and days of thirst” (215), “for great is the punishment of guilt” (216). This leads to a partial exile as well when “The drought-stricken Maya, driven from their towns, roam through the forest in search of food and water” (230–1). However, eventually, the “ceremonial penances” surrounding “the human victim,” who “is purged,” make “the heavens open and the (fruitful) rains fall” (237). Specifically, “Bound are his eyes . . . the poison is summoned . . . [a]nd he will rise up to a different world” (220–1). Though Bierhorst refers to a human victim, the poem subsequently refers to “the death of . . . the deer, a/sudden death” (221), the deer “stand[ing] for the human surrogate” (265), in Bierhorst’s view. In any case, this is a straightforward sacrificial narrative, with sin giving rise to famine as well as exile, and food restored only after a ritual sacrifice. As to Quetzalcoatl, Bierhorst explains that there are five extant fragments of the poem. The first fragment concerns the origin of food and the theft of food by the rain gods (20–1). As Bierhorst notes, this “explains how agriculture became dependent upon the rain gods” (71n.12), which is to say, how the satisfaction of hunger became dependent upon rain and thus on the will of deities. This clearly prepares for the sacrificial narrative. The second fragment is not really part of the same sacrificial sequence. It is a brief heroic plot in which Quetzalcoatl’s father, the Sun, has been murdered by Quetzalcoatl’s uncles. There is a sort of feint, suggesting sacrificial narrative, when Quetzalcoatl asks “What shall I sacrifice to the temple?” (22). However, the sacrifice is a trap by which he captures and kills the usurping uncles. He goes on to conquer many lands. In a

Testing, Revision, and the Program of Research

193

very truncated version of the epilogue of suffering, he falls ill when he reaches Tlapallan, “the mysterious land of the rising sun” (74), where he dies quickly and is “consumed in flames” (24). The third fragment is more complex. It focuses on Quetzalcoatl’s new religion, including his opposition to human sacrifice and his practice of offering “only snakes, birds” instead (28; “butterflies” are also added to the list). The important point here is that Quetzalcoatl himself becomes the perfect sacrifice in this scheme, for he is the “plumed serpent” (4), both snake and bird. Subsequently, Quetzalcoatl is seduced into drunkenness and perhaps incest (81n.45). He is “filled with remorse” (34). This transgression (involving food – or, rather, drink) leads to his exile and self-sacrifice. Specifically, he “traveled” and “ventured widely.” Eventually, “he set fire to himself” (36). In this fragment, the sin and reparation are presented as wholly individual. They do not appear to have any social consequences. In the fourth fragment, however, the social implications become clear – and they are precisely the sort we would expect. Initially, during Quetzalcoatl’s time there was “great abundance” of “foods” and of “all the crops” (39). His people “never wanted. There was/nothing lacking in their houses, they were never/hungry” (40). However, due to the fact that “Quetzalcoatl and all the/Toltecs grew lax” (41), this period of plenty and ease ended. Most importantly for our purposes, “Stones . . . rained down” (55). Bierhorst explains that “A rain of stones means barrenness, presumably brought about by drought” (88–9n.80). Bierhorst’s explanation is confirmed by the following events, in which people are sacrificed according to “the drought rituals held in Aztec times” (89n.81). The sacrifices are only partially successful. There is food, but it is “bitter,” so much so that “it could no longer be/placed on the lips” (56). The idea relates to the Quetzalcoatl’s forbidding of human sacrifice and substitution of birds and snakes – which is to say, himself. (The metaphorical point of the food being inedibly bitter after the human sacrifice seems too obvious to require spelling out.) The following section of the poem concerns Quetzalcoatl’s exile and death – or, rather, his implied death, for the fourth fragment does not fill in the death itself. Thus, we must return to the third fragment to see this as a ritual sacrifice. In the fifth fragment, which is clearly the end of the poem, we are told that Quetzalcoatl has departed and that “the variegated maize . . . came forth” (65).

194

The Mind and Its Stories

Related to the cross-cultural repetition of this plot structure, in researching lyric, I had come upon quite a few poems, especially from oral cultures, that addressed agricultural fertility or livestock. For example, Firth and McLean cite a number of Tikopia songs that deal with great abundance of fish (134), or with herons depleting the fish supply (133), with “a time of grave food shortage” (130), and so on. Deng notes the centrality of cattle to Dinka life. They “provide an essential part of their food supply” (30) and also serve “as the medium of reconciliation between man confronting God and ancestral spirits through sacrifice” (31). A very large number of Dinka lyrics concern cattle (see 96–158 – over one-third of Deng’s collection). This distribution makes sense, and fits the narrative hypothesis perfectly, given our third prototype narrative. Otherwise, it would appear surprising. In addition to these cattle songs, there are songs of “offerings,” sung “during sickness, war, drought, famine” (238). One of these lyrics makes the general point of sacrifice, “If I wrong Him/I make it right” (239). Deng explains that this “making right” is accomplished “by sacrificing or consecrating an animal to the offended spirit” (239n.6). Numerous Native American poems address deities of corn or call for rain. For example, Cronyn’s classic collection (see Lincoln) presents many instances. From the Sia, he includes a “Prayer for Rain,” a “Rain Song of the Giant Society,” and an “Invocation” that adjures “Bring your showers and great rains . . . /I throw out to you my sacred meal . . . /All come out and give us your showers and great rains” (64–5). There is a Pueblo poem “announc[ing]” a “feast” and calling on the clouds “that the planting may yield abundance” (81). ˜ “Song of the Blue-Corn Dance,” also calling on the There is a Zuni clouds (82), and a “Corn-Grinding Song” (123). There are Hopi lyrics about “the blossoming virgin corn” (124) and “Patches of beans in flower” (125), a Teton Sioux “Song of the Famine,” and so on. A very fitting sequence of lyrics may be found in the renowned Navajo Night Chant. Bierhorst explains that “In its narrowest use,” the Night Chant “serves as a form of therapy, conducted by a shaman . . . for the benefit of a principal communicant” (281). The poems, directly integrated into the loosely narrative structure of the ritual, recur continually to rain and agricultural fertility. The “first morning prayer ritual” refers to rain in the opening stanza (293). The communicant’s first speech begins “Owl!/I have made your sacrifice” (294). After calling for a

Testing, Revision, and the Program of Research

195

restoration of his/her physical well-being, he/she goes on to call for “abundant dark clouds,” “abundant showers,” and “abundant vegetation” (295). This mixing of the physical well-being of the individual with agricultural plenty for the society indicates again that the prototype for physical happiness is having plentiful food. Subsequently, the various poems make repeated reference to rain, the “sacred blue corn-seed” (313), “The white bean/And the great corn-plant” (314), “The voice of the thunder . . . that beautifies the land” (318), and so on. The Night Chant ends when the deity “takes up his sacrifice” (331; see 346n.67) and “The corn comes up, the rain descends. . . . /The rain descends, the corn comes up” (346n.65), and of course the well-being of the communicant is assured. As this suggests, lyrics are often explicitly embedded in sacrificial narratives, including rituals. A striking case of this may be found in the Kondh lyrics used in the ritual of human sacrifice. (The Kondh are one of the “tribal” groups of India.) These poems punctuate the ritual, which is itself an implicit plot. They include a song concerning the acquisition of the scapegoat, the “first stab by the priest” (Mahapatra 1), and, most revealingly, the “Invocation to Earth,” that begins “Let no famine/Visit our land”; instead, allow “Our crops” to “flourish” in “plenty” (2). (For a discussion of the sacrifice itself, which has not been practiced since the nineteenth century, and its agricultural purposes, see Mahapatra xviii–xlii.) Other ritual examples may be found in Christian services and various Central Asian shamanistic rites (see, for example, Chadwick and Zhirmunsky 248–9). Perhaps even more interestingly, like heroic and romantic structures, sacrificial plots are also to be found in historiography. A ˜ ına in Kalhan.a’s twelfthstraightforward example is that of Tunj¯ ˙ . ¯ı. There is famine in century history of Kashmir, the R¯ajatarangin Kashmir. The king says to his queen, “it is through some transgression of ours . . . that such a calamity . . . has befallen the innocent people” (53). His solution is just what we would expect, “Therefore . . . I shall now sacrifice my body in the blazing fire” (54). As it turns out, this is unnecessary, for the queen is able to end the famine, evidently through her “pious conduct” (55). This pious conduct is centrally a matter of “Devotion to the husband” (55). In keeping with this, many years after the famine, the queen commits sat¯ı, which is to say, throws herself into the funeral pyre of her

196

The Mind and Its Stories

husband (55) – thus, in effect, performing the sacrifice that he had not performed at the time of the famine. No doubt the most commonly known sacrificial narrative, worldwide, is the Judeo-Christian story of the Fall and Redemption of humankind. The story of Jesus might at first seem quite different from the Yoruba story of Land and Heaven or the songs of the Water Goddess and the Owl God. However, it is, in fact, closely related to them. It, too, is a variation on the basic sacrificial structure. The full story begins in “Genesis.” Adam and Eve offend God when, acting from pride and greed, they violate his one prohibition, a prohibition regarding food: “of the fruit of the tree in the middle of the garden God said, ‘You must not eat it, nor touch it, under pain of death’“ (3.3). In keeping with the standard pattern, their individual violation has broad, communal consequences – indeed, universal consequences, for it affects all of humankind. First, it “Brought Death into the World,” as Milton had it (1.3). While the Ainu and Yoruba stories concern particular deaths, the Judeo-Christian story concerns death in general, death in all its forms. Moreover, this death is bound up with hunger. When God created the garden of Paradise, the first thing he did was produce food: “there was as yet no wild bush on the earth nor had any wild plant yet sprung up, for Yahweh God had not sent rain on the earth.” Then “Yahweh God planted a garden in Eden. . . . Yahweh God caused to spring up from the soil every kind of tree, enticing to look at and good to eat” (2.5, 8–9). God’s first words to Adam and Eve concern food – along with sexual union and social domination, thus covering our three prototypes: “Be fruitful, multiply, fill the earth and conquer it. Be masters of the fish of the sea, the birds of heaven and all living animals on the earth. . . . See, I give you all the seed-bearing plants that are upon the whole earth, and all the trees with seed-bearing fruit; this shall be your food” (1.28–30). In keeping with the general structure of the primary narrative modes, the way in which death enters the lives of Adam and Eve is, first of all, through exile – exile from this garden, which is defined by its abundance of food. The words in which “Genesis” expresses this exile stress precisely the loss of this food: “So Yahweh God expelled him from the garden of Eden, to till the soil” (3.23). Indeed, life itself is identified with a particular food in this lost garden, the fruit of the “tree of life,” from which humankind is particularly forbidden (3.22).

Testing, Revision, and the Program of Research

197

Finally, this exile involves two particular curses from God. The second is the crucial one for our purposes. Though not a curse of famine, it is a curse that allows famine: “Accursed be the soil because of you./With suffering shall you get your food from it” (3.17). In Christian teaching, this expulsion was so absolute that it continued not only in earthly life, but after death as well. It was an absolute and all-encompassing condemnation. Despite this, eternal life could still be regained. God could grant the gift of life, return us to the plenteous garden. But this could occur only after a sacrifice. The narrative continues millennia later with Jesus. Though God, Jesus is incarnate. Thus, his sacrifice is a sacrifice on behalf of humanity. Typically the sacrifice itself has to come from humans, though it is often bound up with the aid of a benevolent deity, as we have seen. Jesus combines these two roles, taking divine benevolence one step further by sacrificing himself. Like the Yoruba carrier, then, Jesus is a sacrificial victim who communicates between earth and heaven, putting an end to the curse of eternal death and exile from abundance, just as the sacrifices of the Ainu and Yoruba stories put an end to the more local and mundane death resulting from a particular famine. As part of this sacrifice, Jesus institutes a ritual in which the sacrifice itself is reenacted, specifically through the consumption of food. (Note that in the Ainu stories, the ritual is initiated by the helper deity – the Water Goddess, in one case, the Owl God, in the other.) Images surrounding these events (for example, the characterization of Jesus as a sacrificial lamb) reenforce the link with more ordinary sacrificial stories treating famine, as do various peripheral elements, such as the common belief that it rained when Jesus died. Finally, Jesus’ resurrection serves as a concrete example of the way this sacrifice has overcome the lapsarian condition of death and led God to give life and plenty back to the damned society – at least in offering the possibility of eternal return to Paradise after death.6 6

Not all sacrificial narratives are ancient. A famous modern instance is William Butler Yeats’s The Countess Cathleen. In this story, the peasants of Ireland are starving (“the land is famine-struck” [2]). In consequence, they are selling themselves to the devil. Cathleen agrees to sacrifice herself in order for the starving peasants to be saved, both physically and spiritually. She sacrifices herself, is transported to heaven, and the peasants are restored. Yeats uses this structure not only to communicate religious points, but to make quite modern social and political points as well. For example,

198

The Mind and Its Stories

a note on ethics and sacriﬁcial tragi-comedy As we saw in Chapter 4, each prototype narrative structure raises particular ethical issues. The personal prototypical narrative, romantic tragi-comedy, raises the ethical issue of the degree to which society can circ*mscribe personal freedom. The social prototypical narrative the demons clearly represent the merchant classes and the capitalist order, while the countess represents the aristocracy and – along with the peasants – the feudal order. Thus Yeats uses the sacrificial structure to comment on capitalism, and on the place of the aristocracy and the peasantry in a mercantile economy. Needless to say, Yeats is not the only important modernist who draws on sacrificial tragi-comedy. For example, Kafka relies on this structure – though his treatment is far more elliptical and ironic than that of Yeats, as one would expect. A still more recent example may be found in Kamala Markandaya’s widely read novel, Nectar in a Sieve. The story concerns Rukmani and Nathan, a poor peasant couple in South India. At one level, their suffering is explained in straightforwardly materialist terms. Specifically, it results from the fact that they do not own their land and from the intrusion of heavy industry into the region. But the plot draws implicitly on the sacrificial structure as well. Rukmani and Nathan begin well enough. But Nathan kills a “sacred” cobra (20) among the ripening fruits and vegetables of the garden (18). After this, their hardships begin – prominently including drought, famine, and exile. The initial state of plenty is not restored for many years. Specifically, Nathan dies, in a temple (thus in a religious context), during a rain storm (recalling, and inverting, the drought that began their suffering long before). After this, Rukmani’s situation improves quickly. She not only has enough food, but she is able to return home as well. Nathan’s death is clearly an implicit sacrifice that allows for her well-being – and indeed for the well-being of the larger community, for when Rukmani returns home, she finds a hospital serving the poor people of the area. This is, of course, a modern version of the sacrificial structure. Though he dies in a temple, Nathan is killed by the wage labor he takes up in the city; he is killed indirectly by capitalists, not directly by a priest. Moreover, the new life brought about by this sacrifice is not merely a matter of agricultural conditions (though there is rain, as just noted). It is equally a matter of cooperative and rational work to help the poor. Yet, despite all this, the family’s improved condition is still a sacrificial restoration. It is a restoration of sustenance brought about by death in a temple. Moreover, this death compensates for an earlier sin against nature, a sin linked with food. In making her modern social and political points, Markandaya has not created a wholly new and modern narrative structure. Rather, she has varied and specified the ancient and universal structure of sacrificial tragi-comedy. These are not isolated cases. As already noted, Amos Tutuola takes up sacrificial narratives explicitly. This is true also for the Nigerian Nobel laureate, Wole Soyinka, who draws on sacrificial motifs overtly in The Strong Breed and Death and the King’s Horseman, implicitly in The Bacchae of Euripides, and still more subtly in The Swamp Dwellers (on the mythic implications of the last play, which indicate its sacrificial links, see my “Particular Myths”). Contemporary films make use of this plot as well. Santosh Sivan’s award-winning The Terrorist in effect sets out to undermine the militaristic use of the sacrificial

Testing, Revision, and the Program of Research

199

stages the contradiction between the ethics of defense and the ethics of compassion. In short, they raise personal and social issues respectively. The sacrificial plot most obviously involves obedience to God. Simply put, defiance brings suffering. In a sense, then, it is a straightforward narrative of moral conformity. However, there are complexities here. Sacrificial plots often raise issues about the proper treatment of nature, which is bound up with divinity in these narratives. They touch on what we would now see as ecological questions regarding humanity’s relation to the rest of the natural world. Specifically, they address the conflict between the human need to consume nature and the ethical imperative to respect or revere nature. In the Judeo-Christian story, for example, God begins by asserting human supremacy over nature – but he immediately circ*mscribes that supremacy. The point is even more obvious in the Ainu stories. Prototypical narratives also define morally admirable acts that balance their particular ethical violations. Heroic plots value loyalty as the reverse of usurping betrayal. Romantic plots value the assertion and pursuit of one’s individual affiliative choice against overbearing authority or social conformism. Sacrificial plots value self-denial as reparation for greed and arrogance. Of course, this is not just any sort of self-denial. Just as loyalty in heroic plots is prototypically a militant or combative loyalty, self-denial here is prototypically self-denial with respect to food. In other words, the ethics of sacrificial tragi-comedy is in part an ethics of self-abnegation that focuses on one’s own body and that, in effect, denies that body or, rather, makes it suffer for a higher good, repudiating the flesh that comes from self-indulgence in food. This ethical prototype manifests itself cross-culturally in

narrative by consistently inverting the plot and opposing sacrifice to fertility. In Luis Puenzo’s The Plague, the plague is ambiguously linked with “collective punishment” by a priest’s sermon. The priest subsequently sacrifices his own life among those who have died. Almost immediately after his death, the cold weather arrives and puts and end to the plague – a point made explicitly by the main character. Note that the shift in weather is directly parallel to the shift from drought to rain, even though the plot here has been varied from the prototype structure, with disease substituting for famine. This variation would not necessarily suggest a weather-based solution to the problem, were it not for the fact that the basic prototype is famine. The list of modern and contemporary sacrificial narratives could be very much extended – not as much as the lists for the other genres, but a great deal nonetheless.

200

The Mind and Its Stories

a range of practices from Hindu asceticism and Jain self-starvation (called “sallekhana” – for discussion, see Dundas 155–6), through Christian and Muslim fasting, perhaps to the modern, secular obsession with slimming, and even to certain aspects of anorexia.7

conclusion Our consideration of Ainu epics has to a great extent provided corroborating evidence for our previous hypotheses concerning narrative structure. However, by rendering salient certain motifs and patterns that are common cross-culturally, it has led to some changes in emphasis and formulation – prominently a greater stress on and elaboration of the love triangle in romantic tragi-comedy. More significantly, it has suggested that there is a third prototype for happiness, a physical prototype (plenteous food) and a correlated, physical prototype for sorrow (famine). This prototype, too, generates a set of universally recurrent narratives. In these sacrificial tragicomedies, some act of human greed or arrogance, usually regarding food, violates the sanctity of nature and offends some deity. In punishment, the society in question suffers famine, perhaps along with exile. The famine ends only when the society makes a sacrifice, often a human sacrifice. The human victim might be killed, thus transporting the sacrifice to the heavens, or he/she might be expelled from the society, symbolically carrying away the initial sin. This structure appears autonomously and in combination with heroic and romantic plots (as when martyrdom – a sacrificial idea – enters heroic tragi-comedy). It raises distinct ethical issues regarding human rights over nature and establishes repudiation of the flesh as a moral value. The isolation and definition of this third narrative prototype is, of course, tentative – like the isolation and definition of the first and second narrative prototypes. It is a step in a research program. It is not a final resolution of problems, but a hypothesis set out to account 7

I am not, of course, trying to explain anorexia through sacrificial narrative. I am simply noting that the ethical component of sacrificial narrative appears to bear on compulsive slimming, insofar as the concern with thinness is driven by ethical concerns, by a view of fleshiness not as unattractive, but as indecent, immoral, “obscene,” as Jean-Paul Sartre puts it (see 401–2 of Being, 407 of L’ˆetre).

Testing, Revision, and the Program of Research

201

for some patterns. As such, it should lead to further questions, further study, further testing, and so on. Indeed, this examination of Ainu epic has been a success – if it has – not insofar as it has reached definitive conclusions, but insofar as it has advanced a research program in such a way that the theories encompassed by that program can be further evaluated, criticized, revised, and extended in the future.

7 The Structure of Stories Some General Principles of Plot

One of the major contentions of the preceding chapters is that narratives are generated, not by complexes of abstract conditions, but by much more concrete prototypes. Obviously, there are exceptions to this. If a playwright determines to follow a five-act structure or an epicist accepts the notion that a heroic poem should begin in medias res, his/her work will be organized in part by reference to abstract conditions. However, the application of such abstract conditions is very limited. Indeed, one might argue that even in these cases we are mistaken in attributing a primary generative function to the general principles (“Organize the play into five acts,” “Begin in the middle of things”). For any author, these general principles are bound up with standard types and with particular instances. If I write an epic in twelve books, I have Milton and Virgil in mind. My conception of the twelve-book structure is shaped by their poems. In other words, here too my work is generated from more concrete precedents – in this case, exempla. Nonetheless, to say that narrative is most importantly derived from prototypes (and exempla) is not to deny the possibility of making general statements about narratives. Indeed, to say that narratives are generated from emotion prototypes is one such general statement. The purpose of this chapter is to articulate and develop some further generalizations about narrative, as they derive from the preceding analyses and bear on our understanding of the relation between narrative and emotion. In fact, it turns out that there is a great deal that 202

Some General Principles of Plot

203

can be said about narrative at this level, a great deal that is, I believe, both surprising and theoretically consequential. Before going on, however, it is important to stress two points. First, the following discussion is in no way intended to be a complete account of the abstract principles of narrative. Again, it is a treatment of some significant generalizations that concern narrative and emotion as analyzed in the preceding pages. Indeed, even on this topic, it barely scratches the surface. Second, these generalizations are not absolute laws giving strictly binding conditions. In fact, they remain closely related to prototypes. Specifically, the following generalizations aim at defining tendencies of narrative production and reception that move toward prototypicality. They do not set out to delimit what narrative is, full stop. They isolate the properties, relations, and structures that make narratives more prototypical. It is perfectly possible to have something that counts as a narrative, but does not fit these generalizations.1 This is not to say that such generalizations do not

It will be clear to readers familiar with “classical narratology” that my approach differs radically from that of “story grammarians,” such as Rumelhart and Prince. The goal of a story grammar is well articulated by Thomas Pavel, who explains that, “Properly constructed, an abstract narrative grammar should generate only the correct narrative strings” (12). There are several problems with this approach. First, it treats texts or utterances as if they are neatly divisible into two categories – stories and nonstories, with stories following the strict rules of story grammar and nonstories failing to follow those rules. This seems to me implausible. Rather, our judgment of “storiness” is a matter of degree. Again, it involves approximation to a prototype. Second, story grammarians regularly begin with the assumption that a particular linguistic theory of grammar, with its particular theoretical organization, applies directly to narrative. There is no reason to make this assumption. Yes, we do tell stories in language. But we also do physics and logic in language and we do not assume that we must use the same type of rules for language and logical inference. Brewer notes that “When psychologists first came into contact with generative-transformational linguistics, an attempt was made to convert it directly into a psychological theory. . . . After about a decade of work it became evident that linguistic theory could not be used in a simple fashion as a psychological theory” (“Literary Theory” 222). Lerdahl and Jackendoff make much the same point. Though they are speaking of music, the problem they identify is generalizable. As they put it, “many . . . applications of linguistic methodology to music have foundered because they attempt a literal translation of some aspect of linguistic theory into musical terms” (5). The claim holds equally if we substitute “narrative” for “music” and “musical” in this sentence. To make matters worse, Chomskyan linguistic theory has changed radically since its early formulations, taken up in narratology. Thus, the linguistic principles assumed by story grammarians are, most often, no longer accepted within linguistics.

204

The Mind and Its Stories

involve rules and principles. They certainly do. These are just rules and principles with a tendential form (for example, one rule might state that a particular property contributes to prototype status). The crucial point is that these are not rules in the sense of statements that give necessary and sufficient conditions for narrativity.

the telic nature of plot The most obvious general point that can be drawn from our analyses is that prototypical plots involve a person and a goal. This fundamental idea has been recognized by writers from the ancient Sanskrit theorists to Keith Oatley. It has also been noted by more narrowly empirical researchers. For example, Katherine Nelson maintains that

Admittedly, these objections apply more to some theorists than others. For example, they apply less to Pavel than to Rumelhart. Nonetheless, the basic difference in orientation remains. Put simply, story grammarians tend to seek necessary and sufficient conditions for storiness and they tend to understand necessary and sufficient conditions in terms of generative rules modeled on Chomskyan grammar. The prototype approach was designed precisely against the assumption of necessary and sufficient conditions. This is not to say that I entirely reject the project of the story grammarians. First, their attempt to examine story structure with precision is extremely valuable. Even if one rejects the claim that they have defined a grammar, they undoubtedly isolate some important patterns. For example, Rumelhart’s analysis of episodes in terms of events and reactions (see 214–18) is insightful and clarifying, no matter what one’s broader theoretical commitments. Second, I am not at all claiming that more general structures do not enter into narratives. They most certainly do, as I have noted repeatedly. For example, it is clear that we require scripts, of the sort famously isolated by Schank and Abelson, to understand both everyday life and stories. Suppose a story involves the statement, “I picked up the receiver. 785–4432. I waited. Nothing.” Clearly, we require a script for telephone use if we are to understand this passage. David Herman has made this case with particular force. But, again, all this is something different from story structure itself. My contention is that story structure most importantly derives, at least in its large contours, from cognitive processes that are prototype based at every stage. One main purpose of the preceding chapters, has been to outline those processes and support this contention. In short, though I find script theory enormously valuable, and I find many Chomskyan principles enormously valuable, I do not believe that the particular narrative phenomena considered here – the literary universals of tragi-comedy – are best understood in terms of scripts or generative rules modeled on Chomskyan transformational grammar. Rather, my argument is that they are best understood by reference to particular emotion prototypes. Moreover, generalizations that bear on these universals are themselves related to prototypes in isolating prototype-related tendencies, not necessary and sufficient conditions, as already noted.

Some General Principles of Plot

205

“At the most basic level, narratives consist of the report of a sequence of actions by actors that are connected in some way, usually because they are organized to achieve a goal or solve a problem” (5; in my usage of “goal,” solving a problem is one sort of goal). If Nelson is correct that the agent/goal structure is the most common structure and the one found in children, it makes sense that this would be prototypical. Of course, there is not only an agent/goal structure, but also some sort of causal sequence – as Nelson indicates, and as virtually every writer on narrative has noted for the last 2,500 years as well. In short, prototypical narratives have a telic structure including an agent, a goal, and a causal sequence connecting the agent’s various actions with the achievement or nonachievement of the goal. We should consider a few complications of each component. The Agent There has been a good deal written on the nature of the protagonist. For example, Bal points out that the hero tends to receive more descriptive attention, to appear more frequently in the narrative, to have relations with a larger number of other characters, and so on (132). At the same time, Frye notes that heroes can be almost null, with all the focus being on subsidiary characters – as in many New Comedies, where it sometimes happens that we hardly know or see the lovers, all our attention being drawn to the wily slave and the senex iratus (167). These are valuable points, worth sorting out in their different emphases and contexts. However, I should like to explore a few aspects of the protagonist that bear more directly on the preceding analyses. Socially Representative Characters and Individualistic Characters A number of writers have noted that we at least partially empathize with the protagonist. Even if we disapprove of him/her, we must in some sense share his/her feelings and interests if we are to be engaged by his/her pursuit of a goal. Again, this empathy is likely to be based on identity categories, such as nationality, or on shared situations, such as being in love. As the preceding examples indicate, romantic stories are likely to rely on situational empathy, while

206

The Mind and Its Stories

heroic stories are more likely to draw on identity categories. In fact, love stories often work against identity categories (such as economic class or nation) by putting the lovers on two sides of some identity conflict. Heroic stories, in contrast, are often formulated in the service of identity categories – supporting the nation, the emperor, or a particular ethnic group. Or, rather, the construction of romantic plots relies from the beginning on situational identification and that of heroic plots relies from the beginning on categorial identification. However, situational identification enters into any engaging literary experience. As discussed in Chapter 2, literature is structured in such a way that its emotional effects are founded upon the priming of emotion-laden memories. Our reading practices, as we approach literary works, reenforce this tendency. In other words, the nature of literature and of the attitudes we adopt to literature tend to stress the personal memories that foster situational empathy. In this way, situational empathy is a likely product of any literary reading. Indeed, this is no doubt one reason why so many writers have felt that literature is humanizing. It humanizes us in the sense that it tends to develop certain sorts of compassionate identification. (Of course, it is not at all clear that this sort of identification extends beyond the literary work to the real world.) This is particularly striking in the case of heroic tragi-comedies, for, once again, in those cases the situational empathy very often runs contrary to the categorial identification that is central to the nationalist or other propaganda purposes of such works. Indeed, the situational empathy is often most significant with respect to the “enemy” and not with respect to the hero, at least in those cases where the hero achieves his/her goal, while the defeated enemy is left impotent and miserable. The point about the relation between hero and categorial identity is worth dwelling on for a moment. To account for the different sorts of protagonist in prototypical stories, we need to draw a distinction between heroes – or, more generally, characters – who represent social categories and the norms they involve and heroes or characters who represent individuality and personal affiliation especially against social categories and the norms they involve. We may refer to these simply as socially normative and individualist characters. The difference between individualist and socially normative characters is not

Some General Principles of Plot

207

that one has individual characteristics and the other does not. Rather, the difference is a matter of whether the character in question stands for or against particular identity categories and the social hierarchies these entail. The hero – or, rather, heroes – in romantic tragi-comedies are most often individualist. We will be turning to other characters in a moment, but even at this point it should be clear that blocking characters, such as the forbidding father in New Comedy, are socially normative. Speaking of romantic comedy, Northrop Frye explains that “The opponent to the hero’s wishes, when not the father, is generally someone who partakes of the father’s closer relation to established society” (164–5). Indeed, as just indicated, it is not merely a closer relation. Characters of this sort represent the rules and judgments of society. They stand for class, nation, race, or religion. Note that the hero in romantic stories may be very nonindividualized (that is, he/she may have few particular, “individualizing” characteristics), but he/she is nonetheless individualistic in this sense. Conversely, a blocking character may be drawn in considerable, idiosyncratic detail, though he/she is socially normative. Things are a little more complicated in the case of heroic tragicomedy. Again, the genre itself tends to be used – cross-culturally and trans-historically – to support particular sorts of categorial identity. For this reason, the hero is almost always to some degree socially normative. However, the hero in this genre often has individualistic elements also. Indeed, sometimes, there is a split between the king or emperor, on the one hand, and the warrior, on the other, as with Kay ´ and Rostám in the Shâhnâme. In these cases, the king is most Kavus often a “pure” representative of the social order. The warrior is also a socially normative character, but he may have consequential individualistic elements as well. For example, Rostám clearly represents Iranian nationality against the Tur¯anian enemy. However, there are sometimes very strong individualistic strains in his character and ac´ bidding, then refuses tion, as when he refuses to come at Kay Kavus’s (however briefly) to serve the Shah in battle. A further complication here is that the pure social representative – whether split off from the partially individualistic warrior or not – is often characterized in such a way as to indicate that he/she does not really merit that role. ´ is made entirely clear by For example, the inadequacy of Kay Kavus Ferdowsi. He is no more worthy to make decisions for other people

208

The Mind and Its Stories

than are most New Comedy fathers. Yet at the same time Ferdowsi clearly supports loyalty to the Shah. Similar points could be made about R¯ama in the R¯am¯ayan.a, the P¯an.d.avas in the Mah¯abh¯arata, and so on. Indeed this inadequacy is part of the conflict between categorial and situational empathy in heroic plots. This suggests that categorial/ situational conflict is more pervasive in the heroic genre than is at first apparent. Though it only becomes obtrusive through the epilogue of suffering, the conflict may be there, in a more submerged form, from the outset. In connection with this, an entire group of prototypical narratives results from shifting the warrior protagonist in heroic tragi-comedy to a fully individualistic character. This is, for example, what happens in stories that fall under the broad category of Romantic Satanism. To take the most famous instance, insofar as Satan is the hero of Paradise Lost, we have a hero who is entirely individualistic. He rebels against the socially normative God the Father in much the way a lover rebels against an oppressive parent in romantic plots. Karn.a’s choice of personal affiliation for Duryodhana over his birth brothers – and his related war against these rightful rulers – is a case of this general sort as well. Though Karn.a is not “the hero” of the Mah¯abh¯arata, his heroic status is indicated at different points in the poem, including the end, where he appears in Heaven with the P¯an.d.avas. This status is also presented in other works, such as Bh¯asa’s early Sanskrit play, Karn.a’s Burden, in which Karn.a is the tragic hero. More generally, Karn.a has a status in Indic literature not entirely unlike that of Milton’s Satan in European literature after Romanticism. I should note that this structure is different from that found in narratives of social rebellion, which provide another variation on the heroic plot, still of very high prototypicality. In stories of social rebellion, the author splits the hero and the emperor or other official leader, making the hero socially normative, but of a different group – a different nation, ethnicity, economic class, religion. Unsurprisingly, narratives of this sort are not common in feudal societies, except in cases where the official ruler is viewed as a usurper. An example of this sort would be Chikamatsu’s The Battles of Coxinga, where the emperor is a Tartar and the heroes, seeking to restore the rightful (Chinese) heir, are Chinese. On the other hand, one might argue that this is true even for, say, Marxist works, in which the bourgeoisie

Some General Principles of Plot

209

holds social power through a sort of usurpation of the rights of the workers. These two variations on the heroic structure are particularly striking because parallel variations in the romantic structure are extremely uncommon. Prototypical romantic plots are already in effect Satanic and rebellious. Thus the only evident parallel to the rebellious variation (in which the roles of rightful ruler and rebel are reversed) would involve making the parents, not the children, into the heroes. The parallel to the Satanic variation (in which the – otherwise socially normative – hero is made individualistic) would presumably involve making the – otherwise individualistic – lovers socially normative. As to the former, there are a few, modern South Asian works that present a child’s desire for love marriage as a mistake and the social preference for arranged marriage as preferable. But these are fairly rare. The second variation, which shifts the lovers from individualistic to socially normative, does not seem to exist – at least not in a prominent enough form to be noticeable. This dissymmetry between heroic and romantic plots suggests that narratives generally drift toward individualistic characterization of heroes and they tend to favor individualistic over socially normative characters. There is also a more limited preference for rebels over rulers, as rebels may be heroes in either genre, but figures of social authority are heroes almost only in heroic plots. The tendency to favor rebels may be simply the residue of the preference for individualism. But what leads to this preference? In part, this results from the situational empathy discussed above. It is also related to the general narrative tendency toward personalization – the transformation of objectified characters into subjective persons (the main reason it is difficult to sustain villainy, as we have noted). Perhaps it is overly optimistic to say so, but this also seems to suggest a tendency of narratives to drift toward expanding individual freedom relative to social norms. This would appear to be an odd tendency in and of itself. However, it is related to the increase in personalization and empathy. As we develop greater emotional identification with a character, we automatically come to view his/her preferences as more human and valid. Thus, we come to see inhibitions on his/her actions as less fair or reasonable – or, at least, as less unequivocally good, even when necessary. In short, authors – especially authors of

210

The Mind and Its Stories

heroic plots – may set out to create works that favor categorial identity and social hierarchy by celebrating socially normative characters and repudiating individualistic characters. However, the development of narrative tends to pull against this. The pull may be weak in many cases and thus have little noticeable impact (that is, it may not overcome an author’s intent of favoring categorial identity). But it appears always to be a pull toward individualistic characters and away from socially normative characters. In ethical terms, it is always a pull toward individual freedom, not toward social constraint. The situation becomes more complex when we consider the sacrificial plot in this context, for in the sacrificial plot these character tendencies are often fused in surprising and complex ways. The crucial characters in the sacrificial narrative are the violator, the offended deity, and the scapegoat. The agent around whom the action is organized and who sets out to save the society may be the violator or the scapegoat or someone else. Perhaps more importantly, the scapegoat may or may not be the violator. If the scapegoat is not the violator, then we have a, so to speak, “pure” sacrificial plot. In this case, the scapegoat is a socially representative character, which is to say, he/she stands for the entire community and purges all their sins. (Socially normative characters are always socially representative, but a character may be socially representative without thereby acting to enforce the norms of the society.) The violator is most often an individualistic character. When the scapegoat and the violator are identified, then the story is, so to speak, secularized. It is a story of retributive justice, not of communal guilt.2 The story of Jesus and the Yoruba myth of Earth and Heaven provide good examples of the pure sacrificial plot. To take the former, Jesus is the scapegoat, while Adam and Eve are the violators. Adam and Eve clearly acted individualistically, 2

It may be that the fullest version of the sacrificial plot incorporates both the pure and the secularized versions, with the former leading to the latter. The influential 1922 F. W. Murnau film, Nosferatu, appears to manifest a structure of this sort. Nosferatu is clearly an evil presence. He brings plague to the city of Bremen and may be seen as the guilty violator. One young woman sacrifices herself to save her fellow citizens. Once she dies, Nosferatu dies also. When these deaths occur, the plague ends immediately. Thus, we have the sacrifice of an innocent scapegoat, in keeping with the pure version, and the death of the violator, in keeping with the secular version; conjunctively, they produce the restoration of physical well-being to the entire society.

Some General Principles of Plot

211

rebelling against God’s law. When sacrificed, Jesus was socially representative, standing for all humanity. But he was not really socially normative – or, rather, he was socially normative for divine society, but an individualistic rebel for human society. (Recall the official reason for his execution – “Jesus was condemned to death . . . as a rebel against the Roman State in one of its subject provinces” [Cullman 6; see also 42–3, and Pagels 6–7].) The second sacrificial structure, in which the violator and the scapegoat are the same, may be found in some of the most promi¨ drama, nent instances of this plot – for instance, the important Yuan The Injustice Done to Tou Ngo, and the paradigm European tragedy, Oedipus the King. In the former, Tou Ngo is executed for a crime she did not commit. The violators are, then, the people who participated in the unjust execution. Before she dies, she prophesies that, due to her murder, “the county of Ch’u shall suffer a drought” (Kuan Han-ch’ing 143). The prophecy comes true. Three years later, as the drought continues, a new magistrate arrives. After investigation, he condemns all those who acted falsely in the prosecution of Tou Ngo (157–8). Thus the scapegoats are the same as the initial violators. Their punishment returns plenty to the land. This play has remarkable complexity in the areas we are considering. Tou Ngo is socially normative in the content of her action, but individualistic in its form. More exactly, Tou Ngo defends traditional Chinese beliefs and practices. But, in doing so, she acts individualistically against her mother and the court. In contrast, her antagonists are socially normative in their formal position – as parent, magistrate, and so on. But their motives are individualistic and their specific actions are contrary to the very social principles they are supposed to represent and enforce. This is like the story of Jesus, then, in having it both ways. The apparent conflict may be resolved by distinguishing between immanent social norms, which may be false, and transcendent principles which are necessarily true and which may be embodied in socially individualistic characters. Thus, the tendency toward individualistic characterization and sympathy is connected with a tendency toward separating ethical principle from common social practice. The cases we have been considering in this section suggest the following broad narrative principles: A standard generative element

212

The Mind and Its Stories

in prototype narratives is a conflict between individualistic and socially normative characters. Ordinary narrative development tends to foster sympathy for the former over the latter. Ethically this is often allowed by an overt or tacit distinction between transcendent and immanent norms, which is to say, ethical principles (or putatively authentic cultural principles) and current cultural practices. The transcendent norms regularly include an implicit maximization of individual freedom, at least for subjectively defined or personalized characters. Empathic and Nonempathic Agents There are, of course, many other ways in which one might organize heroes. Any system of organization will be more or less valuable depending on one’s purposes and the body of literature one is considering. The division between the individualistic and the socially normative, however, seems broadly consequential – and, at the same time, relatively underrecognized. A second division that is consequential and largely unnoticed is, in some ways, a version of Aristotle’s distinction among characters who are like us, characters who are superior to us, and characters who are inferior to us. Aristotle has touched on a significant principle of differentiation here, but has not, I believe, drawn it precisely. The important division has to do not with the intrinsic value of the characters, but with empathy. It concerns characters with whom we can in principle empathize situationally and characters who are, for some reason, beyond situational empathic identification. This is related to Aristotle’s division because our ability to identify with characters has at least something to do with whether or not the characters are “like us.” Indeed, perhaps the most obvious cases of characters beyond empathic identification are absolutes – absolutes of good and evil, power, knowledge, and so on. The initial impulse of much heroic literature is to put the enemy beyond identification in a demonic direction. There is also an impulse (encouraged no doubt by royal patronage) to put the king beyond situational identification in a divine direction. In keeping with this, heroic tragi-comedies regularly link the ruler with divinity and the enemy with Satanic or related powers of evil. This division not only inhibits situational identification, but encourages categorial identification (via the unifying, divinely chosen ruler)

Some General Principles of Plot

213

along with a sense of group threat (from the demonized enemy). In this way, it fosters an invigorated ethics of defense. However, as we saw in Chapter 4, one striking thing about paradigm literature cross-culturally is that it does not maintain this division at all consistently. In fact, literary works routinely undermine this division. In this regard, narratives regularly begin with an opposition between characters with whom we can identify and characters with whom we cannot identify. But the development of narratives itself conduces toward the extension of empathic identification to increasing numbers and increasing varieties of characters. Put differently, narratives might begin by setting up certain characters as objects of awe or disdain. The development of narratives, however, pushes away from both awe and disdain toward a sort of subjective normalcy. Another way of thinking about the issue is in terms of suffering. Situational empathy is most significantly triggered by suffering. There is, of course, a difference between deserved and undeserved suffering. However, the fact that the victim of suffering deserves punishment does not mean that situational empathy is impossible. Rather, there are several variables governing when we feel situational empathy for a character who merits punishment. First, there is the intensity and extent of the punishment. It should be proportionate to the crime, not excessive. Second, there is the degree of threat; insofar as the character is weak and unthreatening, we are more likely to feel situational empathy, no matter how deserving the punishment. Third, there is the presence or absence of remorse. A character’s feeling of remorse seems to be a particularly powerful factor in inspiring our situational empathy. Fourth, there is the salience of the pain. We are more likely to empathize with suffering that we experience as salient – for example, suffering that is represented in vivid and specific detail that encourages visualization and recollection. Finally, there is the extent of individualism. Again, there appears to be a pull toward individualistic characters, whatever the initial intent of the work. Most obviously, then, we are unlikely to have situational empathy with those who do not suffer or who deserve unqualified suffering – thus divine and demonic absolutes. But, again, the idea that there are such characters is continually undermined by practical representations. As Romantic Satanism attests, this undermining occurs even in

214

The Mind and Its Stories

myth, where literal gods and demons are in question. The point holds still more obviously for cases where the deification and demonization are metaphorical. The enemy soldier in pain looks just like the comrade in pain. It is difficult to sustain a distinction between them. The image tends to trigger personal memories and corresponding identifications even when the enemy has been demonized. In sum, a second important axis of character differentiation is that between characters with whom we can identify situationally and those with whom we cannot. This division contributes to many common propagandistic functions of literary works (for example, nationalistic functions). The nonempathic characters are often rendered nonempathic by some sort of absolutization of their character (for example, absolutization of their moral excellence or depravity). However, as a result of individual particularity in characterization, the course of narrative tends to undermine idealization and dehumanization, if these are established initially. Finally, we identify most readily with characters who are suffering, especially when that suffering is salient, excessive, and remorseful and the characters are individualistic and unthreatening. Ancillary Characters In the preceding pages, I have made reference to characters other than the hero, but unsystematically. In fact, the agent/goal structure implies that there will be two types of ancillary character, types widely identified by writers on narrative structure (see, for example, Greimas 205–7): helpers and opponents, characters who facilitate the hero’s achievement of a goal and characters who inhibit that achievement. There are many ways of organizing ancillary characters in a plot. Again, these will be more or less valuable, depending upon one’s purposes and depending upon the body of works under consideration. However, helper and opponent are the only characters that result directly from the agent/goal structure of prototype narrative. Thus, they are the two types that have most obvious significance in the present context. These two types are relatively clear in themselves. Helping figures seem particularly straightforward. To a certain extent, they are versions of the hero. In fact, sometimes it is difficult to distinguish one from the other. In some heroic stories, the leader (for example, the

Some General Principles of Plot

215

king) is the hero and any character who acts on his/her behalf (for example, a warrior) is really a helper figure. In other stories, both are heroes. In keeping with this, the variables mentioned previously apply to helpers. Thus, helpers may be elevated above situational empathy, most obviously when they are divine, but they may also elicit pathos. They may be individualistic, but they may also be socially normative. This does not often seem to be significant, for the helper’s character traits tend to repeat those of the hero. It does sometimes happen that the helper is socially normative while the hero is individualistic. But, in these cases, the helper typically stands for a distinct set of social norms from those to which the hero is opposed. This is probably most common in romantic works, but it occurs elsewhere as well. One heroic example may be found in the Malian Epic of Son-Jara, where Son-Jara is in conflict with his royal half-brother, but receives aid from his aunt. While the half-brother represents the male ruling hierarchy, the aunt represents a distinct, female social organization, more ancient, more mystical, and more purely a matter of pre-Islamic Mande belief. (Of course, as is common in heroic plots, Son-Jara himself is socially normative on the male side, though the form of his relation to that hierarchy is individualistic. Here, as elsewhere, we see a distinction between false, immanent practices and true, transcendent principles.) Distinctions of this sort sometimes have consequences for our understanding of a work’s political attitudes and its ethical implications. It is also worth noting that helping characters may be transferred from one prototype structure to another or may take on characteristics from different structures. For example, helpers in heroic plots may be self-sacrificing. Consider an important scene from The Epic of Son-Jara. Without the fruit of the shea tree, Son-Jara cannot perform the magic necessary to succeed in achieving heroic greatness. He can find only “one old dry Shea tree” (Fa-Digi l. 2418), which will not supply the necessary fruit. Here we have a version of the famine scenario, writ small, and integrated into a heroic plot. Son-Jara’s mother calls out, “Before the break of day/That dried up shea tree here,/Let it bear leaf and fruit” (ll. 2429–31) explaining that she will “change [her] dwelling” before dawn (l. 2443). The next morning, she is dead and the tree bears fruit. Thus, the sacrifice has produced food, via the standard sacrificial sequence. However, in this case that sequence

216

The Mind and Its Stories

has been integrated into a heroic narrative. This cross-genre use of helping characters is obviously an instance of the more general crossgenre borrowing of structural elements, motifs, and so on. Blocking figures follow the same general patterns. As we have already discussed, blocking figures are regularly demonized or otherwise set outside the realm of situational identification. However, once again, this deempathization is regularly undermined in the course of paradigm narratives. Thus, enemies often develop from nonempathic to empathic characters. The division between socially normative and individualistic characters seems particularly consequential with respect to blocking characters. Just as there are socially normative and individualistic heroes, there are socially normative and individualistic opponents. In heroic plots, the enemy king is socially normative. But the enemy warrior is likely to be individualistic in just the way that the hero warrior is individualistic. Indeed, the individualistic opponent is often more an admired rival than a simple enemy. Moreover, typically, our situational empathy with the enemy in heroic plots is focused on characters of two sorts: innocents, which is to say, characters who cannot be seen as bearing the guilt of the enemy (most obviously, children), and individualistic characters, characters who are to some degree at odds with the enemy, even though they are ultimately loyal. Hector in the Iliad is a good example of the latter. It is fitting that the epilogue of suffering should concern him, for he was always uneasy with Trojan policy. Though he never betrayed his side, but rather fought with exemplary bravery, he was nonetheless the dissenting voice. As such, the individualistic enemy is in some ways more of a hero than the individualistic hero, for the hero is ill at ease with the “right” side. It is as if Hector simply happened to be born into the wrong camp, and thus had no choice but to fight on its behalf, though his heart and mind would have led to the right decisions. It is not clear that Achill¯es’ propensities are equally admirable. There is a similar division in the Rostám and Sohráb episode of the Shâhnâme. There the tragedy results in part from the fact that Sohráb is very young (thus partially innocent), and very much at odds with the Tur¯anian ruler whose troops he leads into battle. The same distinction between socially normative and individualistic applies perhaps even more obviously to the romantic antagonist.

Some General Principles of Plot

217

The father in New Comedy, the censorious shaikh in the Persian ghazal, the brothel-keeper (along with the parents) in certain kabuki plays (see Halford and Halford 447), are all socially normative. The rival, in contrast, may be either socially normative or individualistic. Here too the socially normative characters are typically placed beyond empathic identification – in part because they do not suffer, or suffer only the disappointment of not having controlled the lovers and prevented their union. However, it is possible for an individualistic rival to inspire empathic identification, especially if he/she is innocent of the lovers’ separation and his/her own feelings are sincere (as with Layl¯a’s husband, in Niz.a¯ m¯ı’s poem). In plots of sacrifice, the punitive deity is typically one opponent. By the nature of the plot, this opponent cannot be defeated – hence the need for appeasem*nt through sacrifice. This deity is almost always socially normative. In this way, the punishing deity is a pure enemy, in the sense of a socially normative opponent, though it is of course theologically incorrect to say this. On the other hand, the punishing god is not the only antagonist of hapless humans, such as Adam and Eve. There is often an individualistic opponent as well. This opponent is typically the figure who leads the violators (for example, Adam and Eve) into transgression. The obvious example is Satan – in Genesis, the serpent. However, the point is not confined to ` . u` is a the Judeo-Christian and Muslim traditions. In Yoruba belief, Es deity whose function is to trick humans into committing violations of divine prescriptions so that they have to make sacrifices to the gods (see Awolalu 29). He, too, is an individualistic opponent or “enemy rival.” Here again we have an indication of how Romantic Satanism arises in its various forms. Romantic Satanism is the result of making a demonized opponent individualistic (and making his/her suffering salient) while making the main deity a socially normative (and nonsuffering) opponent. Whatever his intent, this is what Milton did in Paradise Lost. The Causal Sequence A story involves not only agents but a series of events. The agent/goal structure implies that the hero must at some point be lacking the goal. Thus, in the comic form, the series of events involves, at the

218

The Mind and Its Stories

very least, the general sequence: Hero lacks goal; hero pursues goal; hero achieves goal. In some cases, the sequence is circular, with two elements preceding those just mentioned: “hero has achieved goal; hero loses goal,” or something along those lines. This sequence must involve causal links. Or, rather, the later events and conditions must follow plausibly from the earlier events and conditions. This is most often a matter of causality in the real world sense. However, there are other ways in which a series of events may be linked together. What is important in stories is not precisely objective causal plausibility. Rather, it is the subjective sense that new events are understandable as the result of preceding events. Aristotle emphasized this when he noted that narrative outcomes may appear plausible or convincing due to “design,” whatever their causal status. Aristotle’s example concerns the death of Mitys’s murderer: “the statue of Mitys at Argos . . . fell upon his murderer while he was a spectator at a festival, and killed him” (39). As Aristotle noted, we accept this sort of sequence – “poetic justice,” as it is called in English. Even though it hardly makes sense in terms of objective causality, it strikes us as plausible. The point is not confined to cases of poetic justice per se. In connection with this, we may distinguish three types of causality that operate significantly in narrative. The first is the ordinary, broadly predictable causality of the macroscopic world. Not only the causality of objects operating on objects, but the predictable causality of human experience and action – certain objects cause fear and lead to flight, and so on. Within this normal or “general” causality, we need to distinguish the ordinary causality that applies to things, “efficient” causality, from the ordinary causality that applies to persons, “final” causality. Final causality is a causality of means and ends. A final causal explanation involves naming a goal and a set of beliefs: Why did Smith go to the store? He wanted to buy some milk. Efficient causality is a causality of objectal events and their effects, prototypically events and effects related to one another by physical laws – unsupported objects fall to the ground; fire burns human flesh; lack of water causes death. It is, roughly, “billiard ball” causality. Perhaps the most important point to make about causality in prototypical narrative is that it pushes inevitably toward final causality. In keeping with its emotive orientation, prototypical narrative tends to maximize explanation in terms of intent. Such narrative not only

Some General Principles of Plot

219

presents the main events in terms of an agent’s goal and actions to achieve that goal. It also tends to transform as many causal sequences as possible into final causal sequences. We may refer to this as the personalization of causal factors. Prototypical narratives, then, tend to personify and individualize conflicts with nature or with broad social structures. The tendency of prototypical narratives is to treat as many events as possible in intentional terms, in terms of human thoughts, feelings, propensities, attitudes, affiliations, and aims. The point is particularly obvious in sacrificial narratives where forces of nature are personalized, such that a famine is the withdrawal of rain by a god. The way to end the famine is to appease the angered deity, compensating for the sin that caused this punishment. The second type of causality is a sort of augmentation of ordinary causality to include common beliefs about causal relations in the real world. In a society that believes in witchcraft, witchcraft may figure in the causal sequences of the narrative. Indeed, once a belief of this sort exists in a society, and people become aware of it, it is almost always available for incorporation into fictional narratives, even narratives outside the culture where the belief is accepted. For example, witchcraft appears commonly in narratives written and read by people who do not believe in witchcraft. One might think that the problem with, say, literary witchcraft is explaining why people would accept such a thing in a literary work when they reject it in real life. But, in fact, it is really not all that surprising that people who do not believe in witchcraft would be able to accept it in fiction, or that they would be able to accept other objectively implausible types of causality. After all, we accept new events, persons, conditions that we know do not exist, why not counterfactual causal principles? Rather, the problem with witchcraft as a causal device is just the opposite – its limited use. The introduction of witchcraft opens the door for a wide variety of causal developments and thus, at a certain point, can render any causal sequence implausible. Once witchcraft or related nonnormal causalities are introduced, their nonuse has to be explained. For example, if witchcraft saves the hero in the end, one might reasonably ask why it was not used to defeat the enemy right at the outset. For this reason, nonnormal causalities operate within an often rather strict set of operational constraints. They form a set of “restricted” causalities, as opposed to “general” or normal causality.

220

The Mind and Its Stories

Perhaps the most theoretically interesting sort of causality is, roughly, what Aristotle isolated with the concept of “design.” In terms of objective properties, it is often based on thematic affinities (to bor¨ row a phrase from Adolf Grunbaum) – similarities (or related associations) that allow us to link two events distinctively. To return to Aristotle’s example, there is a thematic affinity between Mitys’s murder and Mitys’s statue falling on the murderer. The distinctive similarity is straightforward. Someone killed Mitys. Now “Mitys” kills that person. The same point would apply if the murderer simply died by some random accident on the anniversary of Mitys’s murder. Here, the similarity would be that Mitys died on this day, then his murderer died on this day. In both cases, the similarity leads us to think of the two events as forming a mutually relevant temporal sequence, which we then understand as a specifically causal sequence. Of course, this is hardly a full explanation. We see many events as similar or thematically related without thereby concluding that they form a causal sequence. In the cases we are considering, we infer a causal sequence in part because the similarity and mutual relevance are integrated into a larger final causal framework. Thematic association becomes causal precisely by being part of a “design,” which is to say through its operation in the intent of a designer. Ultimately, this form of causality is bound up with providence. It takes part in the causality of a divine plan. Often, the intervention of divinity is explicit in narrative. In romantic plots, such as Niz.a¯ m¯ı’s Layl¯a and ¯ or K¯alid¯asa’s Abhijn¯ ˜ ana´sa¯ kuntalam, a deity takes a direct and Majnun active role in uniting the lovers. The importance of divine intervention in classical European drama is suggested by the fact that our term for a plot-resolving coincidence is “deus ex machina.” This phrase refers to “the practice of some Greek playwrights . . . to end a drama with a god who was lowered to the stage by a mechanical apparatus and, by his judgment and commands, solved the problems of the human characters” (Abrams 42). The causality of Aristotelian design is merely the same sort of providential intervention, now rendered implicit – or, rather, the same sort of providential and punitive intervention. Divine shaping of life is always conceived of as a matter of both reward and chastisem*nt. Moreover, this causality of design is of a piece with the ethical causality that underlies sacrificial narrative – the explanation of extreme conditions (such as drought)

Some General Principles of Plot

221

not by ordinary causal principles but by principles of reward and, more commonly, punishment. Of course, it is characteristic of the other two prototype genres as well. For example, in heroic works, God is continually intervening on the side of the in-group to defeat the demonic enemy. This causality of design is, then, an extension of final causality to distinctively associated (usually distinctively similar) sequential events. In that way its importance is unsurprising, for it is another instance of the maximization of final causality. Indeed, the same is true for restricted causalities such as witchcraft. All such nonnormal causal principles operate to personalize sequences of cause and effect. Goals The final component of our preliminary definition is goals. In many ways, the idea of goals in the agent/goal structure is straightforward. A goal is anything an agent might strive to achieve. As we have already discussed, what an agent strives to achieve is happiness. This is trivially true. It is part of what we mean by “happiness.” Even when a masoch*st strives for the experience of pain, he/she is striving for happiness. Thus, technically, an agent’s goals are always imagined to be eliciting conditions for happiness, or means to the eliciting conditions for happiness. However, this already brings us to a complication. The very nature of final causality is, in a sense, paradoxical. The pursuit of a goal sustains the hero, not only in narrative, but in life. Achieving the goal leaves one in a sort of quandary as to what one should do next. Put differently, life is, to a great extent, decision and action aimed toward the achievement of goals. Our daily lives are animated by a sense of future possibilities. We move through the seeming trivialities of ordinary existence in the hope of reaching some more encompassing goal in the longer term. In a sense, the very lack of happiness is what keeps us going. Sustaining emotions sustain us toward the achievement of some goal. Deprivation, combined with a sense of possibility, is what gives us hope for the future. Once the goal is achieved, however, we risk losing a sense of direction, and possibility. Though it seems counterintuitive, success itself can bring with it a form of despair, for it leaves an agent without a goal, and thus with nothing to

222

The Mind and Its Stories

motivate and sustain him/her. Moreover, ordinary human goals are, as they say, not everything they are cracked up to be. We have already seen that individual and group domination are highly dubious aspirations, likely to bring sorrow and remorse even to those who achieve them. Though romantic plots almost always treat romantic union as an unequivocal good, the actual situation of married people – as well as some of the Ainu poems – suggest that this is not true, even in societies where marriages based on love and individual choice are the norm. Of course, one might respond that this all may bear on human life, but it does not really have any consequences for stories. Despair from success does contribute to the epilogue of suffering, in the way already discussed, but its narrative importance is, it might seem, confined to that one area. However, this is not true. The ramifications of despair are much more general, and not only because we explicitly or implicitly understand our actual lives in terms of narrative sequences. In order to clarify this relation between plot structure and the despair of success, we need to draw a few distinctions. First, we need to distinguish between short-term aims and more encompassing projects. Projects are the long-term goals that we see as fostering happiness. In other words, projects are the eliciting conditions for happiness – or, rather, for a particular type of happiness, which itself may be called “continuing” or “condition” happiness. Specifically, we may distinguish between happiness as an experience, which may be momentary, and happiness as a condition, which is, by definition, long lasting and continuous. The latter is what the Greeks called “eudaimonia,” the good life. The two are not unrelated. Sexual union is one of our primary prototypes for experiential happiness, or enjoyment. It is not, and cannot be, a prototype for condition happiness or eudaimonia. However, it is related to, and part of, one prototype for eudaimonia – romantic union. A similar point could be made about various sorts of limited success (for example, in games, exams, battles) and long-term social standing. Of course, we may pursue enjoyment, just as we may pursue eudaimonia. Thus, at this point, we may distinguish three types of goal: short-term aims that serve as means, intrinsically pleasurable pursuits (such as sexual union) that yield enjoyment, and projects that, when achieved, define condition happiness or eudaimonia.

Some General Principles of Plot

223

Short-term aims may conduce toward pleasurable pursuits or toward long-term projects. Note that projects are limited. Indeed, they are largely limited to the prototype eliciting conditions for the outcome emotion of happiness as discussed above. In other words, prototypical happiness, as treated in chapter three, is eudaimonia, not simple enjoyment. The happiness of enjoyment may also be an outcome emotion for a given narrative (fictional or lived), but it cannot be prototypical for happiness due to its ephemeral quality. Moreover, insofar as it contributes to eudaimonia – that is, insofar as it is subsumed under a project – it is junctural in the way that means are junctural. In these terms, the problem of despair may be seen as taking two forms. The obvious form is a loss of hope regarding the achievement of a particular project (for example, the loss of hope that one will ever marry one’s particular beloved). The second form is a loss of projects – the sense that there is no project worth pursuing and thus nothing that would define one’s short term aims and organize one’s temporary pleasures. Again, this can occur through the achievement of a project. Specifically, once one accomplishes a project, one is likely to be faced with several problems. First, there is the problem that one has been deprived of the goal that animated one’s activities, gave meaning to one’s aims. The sense and purpose of short-term aims seems lost. Second, the joy we experience in a long term condition will fade. This results from a general psychological principle, the “law of habituation,” as Frijda calls it (“Laws” 277). We become accustomed to positive emotional experiences that are continuous or predictably recurring. In other words, the intensity of happiness wanes as the state of romantic union or social domination becomes habitual. Finally, it is simply a fact about human life that everyone’s future holds physical decline – thus decrease in the success and joy of project-related enjoyments (such as sex) – and, ultimately, death. Again, the problem with accomplishing a project is that one must then ask, “What now?” In addition, the answer is not terribly cheering – “Well, there’s sickness, decrepitude, and death.” The problem recurs with any achievable project one might undertake. This is the familiar problem of ambition and greed. They are, in principle, boundless in that each success proves unsatisfactory. As a matter of logic, there is only one type of goal that would solve this problem. It would have to encompass and render complete all normal

224

The Mind and Its Stories

actions. It would have to absolutize the enjoyments of our pursuits and the eudaimonia of our condition while, at the same time, rendering those enjoyments and that eudaimonia impervious to the effects of habituation, as if they were always new. Finally, it would have to prevent both decline and death. In other words, it would have to be a form of outcome happiness that actually does conform to the enduring and all-encompassing qualities of its prototype. Unsurprisingly, most societies have imagined just such a goal into existence. In Christianity, it is Heaven. In Islam, it is the Garden. In Ved¯antic Hinduism, it is moks.a, release from rebirth and reunification with divinity. In each case, it is a spiritual state of absolute bliss that is undecaying and eternal. One function of such an imagined goal – not simply a project, but a culminating telos of human life – is to provide something beyond projects, a principle that motivates and gives sense and direction to all ordinary life. Of course, as an imagined goal, it is vague, difficult to specify. In consequence, it is routinely understood in terms of more concrete, prototypical projects – primarily romantic union, as we have already stressed, but also including plenty and even in-group domination (insofar as one’s in-group is, precisely, the saved). This notion of a transcendental goal or telos, modeled on prototypical projects, but at the same time extending and intensifying them, allows us to explain an otherwise peculiar feature of prototype narratives. We have noted repeatedly that both romantic and heroic tragi-comedies tend to culminate in a transcendental or spiritualized ¯ are joined version of their respective goals. Thus, Layl¯a and Majnun ´ in the divine Garden; Sakuntal¯a and Dus.yanta meet again in a heavenly hermitage; R¯ama leaves his kingdom to return to his place with the gods; the heroes of Gilgamesh and The Mwindo Epic return to their kingdoms with spiritual wisdom and a sense of religious resignation. This transcendentalizing of the conclusion seems to derive from the tacit incorporation of the transcendental goal or telos into the realized project. Specifically, in our understanding of real causal sequences, we tend, explicitly or implicitly, to imagine a telos beyond even our projects. This telos serves to structure and give sense to our projects in the way that our projects serve to structure and give sense to our short-term aims. In that way, it is always there, beyond the projects. (This seems true, at least in some degree, even for those of us who

Some General Principles of Plot

225

are atheists. In general, our minds operate with a wide range of beliefs, expectations, etc., that we would self-consciously repudiate.) In consequence, the telos – itself imagined on a romantic, sacrificial, and/or heroic model – tends to interfere or merge with the project in a narrative, for the telos is what makes the project worth achieving. Without the telos, the project risks becoming hollow. In short, it risks despair. One result is the transcendental or spiritualized conclusion – which may, in fact, recognize and respond to that despair directly, as in the epilogue of suffering, some of the Ainu romantic poems, and such sacrificial narratives as the story of Jesus. The transcendentalizing of the project in narrative is related to two other tendencies as well – the social generalization of the project and the idealization of the primary object. Social generalization is found prominently in heroic narratives. As we have seen, these stories standardly aim toward group domination, especially national domination. There are, of course, dominant individuals in these stories. But their accession to power regularly occurs in the context and service of group domination. The triumph of the hero is a triumph of the entire nation. The condition of the individual is generalized to the group. The same sort of social generalization is obvious in sacrificial stories as well, for the way to overcome individual hunger is through overcoming collective famine. The sacrifice that restores food to the individual does so insofar as it restores food to the group as a whole. Romantic plots may seem to be an exception here. However, there is an impulse toward social generalization in romantic plots also, most obviously in marriage. Thus we sometimes find romantic plots including not only the marriage of the hero and heroine, but the romantic union of various ancillary characters as well. These multiple marriages extend personal happiness beyond the main couple to a broader society. Cases from Shakespeare come to mind immediately, but the same generalization occurs elsewhere. In the Sanskrit comedy, The Little Clay Cart, C¯arudatta and Vasantasen¯a, the main lovers, ´ are able to wed, but so are Sarvilaka and Madanik¯a (Vasantasen¯a’s servant), a couple nearly cheated of their chance for happiness due to poverty and Madanik¯a’s servitude. When R¯ama wins the hand of S¯ıt¯a, they are not married alone; R¯ama’s brothers receive spouses as well and all are wed in a collective ceremony. The story of Bamsi Beyrek, from the Turkish Book of Dede Korkut, involves a still more

226

The Mind and Its Stories

obvious social generalization. When Beyrek marries, he and Prince Kazan arrange brides for his thirty-nine warriors also, and the forty couples share an extravagant, forty day wedding feast. Finally, the prototype conditions for happiness standardly include not only a situation, but an object or a person as well (a number of writers have emphasized the importance of the “sought for” object or person; see, for example, Greimas 202–3 and citations). In itself, the point is obvious and seemingly inconsequential. The sacrificial plot involves food; the heroic plot concerns the kingdom or nation; the romantic plot aims toward the beloved – the prototypical physical, social, and personal object, respectively. Everyone is aware that the beloved is idealized in romantic plots, in the lyric poems that derive from such plots, and so on. In fact, all three objects are regularly idealized. Of course, this idealization is likely to be limited by other tendencies, such as that toward empathy-producing humanization in the case of the beloved. Nonetheless, it is clear that idealization of the object is important to prototype narratives of all three sorts. Thus we find that heroic tragi-comedies regularly establish the kingdom as a utopia. The R¯am¯ayan.a informs us that “When R¯ama ruled, nowhere was the widow’s wail heard . . . and no one was overtaken by calamities. It did not fall to the lot of old men to do the obsequies of the young. Gladness was everywhere and every one was keen on Dharma [ethical duty]” (V¯alm¯ıki, vol. III, 371). When Son-Jara reigned, “peace and happiness entered” society; “justice prevailed everywhere” and “the villages knew prosperity again” (Kouyat´e 81). When the society is not established as a utopia, it is nonetheless often elevated as a place of particular prosperity, virtue, and valor. Again, it is regularly a society special to God. As to sacrificial narratives, the result of sacrifice is often abundance, sometimes marked by a sort of perfect food. Consider again the story of Jesus. The salvation brought by his crucifixion is eternal bliss – and it is commemorated by an unparalleled spiritual food, the Eucharist. (I take it that examples of idealizing the beloved in romantic tragi-comedy are too extensive and evident to require citation.) There are several sources for this idealization. Most obviously, we tend to idealize what we want and do not have, whether it is a person, a political position, or what someone at the next table is eating in a restaurant. Unsurprisingly, this broad human tendency turns up in

Some General Principles of Plot

227

stories. Of course, in stories, the object really is idealized; it is not, most often, presented as idealized only in the imagination of the agent. This brings us to a second reason. Objective idealization fosters an appropriate response on the part of the reader, for it tends to repeat, and thus prime, the reader’s own idealizations – most obviously, in the case of the romantic plot, his/her idealizations of some beloved person at the time he/she was most in love. Finally, as the preceding examples suggest, idealization is bound up with spiritualization. The sacrificial story not only results in food, but in heaven. The kingdom not only wins, it is selected by God. The hero who rules does so with the dispensation of God and, after the epilogue of suffering, with divine wisdom. The beloved is not only beautiful and charming, but morally and spiritually elevated – so much so that the final union of lover and beloved may take place only in heaven. Thus, idealization of the object is, among other things, a consequence of the spiritualization of projects via the telic trajectory in which they are implicitly embedded.

a note on emotional intensiﬁcation Both idealization and spiritualization are in part a matter of strengthening characteristics that foster an emotional response. In other words, both are in part a matter of intensification, which is to say, the enhancement of some aspect of a work that produces emotional impact. Intensification bears not only on agents (through idealization) and goals (through telic spiritualization), but also on causal sequences. We have already noted that the initial conflict in prototypical narratives is often familial – parents forbid the lovers’ union; a brother usurps the hero’s throne. This familializing of the conflict is clearly a way of intensifying that conflict, both ethically and emotionally. In part, this is the result of “role expectations.” As Ortony, Clore, and Collins point out, it is more shocking to discover that a nun has murdered someone than to discover that a gangster has done so (see 79–81). The same point holds for near relatives. More broadly, the minimum requirement for the agent/goal structure is that the agent lacks the goal, strives for it, and (in the full comic version) achieves the goal. In prototypical narratives, however, lack of the goal is intensified into the opposite of the goal. The prototype

228

The Mind and Its Stories

conditions for happiness are preceded by the prototype conditions for sorrow. The opposite of spousal union is the death of the beloved. The opposite of social domination and esteem is ostracism and disgrace; the opposite of group domination is defeat and conquest. The opposite of being well fed is not simply waiting for dinner; it is famine. Note that this also intensifies the goal itself, in part by increasing the effort necessary to achieve the goal. As Ortony, Clore, and Collins point out “increases in effort tend to increase the degree to which goals are positively valued” (73). In addition, while each prototype narrative draws primarily on emotion prototypes of its own category (physical, personal, or social), the different prototypes for both sorrow and happiness may be conjoined as well. This occurs most obviously when elements from different prototypical and narrative structures are combined into a single story – when the R¯am¯ayan.a intertwines a romantic plot with a heroic plot or when Oedipus the King mixes the epilogue of suffering with a sacrificial narrative. It can also happen, however, that one type of story draws on another for structural relations, imagery, or plot elements, without the main concerns of the other narrative being literally incorporated as a plot line. Lovers in Sanskrit romantic tragi-comedies waste away like famine victims. Exiled princes long for their people like lovers separated from their beloveds, and so on. Conversely, Son-Jara’s reign, once he is rightfully placed on the throne, is marked by plentiful harvests: “Vast fields of millet, rice, cotton, indigo and fonio surrounded the villages” (Kouyat´e 81). Jesus’ resurrection is frequently represented as a military triumph over Satan. Combining prototypes for happiness and sorrow from different prototype stories is another way of intensifying the event sequence of the plot. Indeed, a number of the patterns discussed previously (for example, the personalization of causality) could be understood as forms of intensification as well – indicating once again the close interrelation between prototypical narrative and emotion.

place, time, and social order Despite the apparent implications of the discussion thus far, character, goal, and causal sequence are not the only constituents of prototypical plots, nor are they the only elements that are subjected to such

Some General Principles of Plot

229

principles as intensification. Three other constituents appear particularly significant. Labov notes that, in most stories, “At the outset, it is necessary to identify in some way the time, place, persons, and their activity or the situation.” “Persons” and “activity” or “situation” refer to the elements discussed previously. Labov adds time and place to the list. These may appear to be mere incidentals, necessary though not intrinsically consequential aspects of narrative. However, this is not true. An understanding of time and place is, in fact, crucial to our understanding of the general principles of narrative. First of all, both time and place in narrative should not to be understood objectively, but subjectively. Like so much else in prototypical narrative, they are personalized, moved away from efficient causality toward final causality. Time is not, most importantly, the time of chronometers. It is lived time. On the one hand, it is the psychological time of individual expectation, remembrance, enjoyment, fear, and so on. On the other hand, it is the naturally organized time that structures human action. The latter may be divided into cyclical and noncyclical structures. Cyclical structures would include, most obviously, days and seasons, both of which figure prominently in the literal development of plots and, perhaps even more, in their associated imagery. Noncyclical time would most obviously include the time of aging. Space, too, is lived space. Not mere objective location, but that location as it relates to one’s experience of life – aspiration, feeling, action, memory, and so on. The most crucial spatial division is between what is “home” and what is not – with “home” understood contextually to refer to a particular house, town, region, country, or whatever. The spatial structure of narratives is, then, most significantly a structure of home and away. “Home” in romantic plots is a permanent place with the beloved. “Home” in heroic plots is the nation. “Home” in sacrificial plots is a paradise of natural comfort and plenty. It is this personalized spatial structure that makes exile (that is, expulsion and exclusion from home) a central element in all three sorts of narrative. Put differently, in stories, time and space are, first of all, emotional time and space. The intensifications of space and time in prototypical narratives are probably clear from what has already been said about the telic conclusions of these narratives. Time is intensified into eternity in

230

The Mind and Its Stories

the spiritualized endings of all three prototype stories (e.g., when the lovers live “happily ever after”) and home is intensified into heaven or some comparable spiritual ideal. Conversely, those who are punished may be sent to an intensified exile – for example, in hell. In addition to these two elements, named by Labov, it is important to add social structure. Social structure provides the social context for the action of the narrative in a way that directly parallels the spatial and temporal structure. However, the social structure is even more crucial, for the conflicts and developments in prototypical narratives continually involve social hierarchies. The lovers cannot marry because of class differences or some internecine conflict in the society; the rightful heir is dethroned and exiled by a usurper; a sin by some members of the community condemns the entire society to hunger. In all three prototypical genres, social structure is of vital consequence. Here, the most important aspect of the social structure seems to be its hierarchical organization. Again, in romantic plots, the lovers run directly against this hierarchy, for that is what causes their separation. Sacrificial plots are most often straightforward treatments of just how a violation of hierarchy leads to massive suffering. In heroic works, the dispossession of the hero is typically a case of some villainous disruption of the hierarchy. The intensification of social structure is found perhaps most obviously in the common tendency to make the heroes of works into nobles or other figures high in the social hierarchy. In some cases, such as narratives of social rebellion, the intensification may push in the opposite direction giving us heroes of little social prestige or position. In either case, the characterization operates to enhance the emotive impact of the work – depending of course on the individual reader (with his/her particular categorial identities, and personal experiences) – because social hierarchy is crucial to our sense of categorial identity and to our personal experience of social happiness or suffering. (The point obviously bears on the ideological functions of literary works as well.)

the common form of prototypical narratives The tendencies outlined thus far set out some of the broad principles that apply across prototypical narratives. However, there are much

Some General Principles of Plot

231

more specific properties shared by these narratives as well. Indeed, the three story prototypes have a common structure that is remarkably detailed. Specifically, each of the genres begins with some sort of social conflict. This is standardly presented as a conflict between the hero and a socially normative opponent, abetted by a rival, who may be individualistic. (For a summary of this analysis, see Table 7.1.) In romantic stories, this is commonly a conflict between the lovers and their parents, who prefer a rival suitor. In the case of heroic plots, we find a usurper, who is often a rival character and thus frequently individualistic, but may be socially normative as well (as when a king tries to exclude his son from kingship). In sacrificial plots, we often find a character “like us” – not precisely a hero – violating some divine (thus socially normative) prescription, perhaps at the urging of an individualistic opponent (such as Satan in the Judeo-Christian story). This is followed by deprivation of the object (the beloved, the kingdom, food) and exile, which is to say, social expulsion of the hero or his/her imprisonment (thus, in either case, his/her exclusion from home). In romantic stories, the lover is often sent away; the beloved may be confined by the father or abducted by the rival. In heroic plots, the hero is often exiled to another nation. The object – which is to say, the kingdom – is taken away through usurpation. In sacrificial plots, food is withheld in obvious ways – and the transgressors may be exiled from a place of plenty, as occurs with Adam and Eve. This deprivation and exile are associated with death – either literal death, a deathlike state, or simply death imagery. This point is obvious enough in all three structures that it does not require elaboration. The next common element (though it need not occur at just this point) is a social extension, in this case a social extension of threat and suffering. The point is obvious in the heroic structure with its threat/defense sequence. In sacrificial narratives, the point is obvious as well, for the initial fault, though individually committed, affects everyone. Romantic stories appear to be an exception, for by their nature they focus on individualistic characters. However, to a remarkable extent, romantic stories do involve this sort of social generalization.

232

Social conflict, especially with a socially representative character, abetted by some individualistic antagonist

Social conflict of lovers with socially representative characters (e.g., parents), often abetted by a rival lover

Social conflict of hero with (socially representative) ruler and/or rival/usurper

General structure

Romantic tragi-comedy

Heroic tragi-comedy

Usurpation (= deprivation of the kingdom or other rightful position), exile or imprisonment of hero

Separation of lovers: exile of lover; confinement or abduction of beloved

Deprivation of object; exile or imprisonment of hero/heroine

Death of hero and/or death imagery

Death of lovers and/or death imagery

Death or death imagery/ deathlike state

Danger to society from invasion (or civil strife)

Generalization of conflict from lovers’ social difference (e.g., class) or generalization of lovers’ pain and mourning

Extension of threat and/or pain from individual to society and/or nature

Division and reconciliation of heroes (e.g., king and warrior); initial hero or heir defeats and punishes the socially representative enemy and/or rival/usurper

Division and reconciliation of lovers; defeat of enemy (e.g., parents), exclusion or punishment of rival

Dissension and reconciliation of heroes or hero and helper; defeat and punishment of the antagonist

table 7.1. The Common Form of Prototypical Narrative Genres

Return of hero, restoration of kingship or rightful place, familial reunion, imagery of rebirth (perhaps birth of an heir), perhaps coronation or related ritual

Return of hero from exile, reunion of lovers, parent/child reunion, imagery of rebirth, marriage

Reacquisition of the object, return from exile, familial reunion, imagery of rebirth, sanctifying ritual

Epilogue of suffering, with final spiritualization (see below)

Spiritualization of lovers (union after death or in a heavenly place)

Spiritualization

233

Violation of divine prescription (with the deity as a socially representative character), often at the urging of some individualistic tempter

Hero, as socially representative, is responsible for the death of innocents

Sacrificial tragi-comedy

Epilogue of suffering

Death of people and/or death imagery

Hero is deprived of Death imagery his/her position and exiled, either due to his/her own remorse or due to actions of some deity

Withholding of food, exile from land of plenty

Society loses ruler or other dominant figure

Danger to society from starvation

The hero, tacitly placed in the position of the enemy or rival, is punished or self-punished; separation from helper

Conflict between society and scapegoat; sacrifice punishes society (pure version) or violator (secular version) and defeats the rival/tempter, who is excluded or punished The kingdom is returned, imagery of rebirth, perhaps a coronation or other ritual

Return of food, return to land of plenty, perhaps resurrection and reunion of families, institution of ritual

Achievement of divine home through apotheosis or suicide or somber return to human home with spiritual wisdom

Reconciliation of society with God/gods, possibility of heavenly life

234

The Mind and Its Stories

One obvious way in which this negative or sorrowful social extension can occur in romantic plots is through an elaboration of the social prejudices that are preventing the marriage. Modern works that treat interracial love are an obvious case of this, for they almost invariably take up the problems faced by the lovers as symptomatic of larger social problems of racism. The point is not confined to the modern period. For example, Terence, himself a former slave, addresses the problem of slavery in several of his plays, sometimes explicitly. In a New Comedy, we are often faced with two lovers who are unable to wed because it seems that the woman is a slave. However, it turns out that the “slave” was actually free born, and thus the ¯ marriage is permitted. In Eunuchus, Terence takes up this standard sequence. However, in this case, the two lovers are actually a young man and the woman he raped. The man excuses the rape by explaining that he thought she was a slave. Moreover, he vows that “I did it for love of her” (207) – though, after the act, the girl “does nothing but cry and can’t bring herself to answer any questions” (194). Of course, this is a “problem comedy.” We can hardly be happy that these “lovers” are united. Nonetheless, the social generalization is clear. In other cases, broader social problems, conflicts, or conditions may grow out of the initial separation of lovers, even if this is not based on any systematic social division, such as race or slavery. In ˜ ana´sa¯ kuntalam, the king forbids the communal festivities of Abhijn¯ spring, and, as a result, even the processes of natural renewal are stopped (K¯alid¯asa, Theater 148–9). Moreover, he partially neglects his duties as king, assigning his minister to make judgments in his ¯ Majnun ¯ convinces one Bedouin place (150–1). In Layl¯a and Majnun, leader to wage war against Layl¯a’s family, thus destroying the peace of Bedouin society. In each case, then, we see the pain of the lovers extended to the larger society. This leads us to the defeat and punishment of the antagonist, often aided by some helper. The lovers are reunited, thus overcoming the “enemy.” The rival may be killed or banished in the process. At the very least he/she is typically removed from the comic community. In the heroic plot, the usurper is defeated, often killed in battle. The point is equally clear in sacrificial tales. In secularized versions, the violator is punished directly. In the pure or religious version, the

Some General Principles of Plot

235

(guilty) community is punished through the representative scapegoat. In connection with this, the tempter is typically defeated as well, often through exclusion from the final comic society (like a romantic rival). Satan, again, is an obvious example. In some instances, the defeat of the rival is preceded by a split between, then reconciliation of, the heroes or the hero and a prime helper. Though I did not discuss the point earlier, it happens with some frequency that the lovers are estranged at some point in the story. Cases range from Gurg¯an¯ı’s V¯ıs and R¯am¯ın to the Japanese ˜ ana´sa¯ kuntalam. We someLove Letter from the Licensed Quarter to Abhijn¯ times find a comparable conflict in heroic plots, as when Rost´am and the Shah divide briefly before the battle with Sohr´ab. Finally, sacrificial narratives may involve some dispute between the sacrificial victim and the social representatives who are there to enact the sacrifice. For example, in the Kondh sacrifice, one segment of the ritual, involves “a very long and somewhat sentimental, but at the same time argumentative, conversation between the priest and the Meriah [sacrificial victim], the object of the one being to show that the victim must calmly submit to suffering in a cause so greatly to the benefit of mankind . . . and of the other, to prove that he had been cruelly deceived” (E. T. Dalton, quoted in Mahapatra xix–xx). Jesus’ “agony in the garden,” where he questions his role and asks not to have to make the sacrifice, provides another instance. The defeat of the rival and enemy leads to the reacquisition of the object and the return home. The lovers are reunited and retrieved from exile. The rightful ruler comes home to the throne. Food is restored to the people, and they are allowed to enter once again the land of plenty, at least potentially. It is also common for other sorts of family reunion to occur in these three structures. In romantic plots, the parents and children, formerly divided over their romance, may be reconciled once the lovers are joined. Moreover, the lovers may have a child from whom they had been separated. The lovers/parents may ˜ ana´sa¯ kuntalam, or the be reunited with this child as well – as in Abhijn¯ ¯ seminal Japanese story of Izaemon and Yugiri. In the latter, the impoverished lovers can at first neither wed nor raise their own child, but in the end they are united and their child is returned to them (for a summary of this story, and a note on its historical significance, see Halford and Halford 210–11).

236

The Mind and Its Stories

Heroic plots may involve the reuniting of families separated during war, especially the royal family, with the rightful heir restored. An obvious case of this is the reunion of the various family members in Ayodhy¯a, when R¯ama returns to rule the kingdom after he has defeated R¯avan.a and rescued S¯ıt¯a. In the Mwindo Epic, Mwindo’s triumph is accompanied by a reunion with his aunt, who had worried that he would die during his final journey, and with his father – though there is the peculiar twist here that Mwindo’s father had fled after trying to kill Mwindo, then had to be tracked down by him, hardly our prototype case of family reunion. The ending of the Odyssey fits here as well. Though not, as a whole, a strict heroic plot, the part of the story that takes place in Ithaca is clearly a version of the heroic plot, with the exiled ruler’s kingdom threatened by usurpers. (There is a version of the romantic plot operating in this part as well.) Chikamatsu’s Battles of Coxinga brings a father and son ¨ play, together at the final point of heroic triumph. The Chinese Yuan A Stratagem of Interlocking Rings, includes a reunion as well, though an unusual one – a reunion of separated lovers, clearly borrowed from the romantic plot. The enormously popular movie, Independence Day, a straightforward heroic tragi-comedy, concludes with the reunion of parents with children and also the reunion of lovers. In sacrificial narratives, it sometimes happens that the dead are resurrected or otherwise spiritually freed (e.g., for Heaven), in a scenario that allows either a wordly or spiritual reunion for those who have been separated. We see this in the general salvation produced by Jesus’ sacrifice in the Christian story. There are also more mundane reunions, such as the reunion of mother and children after the sacrificial death of Nathan in Markandaya’s Nectar in a Sieve. As the sacrificial case suggests, this is all typically accompanied by imagery of rebirth, paralleling the imagery of death in the first half of the structure. In all three prototype structures, this may be marked by a sanctifying ceremony. The most obvious case of this is the sacrificial narrative, for it frequently includes an account of the origin of a standard ritual (for example, the Eucharist in the story of Jesus). But the romantic story too culminates, explicitly or implicitly, in a ceremony – marriage. And the heroic narrative often leads up to a coronation or some similar ritual, as, for example, in The Mwindo Epic.

Some General Principles of Plot

237

Finally, prototypical narratives end in some sort of social generalization and spiritualization, as discussed previously. The only oddity here is that the heroic narrative often goes through an entire further plot sequence before this occurs – the epilogue of suffering. But the epilogue of suffering is not as anomalous in this respect as it may seem, for it follows the common structure as well, more or less running through the sequence a second time. The epilogue begins with a social conflict in which the hero is the socially normative character responsible for the death of some innocent. This leads to deprivation of the object and exile – voluntary or forced abandonment of home – as the hero gives up the kingship or other rewards of victory and goes in search of knowledge (as in the case of Gilgamesh) or atonement (as in the case of Kumagae) or is deprived of the kingship and exiled in punishment (as in the case of Mwindo). This involves considerable imagery of death, and is directly consequential for the society (thus involving social extension). While there is not really an issue of defeat here, it is clear that the entire sequence of events serves to punish the hero as if he/she were the antagonist. This sequence also frequently involves a graphic separation of the hero from a prime helper (as with Gilgamesh and Enkidu or Yudhis.t.hira and his brothers in the Mah¯abh¯arata). All this is often followed by the return of the object (for example, the return of the kingship to Gilgamesh or to Mwindo), imagery of rebirth (as in Zeami’s Atsumori, where Kumagae and Atsumori “shall be re-born together/On one lotus-seat” [73]), and/or ritual (as in the case of Mwindo [see Rureke 143], where the ritual is strikingly close to the rituals of sacrificial narratives). As we have repeatedly noted, this final return, even if it is a return in defeat or loss (as with Gilgamesh), implicitly includes a sense of spiritual elevation, if in a somewhat somber way. In addition, some plots of this sort have a final moment when the hero – perhaps without returning to the kingship – achieves a divine home through apotheosis (as in the case of Oedipus) or suicide (R¯ama). This outline of common structures makes one more thing salient about the organization of prototypical narratives – they are symmetrical. The initial conflict is paired with the final defeat of the enemy and punishment of the rival; the deprivation of the object and the exile are paired with the restoration of the object and the return of

238

The Mind and Its Stories

the hero; the social generalization of the danger is paired with the concluding social generalization of happiness.

conclusion In sum, though the generation and reception of paradigm narratives is a matter of prototypes, the various prototype structures have many things in common. They share a set of strong tendencies – toward personalization and final causality, individualistic characterization, social generalization of both the sorrow and the happiness, concluding spiritualization, idealization of the object, intensification, and so on. They also share a remarkably detailed and complex structure, going well beyond the minimal pattern implied by agent, goal, and causal sequence of events. These common features are bound up with the emotional premises, components, and consequences of narratives. But only part of this is explained by emotion prototypes. There is clearly a great deal of material here, not only for further descriptive specification, but even more so for further psychological explanation – in short, for a continuing research program in narrative, emotion, and cognition.

Afterword From the Emotional Nature of Narrative to the Narrative Nature of Emotion

Up to now, we have treated emotion fairly narrowly, focusing in particular on emotion prototypes as generative of stories, though also treating emotive response to literature. In these final pages, I should like to consider emotion itself, outside literature. Specifically, I should like to reverse the direction of study and examine the ways in which emotion is a function of narrative – not emotion in literary response, but emotion as we experience and understand it in ordinary life. First, I shall discuss some of the ways in which the eliciting conditions, actional outcomes, and even phenomenological tone of emotions are shaped and oriented by stories and lexical prototypes. In connection with this, I shall touch on some political consequences and ideological functions of prototypical narratives in their various social specifications. Second, I shall consider the biological givens of emotion, just how these are reshaped socially, and how both the biological and social components relate to narrative, including the prototypical narratives we have been considering.

convention, feeling, and social function Our ordinary view of emotions is that they preexist prototypes and stories. Our lexical items name them. Our stories recount and, when successful, trigger them. In part, this is no doubt true. It is simply a pattern of nature that people run from packs of wolves, scream, contort their facial features in certain ways, and so on. In other words, in 239

240

Afterword

part our prototype for “fear” simply captures a natural phenomenon. But there is more to it than that. First, as discussed briefly in Chapter 3, we do not simply have some direct intuitive way of identifying emotions. The point is obvious with other people. We do not have access to anyone’s experiential subjectivity. We cannot feel anyone else’s emotions. Thus, to identify their emotions, we must rely on clues from what we can experience. Prototypes are extremely helpful in this regard. How can we tell if someone is in love? Well, largely the same way we decide whether something is a bird. In the case of a bird, we compare it with a prototype. In the case of romantic love, we compare the person’s actions and expressions with our prototype for actional/expressive outcomes of love. Indeed, this is how we come to decide that sometimes people are mistaken about being in love or not being in love. We compare one part of their behavior (the verbal expression of love or lack of love) with other parts, using our lexical prototype as a guide. What is most important for our present concerns is that this sort of comparative judgment applies not only to other people, but to oneself. Indeed, even where there is no question of my being mistaken, it is nonetheless just these prototypes that allow me to judge whether I have particular emotions. Since I do not have direct access to the “qualia” of another person, the “raw feel” of his/her emotions, how can I judge that the two of us are feeling the same thing? For example, how can a child judge that he/she is feeling something that is aptly referred to as “jealousy” or “suspicion”? In part, it is a matter of parental attribution. The parent compares the child’s situation and actions with prototypes and says, “Oh, Johnny is jealous.” In part it is also a matter of the child forming such prototypes and comparing his/her own situation and responses with them. In both cases, the learning process is the result of inference from prototype eliciting conditions and outcomes. (This general point was argued most famously by Wittgenstein.) This may seem to be a fairly simple and psychologically inconsequential process of acquiring vocabulary. But it is not. Emotional inference is not just a matter of learning the right labels for experientially self-evident and fully pregiven emotions. Some studies suggest that we must infer what we are feeling even after we have learned emotion terms. Daryl Bem’s self-perception theory contends

Afterword

241

that “Individuals come to ‘know’ their own attitudes, emotions, and other internal states partially by inferring them from observations of their own overt behavior and/or the circ*mstances in which this behavior occurs” (222), which is to say by inference from expressive/ actional outcomes and eliciting conditions. Stanley Schachter’s work is particularly relevant here. Schachter has argued that an emotion involves some state of physiological arousal plus some cognitive interpretation of that state. Thus, given “a state of arousal . . . one labels, interprets, and identifies this stirred-up state in terms of the characteristics of the precipitating situation and one’s apperceptive mass. . . . It is the cognition which determines whether the state of physiological arousal will be labeled ‘anger,’ ‘joy,’ or whatever” (402–3). In keeping with this, Nisbett and Ross argue that we judge our own inner states by much the same process of reasoning through which we judge other people’s inner states: “knowledge of the self is produced by the same strategies as knowledge of other social objects, and is thus prone to essentially the same sorts of bias and error” (195). With respect to emotions, they argue that “Once the individual becomes aware of his own state of physiological arousal, the labeling of that state – and the subjective experiences, self-reports, and emotionally relevant behavior that accompany such labeling – is the result of a search for a plausible cause of the arousal” (199). Indeed, “People’s labeling of their emotional states . . . depends on an analysis of evidence conducted in the light of preconceived theories about which antecedents produce which states and which states are the product of which antecedents” (200). I suspect that these conclusions are overstated. Indeed, there are problems with much of Schachter’s research. In one famous study (see section II of Schachter), he injected subjects with a stimulant. He then placed the subjects, one by one, in a room with a confederate who behaved either angry or euphoric. Some test subjects were told that they would feel a degree of physical arousal due to the injection. These subjects did not judge themselves either angry or euphoric after the session with the confederate. Other subjects were not told about the effects of the injection. They tended to judge themselves euphoric after some time with the euphoric confederate and angry after some time with the angry confederate. (There was also a placebo group.) Prima facie, this may seem to suggest that people

242

Afterword

interpret their emotions and do not simply feel them. However, there are other possible explanations. For example, it may be that members of the “informed” group judged the feelings of the confederate to be false, euphoria or anger resulting from the injection only. Thus, they distanced themselves from the confederate’s emotions. In contrast, members of the uninformed group were unguarded in their response to the confederate and thus tended to experience a sort of emotional contagion (a well-known and widely attested phenomenon). By this interpretation of the data, the difference in the groups is that the explanation actually changed the eliciting conditions and thus changed what emotions the test subjects felt. The informed group really did not feel anger or euphoria. The uninformed group did. It was not a matter of their interpretation of the physiological arousal produced by the injections, but of the test subjects’ actual emotive state. Nonetheless, if Schachter’s conclusions appear to be somewhat exaggerated, they are not entirely mistaken. As Ross and Nisbett explain, “While many contemporary theorists would challenge Schachter and Singer’s ideas about the lack of physiological specificity in emotional experience, few now would deny that we can be led to mislabel our feelings and to reach erroneous conclusions about the source of such feelings” (80). In short, there is an interpretive component to emotional experience, and that component is at least in part a matter of relating a raw feel to eliciting conditions and actional/expressive outcomes. On the other hand, putting the matter this way may understate the case. This formulation appears to return us to a basically preSchachterian position that the emotions are fixed in themselves. It only adds the qualification that they are not self-evident. But, in fact, emotions do not appear to be entirely fixed, even if they are also not entirely malleable. Again, Nisbett and Ross note that our inferential reasoning about our own emotions is not simply a matter of labeling, but of “the subjective experiences . . . and emotionally relevant behavior that accompany such labeling” (199). Extending Nisbett and Ross, we might say that emotions are not only identified by reference to emotion prototypes. Specific emotional experiences and outcomes are also, to some degree, shaped by our ideas about emotion, particularly by our emotion prototypes and our prototype narratives. Most

Afterword

243

obviously, we construe our experiences in light of eliciting conditions for emotions and thus in light of stories, in a process that reacts back on those experiences. We do not merely see a particular event (for example, a spouse’s behavior) in its own terms. We see it as approximating or not approximating relevant prototypes and our emotional response is guided by this. Moreover, we place ourselves in standardized narrative sequences when we act on emotions (for example, in expressing romantic love). This returns us to Keith Oatley’s point that emotions are embedded in narratives and to our development of Oatley’s ideas in Chapter 3. Once again, we have short-term aims and long-term projects. We continually evaluate events in our lives in relation to these aims and projects. Our emotions are, largely, responses to those evaluations.1 As discussed in the third chapter, these may be divided into junctural, sustaining, and outcome emotions, which are narrative categories. Moreover, our ultimate evaluative criteria, projects, are nothing other than micro-narrative eliciting conditions for happiness–or, rather, they are nothing other than the micronarratives we take to be eliciting conditions for happiness.2 As this final point suggests, the relation of emotion to narrative is perhaps most obvious with regard to eliciting conditions for happiness and our ideas about these eliciting conditions. We might, then, begin with eliciting conditions, and turn subsequently to phenomenological tone and expressive/actional outcomes. 1

I am leaving aside emotions that operate spontaneously, without significant cognitive mediation. As, for example, LeDoux points out, some perceptions directly excite fear – or do so without any higher cortical involvement, thus presumably without the sort of narrative appraisal I am discussing (see Emotional 164). For an attempt to synthesize appraisal with an account of emotions in terms of perceptual triggers, see Chapter 7 of my Cognitive Science. Indeed, emotions are narrative not only in their direct experience, but in retrospect as well, and the retrospective narratives may be quite different from the ongoing or direct, experiential narratives. Tomkins points out that someone may have a terrible experience, leaving him/her with a series of bad memories. “But what of the more permanent effect of these bad scenes on the quality of [a person’s] life?,” he asks. This effect “will depend critically on the degree of magnification which follows” the experience. Specifically, “rehearsals” of the bad memories “co-assemble them in such an order and with such spacing that they are experienced as magnifying or attenuating the negative affects” (215). In other words, the long-term emotional effects of our experiences appear to be, at least in part, a function of the narratives through which we organize and understand these experiences post facto.

244

Afterword

We do not really have very good ways to determine what will bring happiness, especially insofar as happiness is understood as an enduring condition. Indeed, we have only limited ways of determining what will bring us simple pleasure or pain. How do we know even what kind of food we will like – not to mind what kind of food will make us healthy in the long term? For the most part, we learn these things by trial and error, or by accepting the judgments of others. If we do not even have a good way of determining what foods we might enjoy, how can we possibly have a good way of determining what will make us happy for the rest of our lives, what will give us eudaimonia? We don’t. Well, we have some clues. We have friends and like spending time with them. This suggests to us that spending time with people we like will be a good thing. But there are many forms that this can take, many details to be developed. Even with regard to new foods, we always know beforehand that eating is better than starving. Moreover, we have certain general ideas about our preferences. But this does not give us some algorithm for determining specifics. Our goals, plans, and projects, then, are based only in part on our own experiences of happiness. To a great extent, they are based on what we learn from others about happiness. In forming our ideas about happiness, we form prototypical eliciting conditions that operate not only for literature, but for our lives. This is true both at the universal level and at the culturally particular level. Most if not all societies appear to establish romantic union as a – indeed, the prototype eliciting condition for happiness. This is no doubt partially based on a biological given, the sex drive. But it is also the result of human society. After all, cross-culturally, romantic union is far more than sexuality. It is a matter of living together, of sharing intimacy and affection, and of having children. It is, in other words, a very complex aspiration, and one that arises socially, even though it arises universally. In all or almost all societies, this sort of relation comes to be prized as an ideal, as the paradigm of personal happiness – even in societies where people have little say regarding their own marriage, where individual choice in marriage is officially condemned, and where romantic love appears to be denigrated or ignored (until one takes a look at the literature of the place). Indeed, there are probably many properties and general structures or functions in this

Afterword

245

area that are cross-culturally constant or near constant. For example, probably all societies envision some sense of home associated with romantic union, some connection with social work, some organized system of child-rearing, and so on. (There is a common view that universals must be biologically innate. This is simply untrue. For example, universals may arise from group dynamics. We shall return to this issue.) Of course, the precise nature of that home or work, the precise relation to the children, will vary at least to some degree from culture to culture. In fact, it is here that we can see where cultural specificity enters. Accidents of geography, economy, political history, and so on, have led to different instantiations of the romantic structure. Thus, in traditional Igbo society, the man will farm yams, the woman will grow other crops and perhaps market some goods, and the couple will, ideally, have many sons; in America in the 1950s the man will have a white collar job, the woman will stay at home with the children and the dog, and the home will be a house in the suburbs, perhaps with a picket fence. Indeed, cultural definition goes further still. We even learn what sorts of men or women we should fall in love with – in one standard clich´e, men who are tall and have good careers, women who are thin and cook well. This is not to say that we do actually fall in love with people according to social criteria of this sort. However, it is to say that we very often determine whether or not someone is a candidate for romantic involvement on the basis of such criteria (that is, on the basis of his/her similarity to the relevant paradigms). We do not allow ourselves the possibility of romantic love with some people; we open ourselves to the possibility with others. Indeed, love is not love of a person anyway, but love of an idea of a person. One’s idea of a person is based in part on the actual personality, intellect, aspiration, feeling, of the man or woman in question. But it is also built up out of various prototypes and different exempla. Thus, we come to love or not to love people in part on the basis of social ideas. Though not accounted for via an adequate psychology, the point has been recognized for quite some time. For example, this is precisely the dilemma of Emma Bovary. Indeed, Emma Bovary’s situation is particularly relevant to our discussion, for Emma comes to define her goals, and to feel romantic love, on the basis of social principles that

246

Afterword

she has learned from reading romances, which is to say, from literary narratives – and, indeed, prototypical literary narratives. Of course, Emma’s case also suggests that the eliciting conditions for happiness cannot be entirely shaped by social prototypes. There is a difference between the real eliciting conditions for happiness and the conditions defined by society. At least to some degree, there are things that genuinely make us happy and other things that do not. There is a joke about how you should never wish too hard for anything, because you might just get it. Very often, we find that the achievement of our goals is disappointing. Part of the reason for this is simply formal. Without the goal to organize our lives, ordinary activities seem to wither into insignificance. However, part of the problem is in the nature of the goals themselves. Often, socially determined eliciting conditions for happiness simply do not bring happiness. Moreover the social specification of prototype eliciting conditions is not confined to happiness. One might imagine that some emotions are simply not open to such alteration, that some are simply too direct, too biologically fixed. But this does not seem to be the case. If there were ever a prime candidate for an evolutionarily fixed emotion, that would be fear. Fear is what leads us to flee dangerous predators, thus surviving and having more offspring. But fear too is easily manipulable. Take, for example, the United States in the 1950s, where many ordinary people clearly came to fear the Soviet Union as if it were the proverbial wolf at the door – or the United States today where a South Asian woman can be reported by a frightened passenger, taken away by the police, and questioned for five hours because she remarked enthusiastically on the sights while looking out an airplane window (see Joseph). One of Margaret Atwood’s characters put the general point well: “what to feel was like what to wear, you watched the others and memorized it” (131). Indeed, contrary to one’s initial expectations, the case of fear suggests that social definitions may have their greatest impact on junctural emotions. Junctural emotions often rely on particular beliefs and expectations that may vary a great deal across cultures or historical periods. In connection with this, one might argue that social definitions have systematically different consequences for different types of emotion, as follows: junctural emotions are deeply affected by social specifications; sustaining emotions are affected primarily in altering our openness to certain goals

Afterword

247

and objects (for example, our openness to certain people as possible partners in romantic love); and outcome emotions are affected only superficially. These points bear not only on eliciting conditions, but on the remaining aspects of emotion – phenomenological tone and expressive/ actional outcomes – as well. The phenomenological tone of an emotion is “what it feels like” to have the emotion. It is the part of an emotion that is entirely inaccessible to other people. It is usually seen as both certain and definitive. However, as we have already noted, we do not simply identify our emotions with a sort of introspective directness. We not only have to determine what caused a particular “raw feel.” We often have to determine just what that feel is. This indicates that phenomenological tone too is not entirely fixed – or, at least, it may be vague, ambiguous. In any case, our understanding of it is not as absolute as we usually imagine. In consequence, the role of that raw feel in our ideas and actions is open to considerable variability. Part of this variability results from the way our society organizes ideas about emotions. For example, in the United States today, we tend to think of romantic love as an emotion not only different from, but opposed to, the feeling of friendship – hence the proverbially shattering utterance from one’s beloved, “I like you . . . as a friend.” This has all sorts of consequences. Most obviously, we may tend to segregate off people who fit our “profile” for friends, separating them from people whom we might consider possible romantic objects. In keeping with this, Forgas and Dobosz argue that “individuals from a given culture have fairly clear implicit cognitive representations of the range of relationship prototypes available in their milieu, and are likely to define their emerging contacts with others in terms of the relationship repertoire at their disposal” (290). Thus, a developing relationship between John and Jane will be understood by both parties at least partially by reference to possible relationships in their society – friendship, marriage, cohabitation, a “fling,” and so on. The relevant prototypical possibilities will help guide their feelings, their understanding of those feelings, and their behavior. For example, these prototypes serve to establish expectations, and thus, to define parameters for interpretation and evaluation of otherwise neutral events or nonevents (for example, if a goodnight kiss is

248

Afterword

expected, then its absence is reason for disappointment and perhaps for a reevaluation of the relationship). As Forgas and Dobosz put it, “relationship scenarios . . . have a crucial influence on how partners come to perceive and define their relationship and its expected progress” (291). Our cognitive organization – including our tendency to oppose feelings of love and friendship – has other effects as well. Most importantly, it may lead to differences in our understanding not only of other people and of our possible relations with them, but of our own emotions themselves. A particular raw feel may not clearly fit friendship or romantic love. We may interpret it as one or the other and act accordingly based on various socially prototypical concerns – the nature of the object (does he/she fit the romantic profile), broader aspects of the situation, and so on. Once we have understood our ambiguous feeling in a certain way, and especially after we have begun to act on that interpretation, our experience of the feeling itself will alter (though not necessarily in any simple, predictable way). Indeed, feelings are in all likelihood not only vague, but complex. We have a range of different feelings at any give time. Many are not clearly attached to one another or to any particular object. For social and other reasons, we cluster some feelings together, separating them from putatively different clusters of feelings (for example, separating sexual desire from friendship), and linking them with particular objects. This interpretive clustering and object orientation almost invariably reacts back on the feelings themselves – intensifying, diffusing, or otherwise altering them. Suppose I am tired and frustrated. Initially, both feelings may be tinged with sadness. Perhaps the tiredness has a physiological cause (for example, I am coming down with a cold) and the frustration relates to work. But I cluster the two together and attribute both to the behavior and habits of my spouse. This changes my experience almost immediately. I begin to feel the tiredness and frustration as a single emotion, now tinged with anger and resentment. In contrast, if I think there is a physiological source for the fatigue, I separate the tiredness from other raw feels, experiencing it as tinged with apprehension. The frustration, now ignored, may fade to insignificance. Clearly, I will behave very differently in each case as well (for example, if I blame my spouse, I am unlikely to respond by gargling).

Afterword

249

This brings us to our final component of emotive structure – expressive/actional outcomes. As with phenomenological tone and eliciting conditions, these are in part simply given biologically. Fear leads to flight or paralysis; mirth tends to manifest itself in laughter. But, also like raw feels and eliciting conditions, outcomes are open to a wide range of social specifications. Consider, again, romantic love. All that it clearly involves biologically is a desire for sexual union. Social conditions appear to give rise to a range of other desires and behaviors, which are themselves universal as well. These include a generalized desire for continuing proximity or living together. There is probably a biological/evolutionary kernel to this also, as I shall discuss in the next section. But the desire to live together, share a home, and so on, is almost certainly a case of a (universal) social specification. This aiming toward a shared life leads to behaviors that conduce toward living together – life choices that lead to continued proximity, and so on. As part of this desire for a shared life, there is commonly a desire for reciprocity and exclusiveness. These are bound up with outcomes such as a tendency toward jealousy (though jealousy may have other sources as well). Another part of the desire for a shared life is a mutual identification of interests. The goals and general well-being of the beloved become, to some extent, part of one’s own goals. This sharing of goals leads to a sharing of means toward the achievement of those goals – for example, a sharing of material resources. Needless to say, things do not always work out this way in practice. For example, patriarchy may be understood as, in part, a systematic skewing of these social desiderata. This leads us to the level of cultural particularity (as opposed to universal social specification). Clearly, the desire to share a life is particularized differently in different cultures. The same point holds for other actional outcomes. For example, in a capitalist economy, one standard way of expressing love is through purchasing commodities, such that more love seems broadly to imply more purchases, as well as more expensive purchases, and vice versa (as Cohen and Rogers have in part suggested [70]). This may seem to be some wholly new idea, unique to late capitalism. In fact it is merely a specification of the socially universal sharing of means. Nonetheless, it is a culturally specific version of that universal, with distinct, culturally specific, consequences.

250

Afterword

In all these cases, we find that social specifications – universal or culturally relative – are bound up with plots. They are part of implicit or explicit prototype narratives – in this case, a narrative of romantic love leading to enduring romantic union at “home,” all duly specified. Moreover, one of the ways in which these social specifications are communicated and internalized is through explicit stories, literature. Of course, another way is through observation, seeing one’s parents, and so on. However, while observation of one’s parents is certainly important and influential, it need not indicate that a particular life defines the eliciting conditions of happiness – for example, their marriage may not be ideal; it may not instantiate romantic love. Indeed, given the vacillations, uncertainties, contradictions, and general imperfectness of ordinary life, one might say that fictional narratives, imagined stories, are necessary to organize and specify our emotions. In Chapter 3, I argued that stories are connected with, indeed, formed out of emotions and emotion ideas. Here we see that emotions and emotion ideas are connected with, indeed, formed out of stories as well. More exactly, we saw previously that prototype stories are in part formed out of emotion prototypes. Those emotion prototypes are in part based in actual emotions. There are facts about emotions. However, at the same time, emotions are not fixed and fully definite. Rather, emotions – specific feelings, as experienced and enacted at a particular time and place – appear to be formed in part from emotion prototypes and from the narratives these prototypes define. Moreover, to a great extent we internalize these prototypes through the stories they have shaped. This brings us to a final topic in this first section – the relation of emotion to politics and ideology, a topic clearly raised by the preceding discussion, though one we can only touch upon in the present context. Much of the preceding discussion may seem to suggest that cultural specification is nothing more than an innocuous process of filling in some missing details in otherwise universal patterns. To a great extent, this is true. The problem, however, comes with the word “innocuous.” Our references to commodification and patriarchy indicate that the process of social specification may be far from innocent and inconsequential. In fact, emotions and standard plot structures are particularized in socially functional ways. Specifically, they tend to be defined and organized to preserve social structure.

Afterword

251

The specification of romantic love in a market economy is a specification that commodifies as many conditions and outcomes as possible. It turns the raw feel of love and the associated universal structures to the purposes of the market. Patriarchal societies develop dissymmetries throughout romantic plots, dissymmetries that operate to extend, preserve, and justify gender hierarchies.3 Of course, this is true generally, not just with respect to emotion and narrative. It is the nature of ideology to coopt whatever it can to its own purposes. Moreover, as we have repeatedly noted, there are elements in literary narrative that work against ideological conformity – the rebellion of the lovers against the parental generation with its traditional strictures of class, caste, or race, the epilogue of suffering with its humanization of and empathy for the enemy. Nonetheless, the prototype narratives, despite their dissident elements, tend to be particularized in ways that preserve social hierarchies. The very complexes we have discussed, the very structure of romantic love that develops with romantic tragi-comedy, the socially universal principles of a shared life, and so on, are broadly socially functional. They are part of a cross-cultural structure of marriage, which is a central organizing principle of human society. Indeed, one could argue that romantic and heroic tragi-comedy are prototypical precisely because they treat the most crucial aspects of social structure – reproduction and group power, the family and the nation (tribe, village, or whatever). Narratives of sacrifice, too, treat a crucial functional aspect of society – agricultural production. Of course, their functionality is not due to the agricultural efficacy of sacrifice. It seems, rather, to be bound up with the social hierarchies of scapegoating and state-authorized killing. In this way, not only cultural particulars, but cultural universals are imbued with ideology from the start. The division of emotions, the definition of emotion prototypes (including eliciting conditions and expressive/actional outcomes, even to a certain degree the raw feel of the emotion itself), and finally the prototype narratives we have been discussing, are the product, not only of cognitive structure (and human biology), but of socially functional categories and relations. 3

They also develop dissymmetries in the conceptualization and normative characterization of emotions. For a discussion of this, see Lutz.

252

Afterword

This ideology is not simple. It is continually troubled by contradictory ethical and empathic impulses. But it is nonetheless functional and pervasive.

biological bases of emotion The socially based account of emotion prototypes sketched in the preceding section is clearly a departure from the cognitive focus of earlier chapters. It may even seem contradictory with that earlier focus. However, it is not contradictory. It may seem so due to the common assumption that a cultural universal is, necessarily, biological – something “hard wired” into our brains. Moreover, in the standard view, this biological endowment is the result of evolution – and we can trace the evolutionary path largely by imagining ourselves in the place of our hunter/gatherer ancestors and guessing what sorts of things we would do in different hunter/gatherer situations. I suppose the tone of the last comment indicates my attitude toward this approach. Evolutionary accounts should, in fact, be far more complex, as such writers as Stephen Jay Gould and Noam Chomsky have argued (for a lucid treatment of the topic, see Jenkins). Despite my skepticism about current approaches, however, it is clear that there is some biological component to emotion, and some evolutionary background to this biological component. Often, writers on the biology of emotion begin with the idea of basic emotions. In Chapter 3, I expressed some skepticism about basic emotions. Specifically, I do not believe that our emotions are structured in so discrete a manner as to allow strict division into emotion categories, followed by some sort of hierarchization in which some of these well-defined emotions form the basis of other emotions. It is, I believe, a mistake to understand emotion on the model of, say, atomic physics or molecular chemistry. Unlike atoms, emotions do not begin with well-defined boundaries and fully specified constituents. On the other hand, it is clear that there are clusters of feeling, behavior, and idea, that are in effect biological givens of emotion – “proto-emotions,” as we might call them. A subset of these acquire prototype status because they are biologically salient – through, for example, spontaneous facial expressions – presumably for functional (thus evolutionary) reasons.

Afterword

253

However, having granted that, I want to stress that what I take to be biological is extremely limited. There is very little that is deterministic in my view of the physiological givens of emotion. Evolution leaves us with broad, partially inchoate, complexes of feeling and action. It does not determine what we do with those complexes. Indeed, the “endowment” itself is partially contradictory, at times adjuring incompatible behaviors and feelings. Moreover, societies can develop in such a way as to socialize emotion in ways that seem directly contrary to the biological givens. Of course, there are limits to this. However, it is my contention that the endowments are much more flexible than they might seem. Or, more properly, the givens of our emotional life – givens that we can assume to be correlated with some sort of evolutionary/genetic inheritance – are not rigidly constraining, and bear on the majority of emotional universals only in limited ways. In this section, I should like to consider what I take to be the likely candidates for our evolutionary endowment in terms of emotion, or proto-emotion. But, at the same time, I wish to stress the limits of that endowment.4 In order to begin an analysis of proto-emotions, we need to draw a few simple distinctions regarding emotions. Perhaps the most crucial division concerns the nature of their object. Specifically, we need to separate dispositional or pseudo-dispositional entities, on the one hand, and nondispositional things, on the other. A dispositional entity is anything to which one attributes intents. People are obviously dispositional entities, as are animals. But storms and sicknesses can also be understood dispositionally. Indeed, they regularly are understood in this way, at least in many societies, as their widespread deification attests. Put differently, the division into dispositional entities and nondispositional things is a division between objects that act on us (dispositional entities) and objects on which we act (nondispositional 4

My account is in some ways related to the “componential approach.” As Panksepp explains, in this view, “emotions are . . . constructed during early social development from more elemental units of visceral-autonomic experiences that accompany certain behavior patterns.” Thus, “componentialists suggest that biologically given subunits are compiled into full-blown emotional systems via cognitive appraisals and learning” (45). Of course, in my view, familial and social interactions and practices result in emotional systems that are themselves to a great extent universal. Moreover, I do not see the “components” as fully specified and constant atoms of emotion.

254

Afterword

things). Leaving aside happiness and sadness, the proto-emotions, at least those that are most salient, may be divided productively according to whether their object is understood as dispositional or not. It is worth noting that this division is salient from early childhood. For example, Steven Pinker points out that “Infants divide the world into the animate and the inert early in life” (322; Pinker divides emotions in a parallel fashion, more roughly distinguishing between “emotions about things” and “emotions about people” [374]). The most important division among dispositional objects is that between malevolent and benevolent agents. We determine malevolence and benevolence relative to our own pursuits, goals, and projects. Malevolent agents inhibit our achievement of goals; benevolent agents further our achievement of goals. Among malevolent acts, we may distinguish those that are threatening from those that are inhibitory. Threat is the eliciting condition for fear (or proto-fear) and flight is the primary outcome. Inhibition – “thwarting and frustrations,” as Panksepp puts it (52) – is the eliciting condition for anger (or proto-anger) and attack is the primary outcome. Panksepp notes that anger is a response to frustration that projects an inhibiting agent – an opponent or enemy – even when there is no such agent behind the frustration. As Panksepp puts it, “frustration emerges from the ability of . . . cognitive systems to monitor the probability of forthcoming rewards. If an expected reward is not registered, the higher cell assemblies send out opponent process messages” (197). Benevolent agents negate such malevolent actions. Thus, they protect us against threat and/or aid us in fulfilling goals. There is a range of emotive responses to benevolent behaviors. When limited, they form the eliciting conditions for gratitude (or proto-gratitude). When extended, they form the eliciting conditions for affection (more properly, protoaffection). The outcome of affection is following and joining with the benevolent agent – in other words, the reverse of flight and/or attack. (The limited/extended variable may affect other proto-emotions as well. For example, extended inhibition may provide eliciting conditions for hate.) Anger and fear are two of the emotions standardly seen as “basic.” This account adds affection (sometimes called “attachment”). On the other hand, all lists include sorrow or sadness, which is separate from this list. This is because, by the preceding account, both sadness and

Afterword

255

happiness are specified by way of these other emotions. Consider, for example, the prototype case of sadness – the loss of a person for whom one has affection. Here, sorrow is clearly reliant on another emotion, specifically affection. (Note that “reliant on” does not mean “composed of.”) This is not to say that sadness (or proto-sadness) is not part of our biological endowment. Happiness and sadness are certainly part of that endowment. But they operate only by reference to these other emotions. Indeed, affection itself is often – perhaps always – reliant on fear and anger. This is broadly in keeping with our analysis of these emotions in relation to narrative. Anger and fear are purely junctural; love, as a sustaining emotion, is more general and encompassing; and sadness or sorrow is one of the outcome emotions, thus still more inclusive, for it must take in the entire preceding narrative. Indeed, what I have just said about these proto-emotions is itself very storylike. Although some proto-emotions are biological givens, their relations are quickly defined by a social narrative, for the subject here is a protagonist seeking a goal and his/her emotions bear on helpers and opponents. Once again, the origins and development of emotion and narrative appear inseparably intertwined. Turning to nondispositional things, we may draw a distinction parallel to that regarding agents. There are things that appeal to us and things that repel us. The prototype of repulsion is decay. The decayed object provides eliciting conditions for disgust, leading to the outcome of avoidance. There is a detailed parallel with fear. The feared object is an active threat, whereas the repulsive object is a passive threat. Flight is, in effect, an intense version of avoidance, appropriate to the difference in object. Objects to which we are drawn are most obviously of two sorts – edible and sexed objects. Edible objects to which we are drawn provide partial eliciting conditions for the desire to consume and lead to the outcome of eating. Sexed objects to which we are drawn – attractive objects – provide partial eliciting conditions for lust and lead to the outcome of pursuing sexual relations. In the last two cases, I have spoken of partial eliciting conditions because lust and hunger arise spontaneously from an internal bodily cycle. Thus the desire for sexual relations and the desire to eat are not solely dependent on their object. Indeed, this is one reason why lust and hunger are often not considered to be emotions. It seems clear,

256

Afterword

however, that they belong in the same proto-emotion category as fear, and so on. After all, the relation of lust and hunger to their objects is parallel to that of disgust and the entire set shows obvious parallels with the more uncontroversial emotions of the dispositional category. Moreover, it is clear that lust and hunger – or, rather, the desire to eat – may be inspired by a particular object. Panksepp explains that “animals take large meals if their food is especially tasty but become finicky nibblers if it is not” (169). On the other hand, one hardly needs to do esoteric laboratory research to discover people saying, “I’m stuffed. But I can’t resist that ice cream.” Moreover, the inclusion of lust and hunger among emotions makes neurobiological sense. As Joseph LeDoux explains, “the brain has a number of emotion systems, including networks involved in identifying sexual partners and food sources, as well as detecting and defending against danger” (Synaptic 321; see also 221, 230, and Brothers 47). In part, the difficulty here is just terminological. First, the common term “emotion” (with the details of folk psychology that it presupposes) does not fully fit the scientifically relevant concept. Second, “hunger” does not really fit the phenomenon. Again, it is more properly called “desire to eat.”5 It is also worth mentioning that other 5

Anna Wierzbicka is probably the writer who has spent the most time cautioning researchers against linguistic biases in work on emotion (see Emotions). Cases such as this show the value of her cautions. On the other hand, Wierzbicka’s arguments go well beyond urging researchers not to over generalize from linguistic peculiarities. Indeed, there is a way in which Wierzbicka herself over generalizes from such peculiarities. For example, her criticism of Ekman (in Chapter 4 and elsewhere) involves looking at the oddities of English emotion terms and using these as an argument that Ekman’s assertions about universal facial expressions cannot be correct. In treating anger, she contends that the peculiarities of the English word “anger” are not present in all languages. Thus there is no cross-cultural concept of anger. Thus Ekman is wrong to assert that a particular facial expression is a universal expression of anger. But why should one assume that Ekman’s claims involve all the cultural uniqueness of a particular lexical item? If followed consistently, Wierzbicka’s arguments would disallow such statements as “Eating food is a human universal.” “Food” in English includes McDonald’s hamburgers. Most languages (for example, Sanskrit) have not had a term that includes McDonald’s hamburgers. Thus, there really is no equivalent for “food” in other languages. Therefore, eating food cannot be a human universal. Indeed, Wierzbicka’s reasoning would even disallow statements of physics, such as “Objects fall to the earth’s surface at a constant rate of acceleration.” After all, how many languages include precisely the semantic range of contemporary English“fall”? It seems that a more fair approach would urge caution on the part of researchers, as noted previously, but at the same time would urge readers to maintain their

Afterword

257

emotions are not so fully object dependent as we might think. Even such emotions as anger and fear are in part dependent on our own prior state when encountering the frustration or threat. However, this is not to say that the differentiae of lust and hunger are unimportant. After all, we become hungry when we have not eaten in a while. But we do not become fearful because we have not run away in a while. In fact, these differentiating characteristics are quite important. Lust and hunger provide us with the only emotions on our list that define independent goals. Hunger defines the goal of eating. Lust defines the goal of sexual union. Moreover, in both cases, this goal is itself understood as defining happiness. These are not the only outcomes that define happiness, of course. However, they are the two that define happiness independently, without reference to prior emotions or contingent circ*mstances. Before going on to this last point, however, we should briefly review the foregoing account of proto-emotions. The following list summarizes what has been said thus far. (The form of the following strings is eliciting conditions → emotion/phenomenological tone → expressive/actional outcome.) ordinary interpretive generosity. In any given case, Wierzbicka may be correct that a researcher has been misled by linguistic idiosyncracies. And, of course, she can always urge greater precision and explicitness in a theorist’s definitions. But one cannot merely assume ethnocentric error. Showing that a concept has peculiarities is not the same as showing that those peculiarities have guided a researcher’s reasoning and led to false conclusions. On the other hand, insofar as Wierzbicka does isolate genuine problems, I hope to have avoided these by linking emotion concepts with relatively concrete prototypes, by distinguishing between emotions and proto-emotions, and by considering the place of culture in the specification of emotion concepts and even in the definition, development, and experience of emotions themselves. In connection with the last point, I should perhaps mention that I am far from being alone in discussing the effects of culture on emotion. The general idea has been central to the work of social constructionists for many years – indeed, that is what makes them social constructionists (see, for example, the essays in Gergen and Davis or Harr´e and Parrott). That said, I should note that my treatment of cultural effects differs significantly from that of most social constructionists. First, I emphasize specific biological givens in emotion. Second, I argue that the social component of emotions develops along largely universal lines and that cultural differences, even in the “socially constructed” part of emotion, are relatively superficial. Finally, I have maintained elsewhere that the idea of social construction, far from being a theoretical advance on prior theories of socialization and ideology, is in fact a theoretically faulty notion that confuses some very different phenomena (see my “Ideological Critique”).

258

Afterword

Dispositional Objects: threatening object → fear → flight (or freezing) inhibiting object → anger, then hate → attack protecting/aiding object → gratitude, then affection → joining (opposite of flight and attack) Nondispositional Objects: decaying object → disgust → avoidance edible/tasty object (and hunger cycle) → desire to eat → consumption sexed/attractive object (and sexual cycle) → lust → courting or related sexual pursuit

Again, sexual relations and the consumption of food are the two independent goals yielded by this analysis. This suggests origins for two of the prototype conditions for happiness and their associated narratives.6 However, while food is as such the prototype for physical happiness (with the slight change that, as “plenty,” it is made extensive and enduring), the prototype for personal happiness is more complex. To understand this, we need to consider the relation of the other proto-emotions to happiness. We would not ordinarily say that successful flight from a threat (due to fear) or successful avoidance of decay (due to disgust) leads to happiness. Success in these actional outcomes is purely negative – unlike success in the outcomes of eating or courting. Joining with a protector/helper, however, is positive, a sort of proto-eliciting condition for happiness. Moreover, it is the most enduring of the resultant conditions. In the personal prototype for happiness, and thus in romantic stories, we find a synthesis of the independent goal of sexual union with this enduring, though partially derivative, goal of joining with a protector/ helper.

One colleague, on reading an earlier version of this argument, expressed concern that I treat emotion in relation to narrative, but children lack narratives even though they have emotions. I hope the preceding discussion of proto-emotions indicates why this objection does not have force. Specifically, it conflates an implicit narrative organization of experience with the act of verbal story telling. Children certainly experience the world in terms of agents and nonagents (for example, parents and food), as well as benevolence and malevolence (for example, feeding and spanking); children certainly have goals (for example, to get food), encounter and recognize impediments to goals (for example, an empty bottle), and so on. That is all the narrative structure required at this point.

Afterword

259

This leaves only anger/hate. Successful attack against some malevolent agent is, it seems, not merely a matter of relief, but of more positive feelings as well. This is clearly one ancestor of the prototype for happiness as social domination and for the associated prototype narrative of heroic tragi-comedy. Thus, given this analysis, we might have predicted four prototype eliciting conditions for happiness and four corresponding narratives, for lust, hunger, affection, and anger all give rise to positive feelings when their actional outcomes are successful. This is more or less what we find, with only the slight difference that the sexual and affectionate goals have been joined. This joining is one of those universals that derives, in part, from the working out of human social relations, in this case due to the social function of sex and affection in marriage systems. The general point is suggested by Panksepp, when he writes that “our cultural evolution has sought to bind our desire for sex and our need for social bonding together in an inextricable whole . . . of marriage” (226). Once this conjunction comes about, one would expect the resultant prototype (romantic love) and the corresponding genre (romantic tragi-comedy) to be exceptionally prominent, for it combines two sets of prototype eliciting conditions for happiness. This is, again, precisely what we find.7 The structuring of anger/hate into domination is complex as well. It is inseparable from the – highly socially functional, and also universal – development of categorial identity and in-group/outgroup relations, which are distinct from the biological givens of emotion per se. Here, as with romantic love – and, indeed, with the sacrificial element in narratives of hunger – societies develop the proto-emotions, both conceptually and practically, by extending and specifying them in ways that contribute to the preservation of the society. Indeed, we arguably have a sort of social evolution here. Societies that develop in-group/out-group definitions, specifying 7

Of course, the eliciting conditions for affection and lust do define narratives separately as well. In keeping with the fact that lust reaches its full development much later than the other proto-emotions, this narrative division is age related. I realized this when Keith Oatley pointed out to me that many Disney films (for example, Bambi) are based on nonromantic “attachment” themes. Such films take up the goals defined by affection without combining them with lust. Complementary to this, “adult” narratives, in the euphemistic sense, focus on the goals of lust, with relative indifference to affection/attachment.

260

Afterword

proto-emotions in relation to these definitions, may be more likely to survive as coherent cultures, especially when faced with external threat. (They are probably more likely to pose a threat to other societies as well.) In this way, the universality of the complex emotion prototypes, and the resultant complex, narrative universals, may be seen in part as outcomes of selection pressures on groups rather than individuals. (In this way, a tendency to value, for example, in-group domination could be universal without being based in some unalterable, biological propensity of all – or even any – individual people.) All this helps us to understand why literature takes particular forms. I hope that it also helps us to understand the human mind more broadly. But at this point one might ask what sort of emotive interest might lead to the development of literature initially. Why would we want to read literature or listen to a story in the first place? The reasons are certainly emotive. But they do not obviously derive from any emotions discussed thus far. Here, we need to return to our set of proto-emotions. The division of emotions based on type of object is clearly incomplete, for it leaves out emotions that do not bear on objects, but on ambient conditions. Indeed, these are perhaps the most fundamental in our experience – for the entire sequence is hierarchical. Every dispositional object is also a nondispositional object. Every nondispositional object is also experienced as an alteration in ambient conditions. Our response to ambient conditions is, then, in some sense the most definitive. The first proto-emotion in this category has as its eliciting conditions an excess of stimulation; its actional response is withdrawal from stimulation. We might refer to this as “sensitivity.” This obviously parallels the withdrawal from nondispositional objects in disgust and the withdrawal from dispositional objects in fear. Temperature regulation appears to be our prototype case for this. The fundamental place of ambient condition emotions is suggested by the fact that thermal regulation is probably the most common source domain for metaphors about more standard or prototypical emotions – anger and lust are heat; affection is warmth; fear is cold. The second proto-emotion of this sort is the reverse – the pursuit of stimulation, parallel to the pursuit of food or aggressive approach. This is roughly what Panksepp sees as a sort of generalized urge to act. He calls it “seeking,” but since there is no particular object

Afterword

261

being sought, this is a somewhat misleading term. In fact, the protoemotion he isolates is more or less what we ordinarily refer to as “boredom,” or “boredom” plus “curiosity.” The eliciting conditions for boredom/curiosity are a lack of stimulation – an obvious parallel to the lack of food, and a less obvious parallel to the success deprivation that triggers anger. The actional outcome of boredom/curiosity is simply seeking stimulation. In the proto-emotion of boredom/curiosity, we have one source of interest in literature. Literature provides stimulation – and precisely the sort of stimulation we lack in ordinary life. Stories about individual and in-group domination, stories about love – in other words, prototype stories – provide us with precisely the sort of stimulation we might seek, but they do so in an innocuous form. In real battles, after all, one can really be killed. The stimulation of an imaginary battle is less than the stimulation of a real battle. But the risks are reduced as well. Note that this division also suggests why we might find some types of movie or novel unpleasant and why we might turn away – put the book down, leave the theater. Relative to one’s individual sensibility, the stimulation provided by, say, a particular horror film might be excessive; for someone else, this may not be the case at all. On the other hand, the pursuit of stimulation is not all there is to aesthetic experience. Moreover, sensitivity and boredom/curiosity should not be the only proto-emotions bearing on ambient conditions. I shall end this section with one last feeling that I take to be part of our emotional endowment. From a very early age, children delight in patterned sounds and images. For example, as to the former, from about seven months, babies begin to engage in spontaneous rhythmic babbling (Locke 283). This undoubtedly has a function in aiding language acquisition, as Locke and others have argued. However, from the baby’s point of view, that is irrelevant, at least initially. Infants engage in rhythmic babbling at least in part out of simple enjoyment. In short, they are not entirely hard-headed pragmatists setting out to practice their language skills. They are also playful children having fun. Propensities to enjoy verbal and visual symmetries, rhythms, and so on, develop as we grow, yielding an emotive relation to objects that involves attraction, but is not a matter of affection or lust. This is a sort of proto-aesthetic sense. Its eliciting conditions are what we

262

Afterword

call “beauty.” Its associated feeling is a proto-form or ancestor of what we awkwardly refer to as aesthetical pleasure, but is perhaps more aptly characterized (following the Sanskrit theorists) as wonder or wonder/delight. Its associated actional outcome is contemplation – just looking or just listening. When successful, its result is not precisely happiness, but rather what Abhinavagupta said is the ultimate goal of all art – peace. The point is no doubt related to Gordon Orians’ idea that we find just those landscapes beautiful where we feel safe, for the feeling of safety is a necessary part of peace, or proto-peace (on Orians’ work, see Pinker 376–7 and citations). Wonder completes the triad of ambient condition proto-emotions. But, of course, it is not confined to ambient conditions. Just as we may have a nondispositional emotion directed toward a dispositional object (for example, disgust with the body of another person), we may have wonder directed toward an object, nondispositional or dispositional (for example, wondrous delight at the beloved). Again, the proto-emotions are hierarchized such that the more “fundamental” proto-emotions apply at all the higher levels. Moreover, like lust, wonder is bound up with affection in complex ways. Even when the object of our contemplation is not dispositional per se, even when we are contemplating nature or some artifact, we regularly transfer feelings of affection to some associated dispositional entity – the person behind the beautiful face, the creator of the beautiful thing (for discussion, see Ong “Jinnee,” and my “Beautiful”; the point has implications for attitudes ranging from fascination with the lives of authors to religious sentiment, as Ong has indicated). The point certainly bears on the personalizing tendency of arts discussed in the preceding chapter, as well as literary empathy and related aspects of literary development and experience. The addition of wonder here suggests that our enjoyment of art, including verbal art, derives straightforwardly from the biological givens with which we start our social and personal lives. Indeed, our delight in art results from the goals of two proto-emotions, goals that are no less independent and positive than eating and sex – the goals defined by boredom/curiosity and wonder. (Perhaps a similar point could be made about science, which is not unrelated to curiosity and wonder as well.) Of course, as in the other cases, the proto-emotion of wonder is vague, inchoate, open to alteration, redevelopment, and

Afterword

263

specification through society and individual biography. But it is part of what we all begin with – and no less universal than other emotions, not only in its biology, but in much of its social development as well. I imagine that most people who would accept this idea would nonetheless say that the propensity to wonder is not functional; it is not a development that directly serves our adaptive needs, but is rather a by-product of some other, adaptive development (for example, delight in rhythmic sounds is merely a by-product of commitment to language learning). Literary critics sometimes take offense when evolutionary theorists claim that literary properties did not develop from the pragmatic survival of the fittest. Personally, I rather like the idea that literature is different from big business and the military, that we are magnificently untainted by the survival of the fittest, with its unpleasant suggestions of large male brutes pummeling their competitors into a bloody mush, then taking all the food and all the women – perhaps with an occasional large female brute doing more or less the same. On the other hand, I do wonder if the ability to contemplate beauty is not adaptive. Once our mental lives are filled with fear, anger, lust, disgust, and so on, the contemplative peace brought by a sense of beauty seems as if it would not be wholly lacking in practical value, even for our ancient ancestors.8

conclusion In sum, it appears that our emotions are much more malleable than we are inclined to believe. They are formed from biologically given protoemotions – proto-fear, proto-anger, proto-affection, and so on – that bear on ambient conditions, nondispositional objects, or dispositional objects. These proto-emotions are specified by socially functional practices and ideas that affect eliciting conditions, phenomenological tone, and actional/expressive outcomes. These specifications are, in part, culturally distinctive. But universals are not confined to biology. Social specifications are to a great extent universal as well. This should not come as a surprise, for the intimate interactions of individuals and 8

Tooby and Cosmides present a very different, adaptational account of aesthetic enjoyment in babbling and elsewhere (16–17). For a discussion of some problems with this account, see Chapter 8 of my Cognitive Science.

264

Afterword

the social relations of groups are patterned in common ways. They are not random and wildly diverse, but quite consistent. Noam Chomsky has repeatedly argued that differences among human languages, though highly salient, are structurally minuscule. As he put the point in one recent essay, a “Martian scientist” who has come to study Earth “might reasonably conclude that there is a single human language, with differences only at the margins” (New Horizons 7). In my view, the same point holds for culture in general, including broad areas of culture that are not simply biologically determined but result from social life as well. Among these are the almost inseparable phenomena of narrative and emotion – prominently including romantic, heroic, and sacrificial tragi-comedy, along with the happiness prototypes they specify and the junctrual and sustaining emotions they elaborate. These derive from universal social developments of biologically given proto-emotions – social developments and proto-emotions that also account for our engagement with verbal art and its unique, powerful, and universal satisfactions.

Works Cited

Abhinavagupta. The Aesthetic Experience According to Abhinavagupta. 2nd ed. Ed. and trans. Raniero Gnoli. Varanasi, India: Chokhamba Sanskrit Series, 1968. ¯ . “Locana.” In The Dhvanyaloka of Anandavardhana with the Locana of Abhinavagupta. Trans. Daniel Ingalls, Jeffrey Masson, and M.V. Patwardhan. Cambridge, MA: Harvard University Press, 1990. Abrams, M. H. A Glossary of Literary Terms. 5th ed. Fort Worth, TX: Holt, Rinehart and Winston, 1988. Abu-Lughod, Lila. “Shifting Politics in Bedouin Love Poetry.” In Language and the Politics of Emotion. Ed. Catherine A. Lutz and Lila Abu-Lughod. Cambridge: Cambridge University Press and Paris: Editions de la Maison des Sciences de l’Homme, 1990. Ahmad, Aijaz. In Theory: Classes, Nations, Literatures. New York: Verso, 1992. Amaladass, Anand. Philosophical Implications of Dhvani: Experience of Symbol Language in Indian Aesthetics. Vienna: De Nobili Research Library, 1984. Amin, Samir. Eurocentrism. Trans. Russell Moore. New York: Monthly Review Press, 1989. Amnesty International. “Deterrence.” http://www.aiusa.org/abolish/ deterrence.html. ¯ Anandavardhana. “Dhvanyaloka.” In The Dhvanyaloka of Anandavardhana with the Locana of Abhinavagupta. Trans. Daniel Ingalls, Jeffrey Masson, and M. V. Patwardhan. Cambridge, MA: Harvard University Press, 1990. Anscombe, G. E. M. “On the Grammar of ‘Enjoy.’” In Metaphysics and the Philosophy of Mind: Collected Philosophical Papers. Vol. 2. Minneapolis: University of Minnesota Press, 1981. Appiah, Kwame Anthony. In My Father’s House: Africa in the Philosophy of Culture. New York: Oxford University Press, 1992. 265

266

Works Cited

Aristotle. “Nicomachean Ethics.” Trans. W. D. Ross. In Introduction to Aristotle. Ed. Richard McKeon. New York: Modern Library, 1947. . “Peri Poietikes.” In Aristotle’s Theory of Poetry and Fine Art. 4th ed. Ed. and trans. S. H. Butcher. New York: Dover, 1951. Ashcroft, Bill, Gareth Griffiths, and Helen Tiffin. The Empire Writes Back: Theory and Practice in Post-Colonial Literatures. New York: Routledge, 1989. Aston, W. G. A History of Japanese Literature. London: William Heinemann, 1899. Atwood, Margaret. Surfacing. New York: Fawcett Crest, 1972. Awolalu, J. Omosade. Yoruba Beliefs and Sacrificial Rites. Harlow, United Kingdom: Longman, 1979. Badawi, M. M. Early Arabic Drama. Cambridge: Cambridge University Press, 1988. Bal, Mieke. Narratology: Introduction to the Theory of Narrative. 2nd ed. Toronto: University of Toronto Press, 1997. Barnard, Philip J. “Interacting Cognitive Subsystems: Modeling Working Memory Phenomena Within a Multiprocessor Architecture.” In Models of Working Memory: Mechanisms of Active Maintenance and Executive Control. Ed. Akira Miyake and Priti Shah. Cambridge: Cambridge University Press, 1999. Barsalou, Lawrence W. “Ad Hoc Categories.” Memory and Cognition 11.3 (1983). Bateson, Mary Catherine. 1970. Structural Continuity in Poetry: A Linguistic Study in Five PreIslamic Arabic Odes. The Hague: Mouton, 1970. Bem, Daryl J. “Self-Perception Theory.” In Cognitive Theories in Social Psychology. Ed. Leonard Berkowitz. New York: Academic Press, 1978. Bernheimer, Charles, et al. “A Statement of Purpose: Comparative Literature at the Turn of the Century.” Report circulated in October 1993 to members of the American Comparative Literature Association. Berggren, Paula. Teaching with the Norton Anthology of World Masterpieces, Expanded Edition: A Guide for Instructors. New York: W. W. Norton, 1995. Berndt, Ronald M. Love Songs of Arnhem Land. Chicago: University of Chicago Press, 1976. ´ astra. Delhi, India: Sri Satguru Publications, n.d. Bharatamuni. The N¯a.tya S¯ Bh¯asa. “The Broken Thighs.” In Thirteen Plays of Bh¯asa. Vol. II. Ed. and trans. A. C. Woolner and Lakshman Sarup. Delhi: Motilal Banarsidass, 1985. . “The Minister’s Vows.” In Thirteen Plays of Bh¯asa. Vol. I. Ed. and trans. A. C. Woolner and Lakshman Sarup. Delhi: Motilal Banarsidass, 1985. . “The Vision of V¯asavadatt¯a.” In Thirteen Plays of Bh¯asa. Vol. I. Ed. and trans. A. C. Woolner and Lakshman Sarup. Delhi: Motilal Banarsidass, 1985. ¯ Ed. and trans. M. R. Kale. ¯ The Uttarar¯amacharita of Bhavabhuti. Bhavabhuti. Delhi: Motilal Banarsidass, 1988. Biebuyck, Daniel. Notes to The Mwindo Epic. In The Mwindo Epic. ShéKárisi Rureke. Ed. and trans. Daniel Biebuyck and Kahombo C. Mateene. Berkeley: University of California Press, 1971.

Works Cited

267

Bierhorst, John, ed. and trans. Four Masterworks of American Indian Literature: Quetzalcoatl/The Ritual of Condolence/Cuceb/The Night Chant. New York: Farrar, Straus and Giroux, 1974. Blair, R. J. R. “A Cognitive Developmental Approach To Morality: Investigating the Psychopath.” Cognition 57 (1995). The Book of Dede Korkut: A Turkish Epic. Trans. and ed. Faruk Sumer, Ahmet E. Uysal, and Warren S. Walker. Austin: University of Texas Press, 1972. Bordwell, David. “Convention, Construction, and Cinematic Vision.” In Post-Theory: Reconstructing Film Studies. Ed. David Bordwell and Noel Carroll. Madison: University of Wisconsin Press, 1996. . Narration in the Fiction Film. Madison: University of Wisconsin Press, 1985. Boulton, Maureen and Barry McCann. The Song in the Story: Lyric Insertions in French Narrative Fiction, 1200–1400. Philadelphia: University of Pennsylvania Press, 1993. Bower, G. H. “Affect and Cognition.” Philosophical Transactions of the Royal Society of London, Series B 302 (1983). . “Mood and Memory.” American Psychologist 36 (1981). Bownas, Geoffrey. “Introduction.” In The Penguin Book of Japanese Verse. Ed. Geoffery Bownas and Anthony Thwaite. London: Penguin, 1967. and Thwaite, Anthony, eds. The Penguin Book of Japanese Verse. London: Penguin, 1964. Brandon, James R. The Cambridge Guide to Asian Theatre. Cambridge: Cambridge University Press, 1993. . “Introduction.” In Kabuki: Five Classic Plays. Ed. James R. Brandon. Honolulu: University of Hawaii Press, 1992. , ed. and trans. Kabuki: Five Classic Plays. Honolulu: University of Hawaii Press, 1992. , ed. Traditional Asian Plays. New York: Hill and Wang, 1972. Brenan, Gerald. St. John of the Cross: His Life and Poetry. Trans. of his poetry by Lynda Nicholson. Cambridge: Cambridge University Press, 1973. Brewer, William F. “Literary Theory, Rhetoric, and Stylistics: Implications for Psychology.” In Theoretical Issues in Reading Comprehension: Perspectives from Cognitive Psychology, Linguistics, Artiﬁcial Intelligence, and Education. Ed. Rand J. Spiro, Bertram C. Bruce, and William F. Brewer. Hillsdale, NJ: Lawrence Erlbaum, 1980. . “The Story Schema: Universal and Culture-Specific Properties.” In Literacy, Language, and Learning: The Nature and Consequences of Reading and Writing. Ed. David R. Olson, Nancy Torrance, and Angela Hildyard. Cambridge: Cambridge University Press, 1985. and Edward H. Lichtenstein. “Event Schemas, Story Schemas, and Story Grammars.” In Attention and Performance IX. Ed. John Long and Alan Baddeley. Hillsdale, NJ: Lawrence Erlbaum, 1981. Brewitt-Taylor, C. H., trans. Romance of the Three Kingdoms (San Guo Yan Yi). By Luo Guan Zhong. 2 vols. Singapore: Graham Brash, 1991.

268

Works Cited

Brothers, Leslie. Friday’s Footprint: How Society Shapes the Human Mind. Oxford: Oxford University Press, 1997. Brower, Robert H. and Earl Miner. Japanese Court Poetry. Stanford, CA: Stanford University Press, 1961. Brown, Donald. Human Universals. Philadelphia, PA: Temple University Press, 1991. Cameron, James, dir. Titanic. Script by James Cameron. Photography by Russell Carpenter. Edited by Conrad Buff, James Cameron, and Richard A. Harris. Paramount Pictures, 1997. Cao Xueqin. The Story of the Stone. Vol. 1. Trans. David Hawkes. New York: Penguin, 1973. Cao Xueqin and Gao E. The Story of the Stone. Vol. 4. Trans. John Minford. New York: Penguin, 1982. . The Story of the Stone. Vol. 5. Trans. John Minford. New York: Penguin, 1986. Carroll, Joseph. “Universals in Literary Study.” In Universalien und Konstruktivismus. Ed. Peter Hejl. Frankfurt am Main: Suhrkamp, forthcoming. Carroll, No¨el. “Art, Narrative, and Emotion.” In Emotion and the Arts. Mette Hjort and Sue Laver, eds. New York: Oxford University Press, 1997. Chadwick, Nora K. and Victor Zhirmunsky. Oral Epics of Central Asia. Cambridge: Cambridge University Press, 1969. ¯ Chaereas and Callirrhoe. Trans. B. P. Reardon. In Collected Ancient Chariton. Greek Novels. Berkeley: University of California Press, 1989. Chaucer, Geoffrey. The Canterbury Tales. Ed. John Halverson. Indianapolis, IN: Bobbs-Merrill, 1971. Chelkowski, Peter, ed. Ta’ziyeh: Ritual and Drama in Iran. New York: New York University Press, 1979. ¨ Plays. Chˆeng Teh-hui. “The Soul of Ch’ien-Nu¨ Leaves Her Body.” In Six Yuan Ed. and trans. Liu Jung-en. New York: Penguin, 1972. Chikamatsu Monzaemon. “The Courier for Hell.” Trans. Donald Keene. In Masterpieces of the Orient. Ed. G. L. Anderson. Expanded Edition, New York: W. W. Norton, 1977. . Major Plays of Chikamatsu. Ed. and trans. Donald Keene. New York: Columbia University Press, 1990. Chomsky, Noam. Aspects of the Theory of Syntax. Cambridge, MA: MIT Press, 1965. . Knowledge of Language: Its Nature, Origin, and Use. New York: Praeger, 1986. . Language and Mind. New York: Harcourt, Brace, Jovanovich, 1972. . New Horizons in the Study of Language and Mind. Cambridge: Cambridge University Press, 2000. Chung, Ling. “Biography of Li Ch’ing-chao.” In Complete Poems. Li Ch’ingChao. Trans. and ed. Kenneth Rexroth and Ling Chung. New York: New Directions, 1979. Clinton, Jerome. “The Book of Dede Korkut.” In The Norton Anthology of World Masterpieces, Volume 1: Beginnings to 1650. Expanded ed. Ed. Maynard Mack et al. New York: W. W. Norton, 1995.

Works Cited

269

Cohen, Joshua and Joel Rogers. On Democracy: Toward a Transformation of American Society. New York: Penguin, 1983. Comrie, Bernard. Language Universals and Linguistic Typology: Syntax and Morphology. 2nd ed. Chicago: University of Chicago Press, 1989. Conway, M. A. and D. A. Bekerian. “Situational Knowledge and Emotions.” Cognition and Emotion 1.2 (1987). Cooper, Arthur, ed. and trans. Li Po and Tu Fu. London: Penguin, 1973. Crane, Mary. Shakespeare’s Brain: Reading with Cognitive Theory. Princeton, NJ: Princeton University Press, 2000. Croft, William. Typology and Universals. Cambridge: Cambridge University Press, 1990. Cronyn, George W., ed. American Indian Poetry: An Anthology of Songs and Chants. New York: Fawcett Columbine, 1991. Cullmann, Oscar. The State in the New Testament. London: SCM P, 1957. ¯ In The Norton Anthology of World Danly, Robert Lyons. “The Kokinshu.” Masterpieces, Volume 1: Beginnings to 1650. Expanded ed. Ed. Maynard Mack et al. New York: W. W. Norton, 1995. Davidson, Donald. Inquiries into Truth and Interpretation. Oxford: Clarendon Press, 1985. Davis, Mark H. Empathy: A Social Psychological Approach. Madison, WI: Brown and Benchmark, 1994. Dawson, Michael R. W. Understanding Cognitive Science. Oxford: Blackwell, 1998. Dawood, N. J., trans. Tales from the Thousand and One Nights. New York: Penguin, 1961. Deng, Francis Mading. The Dinka and their Songs. Oxford: Clarendon Press, 1973. de Sousa, Ronald. “The Rationality of Emotions.” In Explaining Emotions. Ed. Amelie Oksenberg Rorty. Berkeley: University of California Press, 1980. ¯ Dhanam A Treatise on Hindu Dramaturgy. Trans. George . jaya. The Da´sarupa: C. O. Haas. New York: AMS P, 1965. Dimock, Edward C., Jr. “The R¯am¯ayan.a Story.” In The Literatures of India: An Introduction. Dimock, Edwin Gerow, C. M. Naim, A. K. Ramanujan, Gordan Roadarmel, and J. A. B. van Buitenen. Chicago, IL: University of Chicago Press, 1978. , Edwin Gerow, C. M. Naim, A. K. Ramanujan, Gordon Roadarmel, and J. A. B. van Buitenen. The Literatures of India: An Introduction. Chicago, IL: University of Chicago Press, 1978. Dissanayake, Ellen. hom*o Aesheticus: Where Art Comes From and Why. New York: Free Press, 1992. Dixon, R. M. W. and Duwell, Martin, eds. The Honey-Ant Men’s Love Song and Other Aboriginal Song Poems. Melbourne: University of Queensland Press, 1990. Dolby, William, ed. and trans. Eight Chinese Plays From the Thirteenth Century to the Present. New York: Columbia University Press, 1978. Doniger, Wendy. The Implied Spider: Politics and Theology in Myth. New York: Columbia University Press, 1998.

270

Works Cited

Duckitt, J. H. The Social Psychology of Prejudice. New York: Praeger, 1992. Dundas, Paul. The Jains. New York: Routledge, 1992. Edmonson, Munro S. Lore: An Introduction to the Science of Folklore and Literature. New York: Holt, Rinehart and Winston, 1971. Egudu, Romanus and Donatus Nwoga. Igbo Traditional Verse. London: Heinemann, 1973. Ekman, Paul. “Cross-Cultural Studies in Facial Expression.” In Darwin and Facial Expression: A Century of Research in Review. Ed. Ekman. New York: Academic Press, 1973. Euripides. The Trojan Women. In Euripides: Ten Plays. Trans. and ed. Moses Hadas and John McLean. New York: Bantam Books, 1960. ` o. ` The Epic of Son-Jara. In The Epic of Son-Jara: A West African TradiFa-Digi Sisok tion. Ed. and trans. John William Johnson. Bloomington: Indiana University Press, 1992. Fanon, Frantz. Black Skin, White Masks. Trans. Constance Farrington. New York: Grove Press, 1963. Ferdowsi, Abolqasem. The Epic of the Kings: Sh¯ah-N¯ama. An abridged version. Trans. and ed. Reuben Levy. Costa Mesa, CA: Mazda Publishers in association with Bibliotheca Persica, 1996. . The Tragedy of Sohráb and Rostám: From the Persian National Epic, the Shâhnâme of Abol-Qasem Ferdowsi. Trans. Jerome W. Clinton. Seattle: University of Washington Press, 1987. Firth, Raymond with Mervyn McLean. Tikopia Songs: Poetic and Musical Art of a Polynesian People of the Solomon Islands. Cambridge: Cambridge University Press, 1990. Fodor, Jerry A. Concepts: Where Cognitive Science Went Wrong. Oxford: Clarendon Press, 1998. Forgas, Joseph Paul. “Affect in Social Judgments and Decisions: A Multiprocess Model.” In Advances in Experimental Social Psychology. Vol. 25. Ed. Mark P. Zanna. San Diego, CA: Academic Press, 1972. . “The Effects of Prototypicality and Cultural Salience on Perceptions of People.” Journal of Research in Personality 17 (1983). . “Mood and the Perception of Unusual People: Affective Asymmetry in Memory and Social Judgments.” European Journal of Social Psychology 22 (1992). and Barbara Dobosz. “Dimensions of Romantic Involvement: Towards a Taxonomy of Heterosexual Relationships.” Social Psychology Quarterly 43 (1980). and Stephanie J. Moylan. “Affective Influences on Stereotype Judgements.” Cognition and Emotion 5 (1991). Fowler, Barbara Hughes, ed. and trans. Archaic Greek Poetry: An Anthology. Madison: University of Wisconsin Press, 1992. Frijda, Nico. H. “Comment on Oatley and Johnson-Laird’s ‘Towards a Cognitive Theory of Emotions.’” Cognition and Emotion 1.1 (1987). . The Emotions. Cambridge: Cambridge University Press and Paris: Editions de la Maison des Sciences de l’Homme, 1986.

Works Cited

271

. “The Laws of Emotion.” In Human Emotions: A Reader. Ed. Jennifer M. Jenkins, Keith Oatley, and Nancy L Stein. Oxford: Blackwell, 1998. Frye, Northrop. Anatomy of Criticism: Four Essays. Princeton, NJ: Princeton University Press, 1957. , Sheridan Baker, and George Perkins. The Harper Handbook to Literature. New York: Harper and Row, 1985. Gadamer, Hans-Georg. Truth and Method. 2nd ed. Trans. Joel Weinsheimer and Donald Marshall. New York: Crossroad, 1989. Gardner, Howard. Art, Mind, and Brain: A Cognitive Approach to Creativity. New York: Basic Books, 1982. Gardner, John and John Maier, eds. and trans. (with the assistance of Richard A. Henshaw). Gilgamesh. (Sin-Leqi-Unninni version.) New York: Vintage Books, 1985. Garman, Michael. Psycholinguistics. Cambridge: Cambridge University Press, 1990. Gathercole, Susan. “Models of Verbal Short-Term Memory.” In Cognitive Models of Memory. Ed. Martin A. Conway. Cambridge, MA: MIT Press, 1997. and Alan Baddeley. Working Memory and Language. Hove, East Sussex: Lawrence Erlbaum, 1993. Gelpke, Rudolf. “Translator’s Preface.” In The Story of Layla and Majnun. Nizami. Ed. and trans. Rudolf Gelpke. New Lebanon, NY: Omega Publications, 1997. Gergen, Kenneth J. and Keith E. Davis, eds. The Social Construction of the Person. New York: Springer-Verlag, 1985. Gerow, Edwin. Indian Poetics. Wiesbaden: Otto Harrassowitz, 1977. Gerstle, C. Andrew. “The Concept of Tragedy in Japanese Drama.” Japan Review 1 (1990). Gibbs, Raymond W. The Poetics of Mind: Figurative Thought, Language, and Understanding. Cambridge: Cambridge University Press, 1994. Gifford, Don with Robert J. Seidman. Ulysses Annotated: Notes for James Joyce’s Ulysses. Revised and expanded ed. Berkeley: University of California Press, 1988. Gillegan, Carol. In a Different Voice: Psychological Theory and Women’s Development. Cambridge: Harvard University Press, 1982. Girard, Ren´e. Violence and the Sacred. Trans. Patrick Gregory. Baltimore, MD: Johns Hopkins University Press, 1977. Gnoli, Raniero. The Aesthetic Experience According to Abhinavagupta. Varanasi: Chowkhamba Sanskrit Series Office, 1968. Greimas, A. J. Structural Semantics: An Attempt at a Method. Trans. Daniele McDowell, Ronald Schleifer, and Alan Velie. Lincoln: University of Nebraska Press, 1983. Grice, Paul. “Logic and Conversation.” In Studies in the Way of Words. Cambridge, MA: Harvard University Press, 1989. ¨ Grunbaum, Adolf. The Foundations of Psychoanalysis: A Philosophical Critique. Berkeley: University of California Press, 1984.

272

Works Cited

Guépin, J. P. The Tragic Paradox: Myth and Ritual in Greek Tragedy. Amsterdam: A. M. Hakkert, 1968. Gurg¯an¯ı, Fakhr ud-Din. V¯ıs and R¯am¯ın. Trans. George Morrison. New York: Columbia University Press, 1972. Hadas, Moses. Introduction to The Trojan Women. In Euripides: Ten Plays. Trans. and ed. Moses Hadas and John McLean. New York: Bantam Books, 1960. ´ “Effect and reminding in literary text processing.” In EmpiriHalász, Lászlo. cal Studies of Literature: Proceedings of the Second IGEL Conference, Amsterdam 1989. Elrud Ibsch, Dick Schram, and Gerard Steen, eds. Amsterdam: Rodopi, 1991. . “Emotional Effect and Reminding in Literary Processing.” Poetics 20 (1991). Halford, Aubrey S. and Giovanna M. Halford. The Kabuki Handbook. Rutland VT: Charles E. Tuttle, 1956. Hampton, James A. “Psychological Representation of Concepts.” In Cognitive Models of Memory. Ed. Martin A. Conway. Cambridge, MA: MIT Press, 1997. Harré, R. and W. G. Parrott, eds. The Emotions: Social, Cultural, and Biological Dimensions. London: Sage, 1996. Harris, Paul, Tjeert Olthof, Mark Terwogt, and Charlotte Hardman. “Children’s Knowledge of the Situations that Provoke Emotion.” International Journal of Behavioral Development 10 (1987). Hart, George. The Poems of Ancient Tamil. Berkeley: University of California Press, 1975. . Poets of the Tamil Anthologies: Ancient Poems of Love and War. Princeton, NJ: Princeton University Press, 1979. Hauser, Arnold. The Social History of Art (Volume I): Prehistoric, AncientOriental, Greece and Rome; Middle Ages. Trans. Stanley Godman. New York: Vintage Books, 1957. Hawkes, David. “The Quest of the Goddess.” In Studies in Chinese Literary Genres. Ed. Cyril Birch. Berkeley: University of California Press, 1974. Heath, Peter. The Thirsty Sword: S¯ırat ‘Antar and the Arabic Popular Epic. Salt Lake City: University of Utah Press, 1996. Herman, David. “Scripts, Sequences, and Stories: Elements of a Postclassical Narratology.” PMLA 112. 5 (1997). Hernadi, Paul. Beyond Genre: Tragicomedies of History on the Modern Stage. Ithaca, NY: Cornell University Press, 1972. . Interpreting Events: Tragicomedies of History on the Modern Stage. Ithaca, NY: Cornell University Press, 1985. Hjort, Mette and Sue Laver. “Introduction.” In Emotion and the Arts. Ed. Mette Hjort and Sue Laver. New York: Oxford University Press, 1997. Hobbs, Jerry. Literature and Cognition. Stanford, CA: Center for the Study of Language and Information, 1990. Hogan, Patrick Colm. “The Beautiful and the Merely Pleasing: Love, Art, and the Jinnee in the Well Wrought Urn.” Time, Memory and the Verbal Arts:

Works Cited

273

Essays on the Thought of Walter Ong. Ed. Dennis Weeks. Selinsgrove, PA: Susquehanna University Press, 1998. . “Beauty, Politics, and Cultural Otherness: The Bias of Literary Difference.” In Literary India: Comparative Studies in Aesthetics, Colonialism, and Culture. Ed. Patrick Hogan and Lalita Pandit. Albany: State University of New York Press, 1995. . Cognitive Science, Literature, and the Arts: A Guide for Humanists. New York: Routledge, 2003. . Colonialism and Cultural Identity: Crises of Tradition in the Anglophone Literatures of India, Africa, and The Caribbean. Albany: State University of New York Press, 2000. . The Culture of Conformism: Understanding Social Consent. Durham, NC: Duke University Press, 2001. . Empire and Poetic Voice: Cognitive and Cultural Studies of Literary Tradition and Colonialism. Albany, NY: State University of New York Press, forthcoming. . “Literary Universals.” Poetics Today 18 (1997). . “Ideological Critique and the Profession of Literature.” The Review of Education/Pedagogy/Cultural Studies 19.1 (1997). . On Interpretation: Meaning and Inference in Law, Psychoanalysis, and Literature. Athens: University of Georgia Press, 1995. . “Particular Myths, Universal Ethics: Wole Soyinka’s The Swamp Dwellers in the New Nigeria.” Modern Drama 41.4 (1999). . Philosophical Approaches to the Study of Literature. Gainesville: University Press of Florida, 2000. . “The Possibility of Aesthetics.” British Journal of Aesthetics 34.4 (1994). . “Shakespeare, Eastern Theatre, and Literary Universals: Drama in the Context of Cognitive Science.” In Shakespeare East and West. Ed. Minoru Fujita. Richmond, Surrey: Japan Library of Curzon Press, 1996. . “Some Prolegomena to the Study of Literary Difference.” Poetics 22 (1994). ¯ . “Toward a Cognitive Science of Poetics: Anandavardhana, Abhinavagupta, and the Theory of Literature.” College Literature 23.1 (1996). Holland, John, Keith Holyoak, Richard Nisbett, and Paul Thagard. Induction: Processes of Inference, Learning, and Discovery. Cambridge, MA: MIT Press, 1986. Holland, Norman. The Brain of Robert Frost. New York: Routledge, 1988. . 5 Readers Reading. New Haven, CT: Yale University Press, 1975. . Poems in Persons: An Introduction to the Psychoanalysis of Literature. New York: W. W. Norton, 1975. and Murray Schwartz. “The Delphi Seminar.” College English 36 (1975). Homer. The Iliad. Trans. W. H. D. Rouse. New York: New American Library, 1938. . The Odyssey. Trans. A. T. Murray and rev. George E. Dimock. Cambridge, MA: Harvard University Press, 1995.

274

Works Cited

Hsia, C. T. “The Military Romance.” In Studies in Chinese Literary Genres. Ed. Cyril Birch. Berkeley: University of California Press, 1974. . “A Dream of Red Mansions.” In Eastern Canons: Approaches to the Asian Classics. Ed. William Theodore de Bary and Irene Bloom. New York: Columbia University Press, 1990. Humayuni, Sadeq. “An Analysis of the Ta’zayih of Qasem.” In Ta’ziyeh: Ritual and Drama in Iran. Ed. Peter Chelkowski. New York: New York University Press, 1979. Huot, Sylvia. “Troubadour Lyric and Old French Narrative.” In The Troubadours: An Introduction. Ed. Simon Gaunt and Sarah Kay. Cambridge: Cambridge University Press, 1999. ¯ All¯ah. Ibn Ish¯aq. The Life of Muhammad: A Translation of Ish¯aq’s S¯ırat Rasul Trans. and ed. A. Guillaume. London: Oxford University Press, 1955. Ilangoˆ Adigal. Shilappadikaram. Trans. Alain Dani´elou. New York: New Directions, 1965. ¯ Ingalls, Daniel H. H., ed. The Dhvanyaloka of Anandavardhana with the Locana of Abhinavagupta. Trans. Daniel Ingalls, Jeffrey Masson, and M. V. Patwardhan. Cambridge, MA: Harvard University Press, 1990. Ingarden, Roman. The Literary Work of Art: An Investigation on the Borderlines of Ontology, Logic, and the Theory of Literature. Trans. George Grabowicz. Evanston, IL: Northwestern University Press, 1973. Irele, F. Abiola. “Africa: The Mali Epic of Son-Jara.” In The Norton Anthology of World Masterpieces, Volume 1: Beginnings to 1650. Expanded ed. Ed. Maynard Mack et al. New York: W. W. Norton, 1995. Jakobson, Roman. Language in Literature. Ed. Krystyna Pomorska and Stephen Rudy. Cambridge, MA: Harvard University Press, 1987. Jankowiak, W. R. and E. F. Fischer. “A Cross-Cultural Perspective on Romantic Love.” In Human Emotions: A Reader. Ed. Jennifer Jenkins, Keith Oatley and Nancy Stein. Oxford: Basil Blackwell, 1998. Jauss, Hans Robert. Toward an Aesthetic of Reception. Trans. Timothy Bahti. Minneapolis: University of Minnesota Press, 1982. Jenkins, Jennifer M., Keith Oatley, and Nancy L. Stein, eds. Human Emotions: A Reader. Oxford: Basil Blackwell, 1998. Jenkins, Lyle. Biolinguistics: Exploring the Biology of Language. Cambridge: Cambridge University Press, 2000. Jerusalem Bible. Gen. Ed. Alexander Jones. Garden City, NY: Doubleday, 1966. Johnson, John William. “The Epic of Son-Jara.” In Oral Epics from Africa: Vibrant Voices from a Vast Continent. Ed. John William Johnson, Thomas A. Hale, and Stephen Belcher. Bloomington: Indiana University Press, 1997. Johnson-Laird, Philip N. The Computer and the Mind: An Introduction to Cognitive Science. Cambridge: Harvard University Press, 1988. and Keith Oatley. “The Language of Emotions: An Analysis of a Semantic Field.” Cognition and Emotion 3.2 (1989). and P. C. Wason, eds. Thinking: Readings in Cognitive Science. New York: Cambridge University Press, 1977.

Works Cited

275

Joseph, George. “Indian Actress Creates Flutter On Board Plane.” India Abroad (July 26, 2002). Joyce, James. Ulysses. Ed. Hans Walter Gabler with Wolfard Steppe and Claus Melchior. New York: Vintage Books, 1986. Kahneman, Daniel and Dale T. Miller. “Norm Theory: Comparing Reality to Its Alternatives.” Psychological Review 93.2 (1986). Kailasapathy, K. Tamil Heroic Poetry. Oxford: Clarendon Press, 1968. ˙ . ¯ı: The Saga of the Kings of Ka´sm¯ır. Trans. and ed. Ranjit Kalhan.a. R¯ajatarangin Sitaram Pandit. New Delhi: Sahitya Akademi, 1968. ´ ˜ ana´sa¯ kuntalam (The Recognition of Sakuntal¯ a).” In K¯alid¯asa. “Abhijn¯ K¯alid¯asa: The Loom of Time. Trans. and ed. Chandra Rajan. New Delhi: Penguin, 1989. . Theater of Memory: The Plays of K¯alid¯asa. Ed. Barbara Stoler Miller. Trans. Edwin Gerow, David Gitomer, and Barbara Stoler Miller. New York: Columbia University Press, 1984. Kaut.ilya. Artha´sa¯ stra. 6th ed. Trans. R. Shamasastry. Mysore: Mysore Printing and Publishing, 1960. Kazin, Alfred, ed. The Portable Blake. New York: Viking, 1968. Keene, Donald, ed. Anthology of Japanese Literature to the Nineteenth Century. New York: Penguin, 1968. , ed. and trans. Major Plays of Chikamatsu. New York: Columbia University Press, 1990. , ed. and trans. Twenty Plays of the N¯o Theatre. New York: Columbia University Press, 1970. King, Bruce. Derek Walcott: A Caribbean Life. Oxford: Oxford University Press, 2000. Kintsch, Walter, Alice Healy, Mary Hegarty, Bruce Pennington, and Timothy Salthouse. “Models of Working Memory: Eight Questions and Some General Issues.” In Models of Working Memory: Mechanisms of Active Maintenance and Executive Control. Ed. Akira Miyake and Priti Shah. Cambridge: Cambridge University Press, 1999. Kiparsky, Paul. P¯an.ini as a Variationist. Ed. S. D. Joshi. Cambridge, MA: MIT Press, 1979. . “The Role of Linguistics in a Theory of Poetry.” In Essays in Modern Stylistics. Ed. Donald C. Freeman. London: Methuen, 1981. . “On Theory and Interpretation.” In The Linguistics of Writing: Arguments Between Language and Literature. Ed. Nigel Fabb, Derek Attridge, Alan Durant, and Colin MacCabe. New York: Methuen, 1987. Klein, Stanley, Judith Loftus, J. Gregory Trafton, and Robert W. Fuhrman. “Use of Exemplars and Abstractions in Trait Judgments: A Model of Trait Knowledge About Self and Others.” Journal of Personality and Social Psychology 63 (1992). Klemenz-Belgardt, Edith. “American Research on Response to Literature: The Empirical Studies.” Poetics 10.4/5 (1981). Knight, W. F. Jackson. “Introduction.” In The Aenid. Virgil. Trans. W. F. Jackson Knight. New York: Penguin, 1958.

276

Works Cited

Knox, Bernard. Introduction and Notes. In The Three Theban Plays. Sophocles. Trans. Robert fa*gles. Intro. and notes Bernard Knox. New York: Penguin, 1984. Kouyaté, Djeli Mamoudou. Sundiata: An Epic of Old Mali. Ed. D. T. Niane. Trans. G. D. Pickett. New York: Longman, 1965. ¨ Kovecses, Zoltán. Metaphor and Emotion: Language, Culture, and Body in Human Feeling. Cambridge: Cambridge University Press and Paris: Editions de la Maison des Sciences de l’Homme, 2000. ¨ Plays. Ed. and Kuan Han-ch’ing. The Injustice Done to Tou Ngo. In Six Yuan trans. Liu Jung-en. New York: Penguin, 1972. Kunene, Daniel P. Heroic Poetry of the Basotho. Oxford: Clarendon Press, 1971. Labouvie-Vief, Gisela. Psyche and Eros: Mind and Gender in the Life Course. Cambridge: Cambridge University Press, 1994. Labov, William. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press, 1972. Lakatos, Imre. “Falsification and the Methodology of Scientific Research Programmes.” In Criticism and the Growth of Knowledge. Ed. Imre Lakatos and Alan Musgrave. Cambridge: Cambridge University Press, 1970. Lakoff, George. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago, IL: University of Chicago Press, 1987. and Mark Johnson. Metaphors We Live By. Chicago, IL: University of Chicago Press, 1980. Lal, P. Great Sanskrit Plays. New York: New Directions, 1964. Lameer, Joep. Al-F¯ar¯ab¯ı and Aristotelian Syllogistics: Greek Theory and Islamic Practice. Leiden: E. J. Brill, 1994. ´ and Uffe Seilman. “Across Time and Place: Larsen, Steen, János Lászlo, Cultural-Historical Knowledge and Personal Experience in Appreciation of Literature.” In Empirical Studies of Literature: Proceedings of the Second IGEL Conference, Amsterdam 1989. Elrud Ibsch, Dick Schram, and Gerard Steen, eds. Amsterdam: Rodopi, 1991. LeDoux, Joseph. The Emotional Brain: The Mysterious Underpinnings of Emotional Life. New York: Touchstone, 1996. . Synaptic Self: How Our Brains Become Who We Are. New York: Viking, 2002. Lerdahl, Fred and Ray Jackendoff. A Generative Theory of Tonal Music. Cambridge, MA: MIT Press, 1983. Levinson, Jerrold. “Emotion in Response to Art: A Survey of the Terrain.” In Emotion and the Arts. Ed. Mette Hjort and Sue Laver. New York: Oxford University Press, 1997. Lévi-Strauss, Claude. The Raw and the Cooked (Introduction to a Science of Mythology, Vol. 1). Trans. John and Doreen Weightman. New York: Harper and Row, 1969. . From Honey to Ashes (Introduction to a Science of Mythology, Vol. 2). Trans. John and Doreen Weightman. New York: Harper and Row, 1973. Levy, Reuben. An Introduction to Persian Literature. New York: Columbia University Press, 1969.

Works Cited

277

Lewis, Geoffrey, ed. and trans. The Book of Dede Korkut. New York: Penguin, 1974. Lichtenstadter, Ilse. Introduction to Classical Arabic Literature. New York: Schocken Books, 1976. Li Ch’ing-Chao. Complete Poems. Trans. and ed. Kenneth Rexroth and Ling Chung. New York: New Directions, 1979. Lincoln, Kenneth. “Foreword.” In American Indian Poetry: An Anthology of Songs and Chants. Ed. George W. Cronyn. New York: Fawcett Columbine, 1991. Ling Chung, “Biography of Li Ch’ing-Chao.” In Li Ch’ing-Chao. Complete Poems. Trans. and ed. Kenneth Rexroth and Ling Chung. New York: New Directions, 1979. Liu Hsieh. The Literary Mind and the Carving of Dragons. Trans. Vincent Yuchung Shih. New York: Columbia University Press, 1959. Liu, James J. Y. The Art of Chinese Poetry. Chicago, IL: University of Chicago Press, 1962. . “Some Literary Qualities of the Lyric (Tz’u).” In Studies in Chinese Literary Genres. Ed. Cyril Birch. Berkeley: University of California Press, 1974. ¨ Plays. New York: Penguin, 1972. Liu Jung-en, ed. and trans. Six Yuan Locke, John L. “Development of the Capacity for Spoken Language.” In The Handbook of Child Language. Ed. Paul Fletcher and Brian MacWhinney. Oxford: Basil Blackwell, 1995. Lord, Albert. The Singer of Tales. Cambridge, MA: Harvard University Press, 1960. Luk´acs, Georg. Essays on Realism. Ed. Rodney Livingstone. Trans. David Fernbach. Cambridge, MA: MIT Press, 1981. Luo Guanzhong. Three Kingdoms: A Historical Novel. Ed. and trans. Moss Roberts. Beijing: Foreign Language Press and Berkeley: University of California Press, 1999. Lutz, Catherine A. “Engendered Emotion: Gender, Power, and the Rhetoric of Emotional Control in American Discourse.” In Language and the Politics of Emotion. Ed. Catherine A. Lutz and Lila Abu-Lughod. Cambridge: Cambridge University Press and Paris: Editions de la Maison des Science de l’Homme, 1990. Lutz, Catherine A. and Lila Abu-Lughod, eds. Language and the Politics of Emotion. Cambridge: Cambridge University Press and Paris: Editions de la Maison des Science de l’Homme, 1990. ¨ Ma Chih-yuan. Autumn in Han Palace. In Six Yuan Plays. Ed. and trans. Liu Jung-en. New York: Penguin, 1972. Mack, Maynard, et al., eds. The Norton Anthology of World Masterpieces, Volume 1: Beginnings to 1650. Expanded ed. New York: W. W. Norton, 1995. Mahapatra, Sitakant, ed. and trans. Staying is Nowhere: An Anthology of Kondh and Paraja Poetry. Calcutta, India: Writers Workshop, n.d. Marcus, Ruth Barcan. “Moral Dilemmas and Consistency.” The Journal of Philosophy 77.3 (March 1980).

278

Works Cited

Markandaya, Kamala. Nectar in a Sieve. New York: New American Library, 1982. Martinovitch, Nicholas N. The Turkish Theatre. New York: Theatre Arts, 1933. Marx, Karl. Economic and Philosophic Manuscripts of 1844. Ed. Dirk Struik. Trans. Martin Milligan. New York: International Publishers, 1964. . Capital, Vol. 1: A Critical Analysis of Capitalist Production. Ed. Frederick Engels. Trans. Samuel Moore and Edward Aveling. New York: International Publishers, 1967. Mathews, A. “Biases in Emotional Processing.” In Human Emotions: A Reader. Ed. Jennifer M. Jenkins, Keith Oatley, and Nancy L. Stein. Oxford: Basil Blackwell, 1998. McClelland, J. and D. Rumelhart. “Distributed Memory and the Representation of General and Specific Information.” Journal of Experimental Psychology: General 114 (1985). . “A Distributed Model of Human Learning and Memory.” In Parallel Distributed Processing, Vol. 2. Ed. J. McClelland and D. Rumelhart. Cambridge, MA: MIT Press, 1986. McLeod, Peter, Kim Plunkett, and Edmund T. Rolls. Introduction to Connectionist Modelling of Cognitive Processes. Oxford: Oxford University Press, 1998. Miller, Barbara Stoler. “Introduction.” In Love Song of the Dark Lord: Jayadeva’s G¯ıtagovinda. Ed. and Trans., Brabara Stoley Miller. New York: Columbia University Press, 1977. Miller, George and Philip Johnson-Laird. Language and Perception. Cambridge, MA: Harvard University Press, 1976. Miller, Peggy J., Lisa Hoogstra, Judith Mintz, Heidi Fung, and Kimberly Williams. “Troubles in the Garden and How They Get Resolved: A Young Child’s Transformation of His Favorite Story.” In Memory and Affect in Development: The Minnesota Symposia on Child Psychology. Vol. 26. Ed. Charles A. Nelson. Hillsdale, NJ: Lawrence Erlbaum, 1993. Milton, John. Paradise Lost. In Paradise Lost, Paradise Regained, Samson Agonistes. Intro. Harold Bloom. New York: Collier, 1962. Miyake, Akira and Priti Shah. “Toward Unified Theories of Working Memory: Emerging General Consensus, Unresolved Theoretical Issues, and Future Research Directions.” In Models of Working Memory: Mechanisms of Active Maintenance and Executive Control. Ed. Akira Miyake and Priti Shah. Cambridge: Cambridge University Press, 1999. Monier-Williams, Monier. A Sanskrit-English Dictionary. New ed. Delhi: Motilal Banarsidass, 1997. Moreh, Shmuel. Live Theatre and Dramatic Literature in the Medieval Arab World. New York: New York University Press, 1992. Morrison, George. “Introduction.” In V¯ıs and R¯am¯ın. Fakhr ud-Din Gurg¯an¯ı. Trans. George Morrison. New York: Columbia University Press, 1972. Murnaw, F. W., dir. Nosferatu, A Symphony of Horror. Screenplay by Henrik ¨ Galeen. Cinematography by Gunther Krampf and Fritz Arno Wagner. 1922.

Works Cited

279

Murphy, Roland E. “Psalms, Book of.” In The Oxford Companion to the Bible. Ed. Bruce M. Metzger and Michael D. Coogan. New York: Oxford University Press, 1993. ¯ ¯ o¯ I, Namiki Geiji, and Asada Itcho. ¯ Chronicle Namiki Sosuke with Namiki Shoz of the Battle of Ichinotani. In Kabuki: Five Classic Plays. Ed. and trans. James R. Brandon. Honolulu: University of Hawaii Press, 1992. Narayan, R. K. The Ramayana. New York: Penguin, 1972. N¯ar¯ayan.a. Hitopade´sa. Trans. and ed. M. R. Kale. Delhi, India: Motilal Banarsidass, 1967. Nelson, Katherine. “Events, Narratives, Memory: What Develops?” In Memory and Affect in Development: The Minnesota Symposia on Child Psychology (Volume 26). Ed. Charles A. Nelson. Hillsdale, NJ: Lawrence Erlbaum, 1993. ˜ ı wa Thiong’o. Moving the Centre: The Struggle for Cultural Freedoms. Ngug˜ London: James Curry, 1993. Nisbett, Richard and Lee Ross. Human Inference: Strategies and Shortcomings of Social Judgment. Englewood Cliffs, NJ: Prentice-Hall, 1980. Nizami. The Story of Layla and Majnun. Ed. and trans. Rudolf Gelpke in collaboration with E. Mattin and G. Hill. Final chapter trans. Zia Inayat Khan and Omid Safi. New Lebanon, NY: Omega Publications, 1997. Oatley, Keith. Best Laid Schemes: The Psychology of Emotions. Cambridge: Cambridge University Press and Paris: Editions de la Maison des Sciences de l’Homme, 1992. . “Emotions and the Story Worlds of Fiction.” In Narrative Impact. Ed. Melanie Green, Jeffrey Strange, and Timothy Brock. Mahwah, NJ: Lawrence Erlbaum, 2002. and P. N. Johnson-Laird. “Toward a Cognitive Theory of Emotions.” Cognition and Emotion 1.1 (1987). O’Connor-Visser, E. A. M. E. Aspects of Human Sacriﬁce in the Tragedies of Euripides. Amsterdam: B. R. Gruner Publishing, 1987. Ohnuki-Tierney, Emiko. The Ainu of the Northwest Coast of Southern Sakhalin. New York: Holt, Rinehart and Winston, 1974. ¯ Okabou Ojobolo. The Ozidi Saga. Ed. and trans. J. P. Clark. Ibadan, Nigeria: Ibadan University Press and Oxford University Press, 1977. Ong, Walter J., S.J. “The Jinnee in the Well-Wrought Urn.” In The Barbarian Within and Other Fugitive Essays and Studies. New York: Macmillan, 1962. . Orality and Literacy: The Technologizing of the Word. New York: Methuen, 1982. Opland, Jeff. Xhosa Oral Poetry: Aspects of a Black South African Tradition. Cambridge: Cambridge University Press, 1983. Ortony, Andrew “Beyond Literal Similarity.” Psychological Review 86 (1979). and Gerald L. Clore. “Emotions, Moods, and Conscious Awareness: Comment on Johnson-Laird and Oatley’s ‘The Language of Emotions: An Analysis of a Semantic Field.’” Cognition and Emotion 3.2 (1989). , Gerald Clore, and Allan Collins. The Cognitive Structure of Emotions. Cambridge: Cambridge University Press, 1988.

280

Works Cited

Owen, Stephen. “The Nine Songs.” In The Norton Anthology of World Masterpieces, Volume 1: Beginnings to 1650. Expanded ed. Ed. Maynard Mack et al. New York : W. W. Norton, 1995. Padgett, Ron, ed. The Teachers & Writers Handbook of Poetic Forms. New York: Teachers & Writers Collaborative, 1987. Pagels, Elaine. The Origin of Satan. New York: Vintage, 1995. Pandit, B. N. Essence of the Exact Reality or Param¯arthas¯ara of Abhinavagupta. Delhi: Munshiram Manoharlal, 1991. Pandit, Lalita. “Caste, Race, and Nation: Dialectic in Rabindranath Tagore’s Gora.” In Literary India: Comparative Studies in Aesthetics, Colonialism, and Culture. Ed. Patrick Colm Hogan and Pandit. Albany: State University of New York Press, 1995. Panksepp, Jaak. Affective Neuroscience: The Foundations of Human and Animal Emotions. New York: Oxford University Press, 1998. Pavel, Thomas G. The Poetics of Plot: The Case of English Renaissance Drama. Minneapolis: University of Minnesota Press, 1985. Philippi, Donald L. Songs of Gods, Songs of Humans: The Epic Tradition of the Ainu. San Francisco: North Point Press, 1982. Pinker, Steven. How the Mind Works. New York: W. W. Norton, 1997. Plantinga, Carl. “Film Theory and Aesthetics: Notes on a Schism.” The Journal of Aesthetics and Art Criticism 51 (1993). Prince, Gerald. A Grammar of Stories. The Hague: Mouton, 1973. . Narratology: The Form and Functioning of Narrative. Berlin: Mouton, 1982. Puenzo, Luis, dir. The Plague. Screenplay by Luis Puenzo. Based on La Peste by Albert Camus. Photography by Félix Monti. Edited by Juan Carlos Mac´ıas. Companie Française Cinématographique, The Pepper-Prince Company, Oscar Kramer Producciones SA, 1992. Pukui, Mary Kawena and Alfons L. Korn. The Echo of Our Song: Chants and Poems of the Hawaiians. Honolulu: University Press of Hawaii, 1973. Quint, David. Epic and Empire: Politics and Generic Form from Virgil to Milton. Princeton, NJ: Princeton University Press, 1993. Racine, Jean. Phèdre. In Théaˆ tre Complet: II. Ed. André Stegmann. Paris: Flammarion, 1965. Richardson, Alan. “Rethinking Romantic Incest: Human Universals, Literary Representation, and the Biology of Mind.” New Literary History 31 (2000). Richman, Paula. “E. V. Ramasami’s Reading of the R¯am¯ayan.a.” In Many R¯am¯ayan.as. Ed. Paula Richman, 1991. . “Introduction: The Diversity of the R¯am¯ayan.a Tradition.” In Many R¯am¯ayan.as, Ed. Paula Richman, 1991. , ed. Many R¯am¯ayan.as: The Diversity of a Narrative Tradition in South Asia. Berkeley: University of California Press, 1991. Ross, Lee and Richard E. Nisbett. The Person and the Situation: Perspectives of Social Psychology. Philadelphia, PA: Temple University Press, 1991. Rubin, David C. Memory in Oral Traditions: The Cognitive Psychology of Epic, Ballads, and Counting-out Rhymes. New York: Oxford University Press, 1995.

Works Cited

281

Rumelhart, David E. “Notes on a Schema for Stories.” In Representation and Understanding: Studies in Cognitive Science. Ed. Daniel G. Bobrow and Allan Collins. New York: Academic Press, 1975. Rureke, Shé-kárisi. The Mwindo Epic. Ed. and trans. Daniel Biebuyck and Kahombo C. Mateene. Berkeley: University of California Press, 1971. Russell, James. “In Defense of a Prototype Approach to Emotion Concepts.” Journal of Personality and Social Psychology 60.1 (1991). Russell, Ralph. The Pursuit of Urdu Literature: A Select History. London: Zed Books, 1992. Said, Edward. “Bookless in Gaza.” The Nation 263.8 (September 23, 1996). ´ ¯ scaryacud ¯ . aman.i. Trans. V. Raghavan. In The Wondrous CrestSaktibhadra. A´ ¯ scaryacud ´ ¯ . aman.i of Saktibhadra Jewel in Performance: Text and Translation of the A´ ¯ . iy¯a.t.tam Sanskrit Drama. with the Production Manual from the Tradition of Kut Ed. Clifford Reis Jones. Delhi, India: Oxford University Press, 1984. Sandars, N. K. “Gilgamesh.” In The Norton Anthology of World Literature. Expanded ed. Gen. ed. Maynard Mack. New York: W. W. Norton, 1997. Sarbin, Theodore R. “Believed-In Imaginings: A Narrative Approach.” In Believed-In Imaginings: The Narrative Construction of Reality. Ed. Joseph De Rivera and Theodore R. Sarbin. 1998. Sartre, Jean-Paul. Being and Nothingness: A Phenomenological Essay on Ontology. Trans. Hazel Barnes. New York: Washington Square, 1966. Sartne, Jean-Paul.. L’être et le néant, essai d’ontologie phénoménologuque. Paris: Gallimard, 1943. Schachter, Stanley. “The Interaction of Cognitive and Physiological Determinants of Emotional State.” In Cognitive Theories in Social Psychology. Ed. Leonard Berkowitz. New York: Academic Press, 1978. Schacter, Daniel L. Searching for Memory: The Brain, the Mind, and the Past. New York: Basic Books, 1996. Schank, Roger C. “Explanation: A First Pass.” In Experience, Memory, and Reasoning. Ed. Janet L. Kolodner and Christopher K. Riesbeck. Hillsdale, NJ: Lawrence Erlbaum, 1986. Schank, Roger C. and Robert P. Abelson. Scripts, Plans, Goals, and Understanding: An Inquiry into Human Knowledge Structures. Hillsdale, NJ: Lawrence Erlbaum, 1977. Schwarz, Norbert. “Social Cognition: Information Accessibility and Use in Social Judgment.” In Thinking: An Invitation to Cognitive Science. Vol. 3. 2nd ed. Cambridge, MA: MIT Press. Seilman, S. and S. F. Larsen. “Personal Resonance to Literature: A Study of Remindings While Reading.” Poetics 18 (1989). Shaver, Phillip, Judith Schwartz, Donald Kirson, and Cary O’Connor. “Emotion Knowledge: Further Exploration of a Prototype Approach.”Journal of Personality and Social Psychology 52.6 (1987). Sherzer, Joel. “Poetic Structuring of Kuna Discourse: The Line.” In Native American Discourse: Poetics and Rhetoric. Ed. Sherzer and Anthony Woodbury. Cambridge: Cambridge University Press, 1987.

282

Works Cited

Sherzer, Joel and Anthony Woodbury. “Introduction.” In Native American Discourse: Poetics and Rhetoric. Ed. Sherzer and Woodbury. Cambridge: Cambridge University Press, 1987. Shakespeare, William. Hamlet. Ed. Louis B. Wright and Virginia A. LaMar. New York: Washington Square Press, 1959. . Macbeth. Ed. Louis B. Wright and Virginia A. LaMar. New York: Washington Square Press, 1959. . The Tragedy of Romeo and Juliet. In The Riverside Shakespeare. Ed. G. Blakemore Evans. Boston, MA: Houghton Mifflin, 1974. Shanks, David R. “Representation of Categories and Concepts in Memory.” In Cognitive Models of Memory. Ed. Martin A. Conway. Cambridge, MA: MIT Press, 1997. Shibatani, Masayoshi. The Languages of Japan. Cambridge: Cambridge University Press, 1990. Shirane, Haruo. “The Poetry of Matsuo Basho.” In Eastern Canons: Approaches to the Asian Classics. Ed. William Theodore de Bary and Irene Bloom. New York: Columbia University Press, 1990. Singha, Rina and Reginald Massey. Indian Dances: Their History and Growth. New York: G. Braziller, 1967. ´ Sissoko Kabine, et al. Da Monzon de Segou: Epop´ ee Bambara. 4 vols. Ed. and trans. Lilyan Kesteloot with A. Traore, J.-B. Traore, and Amadou Hampate Ba. Paris: Fernand Nathan Editeur, 1972. Sivan, Santosh, dir. The Terrorist. Story by Santosh Sivan. Screenplay by Santosh Sivan, Ravi Deshpande, and Vijay Deveshwar. Photography by Santosh Sivan. Edited by Sreekar Prasad. Moderne Gallerie Motion Picture – Wonderfilms, 1998. Sommers, Christina. “Filial Morality.” The Journal of Philosophy 83.8 (August 1986). Sophocles. The Three Theban Plays. Trans. Robert fa*gles. Intro. and notes Bernard Knox. New York: Penguin, 1984. Soyinka, Wole. “The Bacchae of Euripides.” In Collected Plays 1. Oxford: Oxford University Press, 1973. . Death and the King’s Horseman. New York: Hill and Wang, 1987. . “The Strong Breed.” In Collected Plays 1. Oxford: Oxford University Press, 1973. . “The Swamp Dwellers.” In Collected Plays 1. Oxford: Oxford University Press, 1973. Sperber, D. and D. Wilson. Relevance: Communication and Cognition. Oxford: Basil Blackewell, 1986. Spolsky, Ellen. Gaps in Nature: Literary Interpretation and the Modular Mind. Albany: State University of New York Press, 1993. ¨ “Stratagem of Interlocking Rings.” In Six Yuan Plays. Ed. and trans. Liu Jung-en. New York: Penguin, 1972. Suso, Bamba. In Sunjata. Bamba Suso. Trans. and annotated Gordon Innes. New York: Penguin, 1999. and Banna Kanute. Sunjata. Trans. and annotated by Gordon Innes. Ed. and intro. Lucy Dur´an and Graham Furniss. New York: Penguin, 1999.

Works Cited

283

The Tale of the Heike. Trans. Helen Craig McCullough. Stanford, CA: Stanford University Press, 1988. Tan, Ed S. Emotion and the Structure of Narrative Film: Film as an Emotion Machine. Trans. Barbara Fasting. Mahwah, NJ: Lawrence Erlbaum, 1996. Terence. The Comedies. Ed. and trans. Betty Radice. New York: Penguin, 1976. Todorov, Tzvetan. The Fantastic: A Structural Approach to a Literary Genre. Trans. Richard Howard. Cleveland, OH: Case Western Reserve University Press, 1973. Tomkins, S. S. “Script Theory: Differential Magnification of Affects.” In Human Emotions: A Reader. Ed. Jennifer M. Jenkins, Keith Oatley, and Nancy L. Stein. Oxford: Basil Blackwell, 1998. Tooby, John and Leda Cosmides. “Does Beauty Build Adapted Minds? Toward an Evolutionary Theory of Aesthetics, Fiction, and the Arts.” SubStance: A Review of Theory and Literary Criticism 94/95 (2001). Trask, Willard R., ed. The Unwritten Song: Poetry of the Primitive and Traditional Peoples of the World, Volume I: The Far North/Africa/Indonesia/ Melanesia/Australia. New York: Macmillan, 1966. Turner, Frederick. “The Neural Lyre: Poetic Meter, The Brain, and Time.” In Natural Classicism: Essays on Literature and Science. New York: Paragon House Publishers, 1985. Turner, Mark. The Literary Mind. New York: Oxford University Press, 1996. . Reading Minds: The Study of English in the Age of Cognitive Science. Princeton, NJ: Princeton University Press, 1991. Tutuola, Amos. The Palm-Wine Drinkard and My Life in the Bush of Ghosts. New York: Grove Press, 1994. Tversky, Amos. “Features of Similarity.” Psychological Review 84 (1977). V¯alm¯ıki. Srimad V¯alm¯ıki R¯am¯ayan.a. 3 vols. Trans. N. Ragunathan. Madras, India: Vighneswara Publishing, 1981. Varley, H. Paul. “The Tale of the Heike.” In Eastern Canons: Approaches to the Asian Classics. Ed. William Theodore de Bary and Irene Bloom. New York: Columbia University Press, 1990. Virgil. The Aeneid. Trans. W. F. Jackson Knight. Revised ed. New York: Penguin Books, 1958. . P. Vergili Maronis: Opera. Ed. R. A. B. Mynors. Oxford: Clarendon Press, 1969. Vi´sa¯ khadatta. Mudr¯ar¯akshasa. 6th ed. Ed. and trans. M. R. Kale. Delhi: Motilal Banarsidass, 1976. Voight, Ellen Bryant. “Narrative and Lyric: Structural Corruption.” The Southern Review 30 (1994). Vy¯asa. Mahabharata of Krishna-Dwaipayana Vyasa. 4 vols. Trans. Kisari Mohan Ganguli. New Delhi: Munshiram Manoharlal, 1970. Waley, Arthur, ed. and trans. The Book of Songs. Boston: Houghton Mifflin, 1937. , ed. and trans. The N¯o Plays of Japan. London: Unwin, 1988. Watson, Burton. Early Chinese Literature. New York: Columbia University Press, 1962.

284

Works Cited

Watanabe, Hitoshi. The Ainu Ecosystem: Environment and Group Structure. Seattle: University of Washington Press, 1972. White, Hayden. Metahistory: The Historical Imagination in Nineteenth-Century Europe. Baltimore, MD: Johns Hopkins University Press, 1973. Wierzbicka, Anna. “Emotions.” In Semantic Primitives. Trans. Anna Wierzbicka and John Besemeres. Frankfurt am Main, West Germany: Athenaum Verlag, 1972. . Emotions Across Languages and Cultures: Diversity and Universals. Cambridge: Cambridge University Press, 1999. Williams, Mark. “Mansfield in Maoriland: Biculturalism, Agency and Misreading.” In Modernism and Empire. Ed. J. Howard Booth and Nigel Rigby. Manchester: Manchester University Press, 2000. Wilson, Edward O. Consilience: The Unity of Knowledge. New York: Vintage, 1998. Wittgenstein, Ludwig. Philosophical Investigations. 3rd ed. Trans G. E. M. Anscombe. New York: Macmillan, 1958. Woolner, A. C. and Lakshman Sarup, eds. and trans. Thirteen Plays of Bh¯asa. Delhi: Motilal Banarsidass, 1985. Wright, Louis B. and Virginia A. LaMar, eds. Hamlet. In Shakespeare, Hamlet. , eds. Macbeth. In Shakespeare, Macbeth. Yarshater, Ehsan. “The Development of Iranian Literatures.” In Persian Literature. Ed. Yarshater. N.C.: Bibliotheca Persica, 1988. . “Foreword.” In The Epic of the Kings: Shah-nama by Ferdowsi. Ed. and trans. Reuben Levy. Costa Mesa, CA: Mazda Publishers with Bibliotheca Persica, 1996. Yeats, William Butler. The Collected Plays of W. B. Yeats. New ed. New York: Macmillan, 1965. Young, Kay and Jeffrey L. Saver. “The Neurology of Narrative.” SubStance: A Review of Theory and Literary Criticism 94/95 (2001). Yuasa, Nobuyuki, ed. and trans. Bash¯o: The Narrow Road to the Deep North and Other Travel Sketches. Baltimore, MD: Penguin, 1966. Zeami Motokiyo. “Atsumori.” Trans. Arthur Waley. In Masterpieces of the Orient. Expanded ed. Ed. G. L. Anderson. New York: Norton, 1977. Zhang Longxi. Mighty Opposites: From Dichotomies to Differences in the Comparative Study of China. Stanford, CA: Stanford University Press, 1998.

Index

Abanaki tradition 165 Abduction 102, 108, 114, 116, 180–2, 231, 232 Abelson, Robert P. 204 ˜ ana´sa¯ kuntalam 21, 105, 167, Abhijn¯ 173, 220, 234, 235 Abhinavagupta xi, 13, 39, 45, 46, 49–54, 57, 62, 66, 78, 262 Abraham (Ibr¯ah¯ım) 21 Abrams, M. H. 220 Absolutization 104, 106, 110, 170, 214, 224 Abstraction 26–8, 32, 37, 44, 51, 61, 96, 97 Abu¯ Lahab 116 Abu-Lughod, Lila 110 Abundance 191, 193–7, 226 Achill¯es 123, 130, 149, 216 Adam 15, 196, 210, 217, 231 Adaptation 263 Ad hoc categories 55, 59, 183 Aeneid 38, 123, 134 Aeschylus 104 Aesthetic experience 18, 23, 30, 36, 39, 41, 43, 47, 48, 51–3,

55, 68, 69, 80, 81, 93, 128, 132, 168, 261–3 Affection 70, 72, 73, 97, 105, 124, 166, 244, 254, 255, 258–62 Agency 23, 76, 77, 91, 186, 205, 210, 212, 214, 217, 219, 221, 227, 238, 254, 255, 258, 259 Aggression 85, 162, 260 Ahim . s¯a 137 Ahmad, Aijaz 16 Ainu epic 14, 15, 21, 28, 172, 175–83, 186–90, 196–201, 222, 225 Akam poems 161 Aladdin (‘Al¯a’ al-d¯ın) 25, 108 Alam . k¯ara 46 Algorithms 59, 63, 244 Al-H . asan 115 Al-H . usayn 115 ‘Al¯ı 115 Allegory 129, 147, 191 Alliteration 21, 22, 27, 28, 32, 33, 35–7, 46, 56, 158 Allusion 154 Amaladass, Anand 48

285

286

Index

Amin, Samir 16 Amnesty International 190 ¯ Anandavardhana xi, 45, 46, 48, 50–2, 54 Andromach¯e 124, 128 Anger 2, 47, 51, 63, 76, 77, 84, 85, 89, 90, 93, 95, 219, 241, 242, 248, 254–61, 263 Anorexia 200 Anscombe, G. E. M. 131 Antagonist 131, 146, 147, 177, 211, 216, 217, 232, 234, 237 ‘Antar 115, 116 Anubh¯ava 82 Anxiety 67, 95, 187 Appiah, Kwame Anthony 9, 16 Arabic tradition 24, 107, 108, 114–16, 129, 161, 162 ¯ Arambha 78 Arawak tradition 109 Arberry, A. J. 162 Archilochos 161 Architecture, cognitive 43, 58, 59, 61 Areal distinctness 17–19, 26, 44, 174, 176 ¯ es 191 Ar¯ Aristotle 13, 24, 77, 78, 89, 102, 132, 133, 144, 212, 218, 220 Arjuna 125 Arnhem Land 26 Artha 93–5, 97, 128 Artha´sa¯ stra 94, 137 Asceticism 200 Ashcroft, Bill 9 Assonance 21, 32–5, 37, 158 Aston, W. G. 168 Astyanax 94, 124, 194 Atonement 237 Atsumori 127, 128, 130, 134, 135, 138, 149, 237 Attention 36, 37, 39, 64, 65, 75

Atwood, Margaret 246 Austen, Jane 174, 179 Avoidance 255, 258 Awe 169, 213 Awolalu, J. Omosade 217 Axelrad, Elise 67, 68 Aztec tradition 191, 193 Babbling 261, 263 Babylonian tradition 21, 38 Badawi, M. M. 23 Baddeley, Alan 39 Baker, Sheridan 152 Bal, Mieke 205 Bambara tradition 113, 165 Bambi 259 Banality 23 Barnard, Philip J. 43 Barsalou, Lawrence W. 58, 59, 183 Bartlett, F. C. 67 ¯ Matsuo 160, 167, 168 Basho, Basicness 184 Basotho poetry 21, 26, 38 Bateson, Mary Catherine 33, 36 Beauty 48, 53, 160, 262, 263 Bedouin poetry 110, 234 Bekerian, D. A. 65, 66, 95 Bem, Daryl J. 240 Beowulf 118 Berggren, Paula 164 Berndt, Ronald M. 26, 38, 165 Bernheimer, Charles 8 Bhakti 129, 170, 174 Bharatamuni 24, 47, 77 Bh¯asa 25, 102, 105, 135, 157, 208 Bh¯ava 47, 76, 81, 82, 93 ¯ 25, 105, 126, 135, 156 Bhavabhuti Bible 155 Biebuyck, Daniel 138 Bierhorst, John 191–4 Bih¯ar¯ı 161

Index Bilhan.a 161 Blair, R. J. R. 143, 144 Blake, William 38 Blending see Conceptual blending Bordwell, David 7, 8 Boredom 96, 261, 262 Bororo 109 Boulton, Maureen 158 Bower, G. H. 63, 64 Bownas, Geoffrey 21, 164 Brandon, James R. 25, 102, 107, 113, 191 Brenan, Gerald 170 Brewer, William F. 7, 28, 29, 87, 203 Brewitt-Taylor, C. H. 21, 159 Broadway musicals 157 Brothers, Leslie 256 Brower, Robert H. 164 Brown, Donald 10 Bunraku 104 Businessman 121 By-products, evolutionary 263 Cadmus 191 Cao Xueqin 159 Capitalism 198, 249 Capital punishment 189 Care 136, 137 Carroll, Joseph 8 Carroll, No¨el 64, 65 Catharsis 189 Causality 5, 23, 33, 77, 91, 138, 153, 189, 205, 217–21, 224, 227–9, 238 efficient 218, 229 final 218–21, 229, 238 general 218, 219 restricted 219, 221 see also Design

287

Ceremonies 105, 123, 128, 165, 225, 236 Chadwick, Nora K. 116, 195 Chang Wen-ch’ien 163 Character 19, 24, 29, 47, 53, 64, 69, 74, 75, 87, 88, 92, 93, 109, 117, 119, 133, 144–6, 149–51, 157–60, 166, 169, 176, 179, 180, 182, 205–17, 220, 225, 228, 231–3, 237, 246 objectification of 145, 209 subjectivity in 145–9, 209 Character types 23, 24, 87 empathic 212–14, 216 individualistic 205, 207–17, 231–3, 238 socially normative 206–12, 215–17, 231, 237 socially representative 205, 210, 211, 232, 233 ¯ 25, 102, 108 Chariton Chaucer, Geoffrey 38 Chelkowski, Peter 115 Chˆeng Teh-hui 25, 102, 104 Chikamatsu Monzaemon 24, 94, 107, 113, 134, 208, 236 Chinese tradition 2, 12, 13, 21, 23, 25, 26, 33, 38, 42, 43, 98, 105, 106, 114, 116, 117, 154, 158, 159, 163, 169, 170, 191, 208, 211, 236 Chippewa 165 Chomsky, Noam 16, 22, 23, 28, 31, 45, 203, 204, 252, 264 Ch’u tradition 170 Cilappatik¯aram 111 Circularity see Plot circularity Cleopatra 123 Clinton, Jerome 116 Clore, Gerald L. 79, 227, 228

288

Index

Cohen, Joshua 249 Collins, Allan 227, 228 Colonialism 8, 9, 19, 23, 51, 118, 121 Color terms 98 Comedy 11, 13, 23–5, 78, 90, 98, 101–4, 107–9, 110, 112, 113, 116, 121, 133, 134, 155, 205, 207, 208, 217, 225, 227, 234, 235 Commodification 250, 251 Compassion 106, 124, 136–50, 199, 206 Comrie, Bernard 19, 28, 29, 98 Conceptual blending 7 Conchubar 181 Concretization 58, 66, 84, 86 Congolese poetry 126 Connectionism 58–60 Connotation 33, 50 Consciousness 13, 35, 36, 52, 53, 56, 57, 62, 81, 89, 141 Contemplation 262, 263 Conway, Martin A. 65, 66, 95 Cooper, Arthur 21, 33, 38, 163 Cosmides, Leda 4, 263 Crane, Mary 4 Croft, William 28, 29 Cronyn, George W. 165, 194 Crucifixion 226 Cuceb 191, 192 Cues and constraints 20, 43 Cullmann, Oscar 211 Culture 1, 2, 7–10, 14–16, 18–21, 23–6, 28, 29, 36, 37, 59, 79, 84, 98, 102, 104, 109, 110, 113, 124, 133, 170, 175–7, 183, 184, 188, 190, 191, 194, 212, 213, 219, 244–7, 249–52, 256, 257, 259, 260, 263, 264

Dalton, E. T. 235 Da Monzon de Segou 113, 114 Danger 16, 70, 111, 114, 116, 119, 139, 142, 177, 238, 246, 256 Danly, Robert Lyons 160 D¯ar¯ab 21 Davidson, Donald 3, 10 Davis, Keith E. 257 Davis, Mark H. 81, 140, 145, 146 Dawood, N. J. 25, 108 Dawson, Michael R. W. 63 Death 5, 24, 25, 52, 53, 62, 71–4, 83, 84, 86, 90, 91, 94–6, 101–11, 113–17, 119, 124, 126, 127, 130, 131, 138, 144, 149, 157, 161, 163, 165, 174, 177, 180, 182, 189–93, 196–9, 210, 211, 215, 218, 220, 223, 224, 228, 231–3, 236, 237 Decay 255, 258 Deconstruction 110 Dede Korkut, The Book of 116, 127, 145, 158, 225 Defaults 20, 22, 58, 60, 79, 86, 129, 140, 184 Defense 112, 115, 116, 137, 138, 147, 150, 199, 213 Dehumanization 177, 214 Deification 214, 253 Deirdre 181 Demonization 108, 110, 111, 114, 126, 177, 198, 212–17, 221 Deng, Francis Mading 38, 165, 166, 194 Design 218, 220, 221 De Sousa, Ronald 82, 83 Despair 73, 110, 119, 124, 125, 128, 137, 168, 170, 221–3, 225 Dhanam . jaya 77 Dharma 93, 128, 131, 226

Index Dhvani 24, 45, 46, 48–52, 55, 57, 62, 66, 71, 72, 74 see also Rasadhvani D¯ıdo¯ 123, 134 Dieting 200 Dimock, Edward C., Jr. 127, 156 Dinka poetry 38, 165, 166, 194 Discourse 7, 87, 100, 121, 177 Disgust 76, 77, 79, 81, 89, 90, 255, 256, 258, 260, 262, 263 Dispositional objects 253–4, 256, 258, 260, 262, 263 Dissanayake, Ellen 8 Dixon, R. M. W. 41 Dobosz, Barbara 247, 248 Dobyns, Stephen 153 Dolby, William 25 Domination 9, 96, 97, 111, 120, 122, 123, 128, 134, 141, 150, 151, 155, 156, 170, 178, 182, 184, 196, 222–5, 228, 259–61 Doniger, Wendy 8 Drama 2, 9, 19, 23–5, 47, 76, 77, 88, 102–7, 111, 113, 115, 127, 128, 133, 152, 156, 157, 162, 164, 167, 190, 191, 202, 211, 217, 220, 234 Draupad¯ı 125 Duckitt, J. H. 141 Dundas, Paul 200 Duryodhana 135, 208 Dus.yanta 105, 224 Duwell, Martin 41 Dyirbal tradition 41, 42 Eat, desire to 255–8, 262 Ecology 199 Eden 196 Edible object 255, 258 Edmonson, Munro S. 18 Edo kabuki 113

289

Egocentrism 53, 70, 71 Egudu, Romanus 21 Egyptian literature 162, 169 Ekman, Paul 76–80, 93, 256 Emotion actional outcome of 84, 85, 88, 96, 239, 240, 242, 249, 258–63 and ambient conditions 260–3 basic 53, 76–81, 89, 252, 254 categorization of 247–8 cause of 1, 46, 62, 69, 70, 82, 84, 85, 92, 124, 218, 241, 247, 248 clustering of 248 and communication 80–2, 89, 242 componential approach 253 eliciting conditions for 82–5, 88–90, 92–7, 102, 103, 110, 118, 129, 131–3, 147, 187, 215, 221–3, 239–51, 254–63 expressive outcomes of 16, 76, 77, 79, 80, 82, 83, 88, 91, 96, 240, 241, 243, 247, 249, 251, 252, 256, 257 factual versus substantial 131–2 labeling of 240–2 malleability of 241–52, 263 object of 69–71, 78, 226, 227, 231, 235, 237, 238, 248, 253–62 phenomenological tone of 81, 82, 85, 87, 89, 90, 109, 239, 240, 242, 243, 247–9, 251, 263 source of 46, 62, 67, 69, 70, 75, 189, 242

290

Index

Emotion prototypes 15, 65, 84–98, 133, 171, 202, 204, 228, 238, 239, 242, 250–2, 260 personal 94–8, 102, 103, 121, 122, 129, 182–6, 198, 199, 225, 226, 228, 244, 258 physical 183, 186, 195, 200, 226, 228, 258 social 94–8, 101, 104, 110–13, 121–5, 128–34, 139, 141, 150, 151, 166, 178, 182–6, 196, 226, 228, 230, 246, 259 transcendental 125–31 Emotion terms 2, 79, 82, 92, 95, 96, 256 Emotion types junctural 91–3, 102, 103, 169, 246, 255, 264 outcome 91–4, 102, 103, 118, 129, 153, 223, 224, 243, 247, 255 sustaining 93, 94, 102, 118, 136, 186, 221, 243, 246, 255, 264 Empathy 54, 67, 70, 71, 75, 80–2, 88–91, 96, 122, 123, 139–50, 177, 205, 206, 208, 209, 212–17, 226, 251, 252, 262 categorial 140, 141, 148, 150, 208 situational 140–50, 205–9, 212–16 see also Identification Emplotment 5, 116, 117, 153 Encoding 35–9 Epic 6, 14, 87, 88, 94, 111–16, 120–3, 126, 127, 130, 138, 152, 156, 162, 172, 173, 175–81, 186, 192, 200–2, 215, 224, 236

Epilogue of suffering 123, 125–9, 133, 137, 138, 142, 145, 148–51, 155, 163, 177, 179, 181, 190, 193, 208, 216, 222, 225, 227, 228, 233, 237, 251 Epithets 28, 43, 162 ¯ Erotici Graeci 25 ` . u` 217 Es Ethical conflict 133–9, 150, 151, 198–200 Ethics 12, 85, 93, 122, 123, 126–42, 145–7, 150, 151, 187, 189, 198–200, 210–13, 215, 220, 227, 252 of care 136, 137 of compassion 136–47, 199 of defense 136–47, 199, 213 of justice 136, 137 of self-abnegation 199, 200 Ethnocentrism 257 Eucharist 226, 236 Eudaimonia 222–4, 244 Eulogy 34, 158, 162, 165, 166, 169, 170 Euripid¯es 94, 128, 148, 190, 198 Eurocentrism 9 Eve 15, 196, 210, 217, 231 Evil 113, 116, 144, 151, 170, 177, 182, 210, 212 Evolution xi, 246, 249, 252, 253, 259, 263 Exempla 57, 59, 60–2, 85, 133, 184, 202, 245 Exemplars 59–61, 83, 85 Exile 11, 15, 94, 95, 101, 104–20, 125, 128, 143, 154, 155, 157, 160, 163–5, 168, 170, 173, 177, 178, 180, 181, 189, 190, 192, 193, 196–8, 228–37

Index Explanatory adequacy 31, 32, 43, 44 Facial expression 76–80, 91, 252, 256 ` o` 94, 215 Fa-Digi Sisok Fall, the 21, 196 Familializing conflict 148, 149, 227 Famine 181, 185–200, 215, 219, 225, 228 Fanon, Frantz 16 Fasting 200 Fear 63–5, 70, 73, 76, 79, 81, 89, 90, 93, 95, 96, 115, 120, 139, 142–4, 218, 229, 240, 243, 246, 249, 254–60, 263 Ferdowsi, Abolqasem (Firdausi) 21, 114, 207, 208 Feudal societies 29, 113, 198, 208 Firth, Raymond 21, 26, 165, 194 Fischer, E. F. 98 Flaubert, Gustave 245 Flight 218, 246, 249, 254, 255, 258 Fodor, Jerry A. 60 Food 86, 166, 182, 187–200, 215, 225–7, 231–3, 244, 245, 256, 258, 260, 261, 263 Foreshadowing 21, 33, 34, 35, 37 Forgas, Joseph Paul 63, 64, 247, 248 Formulae 28, 43 Fowler, Barbara Hughes 161 Freedman, J. L. 95 Freud, Sigmund 1 Friendship 247, 248 Fright 52, 73, 76, 77, 79, 189, 246 Frijda, Nico H. 103, 223 Frustration 170, 248, 254, 257 Frye, Northrop 13, 24, 134, 152, 205, 207 Fuhrman, Robert W. 61

291

Funeral 86, 123, 128, 165, 195 Fung, Heidi 67 Fusion of horizons 140, 142, 144 of narrative structures 228 Gadamer, Hans-Georg 140, 142, 144 Gao E. 159 Gardner, Howard 4 Gardner, John 138 Garman, Michael 39, 56, 57 Gathercole, Susan 39, 42, 43 Gelpke, Rudolf 108 Gender 137, 176, 179, 180, 182, 251 Gendler, Tamar 143 Generalization, social 225, 226, 231, 234, 237, 238 Genetic distinctness 17–19, 26, 32, 44, 122, 174–6 Genji 112, 160 Genre 2, 13, 15, 19, 21, 23, 59, 65, 75, 87, 104, 107, 109, 115, 118–20, 134, 136, 147, 151, 153, 154, 161, 162, 166, 168–70, 176, 185, 199, 207–9, 230, 231, 259 Gergen, Kenneth J. 257 Gerow, Edwin 46, 156 Gerstle, C. Andrew 104 Ghazal 129, 134, 162, 170, 217 Gibbs, Raymond W. 33 Gifford, Don 120 Gilgamesh 126, 130, 131, 138, 149, 162, 224, 237 Gilligan, Carol 136, 137 Girard, Ren´e 189, 190 Gnoli, Raniero 39 Goals 65, 76, 77, 88, 89, 91–4, 102, 103, 111, 113, 116, 122, 125, 129–33, 140, 178,

292

Index

Goals (cont.) 182, 185, 189, 203–6, 214, 217–24, 227, 228, 244–6, 249, 254, 255, 257–9, 262 short-term 222–4, 243 transcendental 182, 224 Goethe, Johann Wolfgang von 104 Gould, Stephen Jay 252 Gratitude 162, 254, 258 Greimas, A. J. 214, 226 Grice, Paul 48 Grief 71, 95, 108, 124, 125, 149 Griffiths, Gareth 9 ¨ Grunbaum, Adolf 220 Guépin, J. P. 191 Guilt 128, 132, 149, 157, 192, 210, 216 Guinean tradition 165 Gurg¯an¯ı, Fakhr ud-Din 115, 157, 158, 235 Habituation 22, 186, 223, 224 Hadas, Moses 148 Haiku 40–2, 160 Halász, Lászlo´ 66, 67, 143 Halford, Aubrey S. 191, 217, 235 Halford, Giovanna M. 191, 217, 235 Hamartia 77, 132, 133, 138, 144 Hamlet 46, 69–74, 118, 120, 179 Hampton, James A. 59 Happiness 11, 15, 47, 52, 63, 65, 76, 90–8, 101–3, 105, 106, 108, 110, 111, 121, 122, 125, 127–33, 137, 141, 150, 151, 166, 182–7, 195, 200, 221–30, 234, 238, 243, 244, 246, 250, 254, 255, 257–9, 262, 264 Hardman, Charlotte 187 Harré, R. 257

Harris, Paul 187 Hars.adeva 105 Hart, George 28, 161 Hate 34, 79, 95, 254, 258 Hauser, Arnold 29 Hawai’ian tradition 38 Hawkes, David 170 Healy, Alice 43 Heath, Peter 116 Hebrew tradition 21, 169 Hector 123, 124, 128, 130, 149, 216 Hecuba 94 Hegarty, Mary 43 Heike, Tale of the 112, 113, 124, 125, 127, 128, 130, 134, 135, 147, 149 Herman, David 204 Hernadi, Paul 4, 13, 116 Hijrah 116 Hinduism 51, 129, 137, 141, 170, 200, 224 Historiography 116, 117, 195 Hitomaro 164 Hitopade´sa 94 Hittite tradition 138 Hjort, Mette 79 Hobbs, Jerry 4 Holland, John 35, 83, 145, 146 Holland, Norman xii, 4, 66 Hollywood musicals 157 Holyoak, Keith 35 Home 95, 108, 111, 113, 119, 120, 127, 129, 154–7, 163, 167, 177, 178, 180, 181, 198, 229–31, 235, 237, 245, 249, 250 Homer 124 hom*osexuality 20 Hoogstra, Lisa 67 Hopi tradition 94 Horror films 65, 261

Index Hsia, C. T. 105, 106, 116 Humanization 150, 177, 182, 206, 226, 251 Humayuni, Sadeq 115 Humbaba 126, 138, 149 Hunger 166, 186, 187, 191–3, 196, 225, 230, 255–9 Huot, Sylvia 153 Iblis 114 Ibn Ish¯aq 116 Idealization 146, 214, 225–7, 238 Identitification categorial 141, 144, 145, 148, 206, 207, 210, 212, 230, 259 situational 144, 206, 212, 214, 216 see also Empathy Ideology 8, 9, 30, 175, 189, 230, 239, 250–2, 257 Ijo tradition 113 Ilangoˆ Adigal 111 Iliad 123, 124, 128, 135, 181, 216 Imagery, 19, 21, 25–7, 34, 36, 95, 101–5, 107, 108, 110, 119, 129, 131, 152, 155, 167–9, 180, 184, 189, 197, 214, 228, 229, 231–3, 236, 237, 261 Imagination 1, 15, 66, 69, 73, 74, 82, 91, 92, 96, 104, 105, 113, 136, 140, 144, 147, 149, 153, 166, 169, 182–4, 221, 224, 225, 227, 246, 247, 250, 252, 261, 263 Immanent norms 211, 212, 215 Implicature 48 Indian cinema 157 Indian tradition 9, 10, 21, 23, 25, 26, 39, 45, 46, 55, 76–8, 90, 111, 113, 123, 137, 161, 165, 195, 198, 208

293

Influence, literary 17–19 Ingalls, Daniel H. H. 78 Ingarden, Roman 66 In-group 97, 111, 121, 141, 144, 170, 176, 221, 224, 259–61 Inhibition 254, 258 Innatism 31 Intensification 23, 86, 88, 103, 153, 160, 223, 224, 227–30, 238, 248 Intentional object 69, 70, 75 Intertwining of narrative structures 228 Iranian tradition 114, 115, 207 Irele, F. Abiola 162 Ise, Tale of 160 Iser, Wolfgang 88 Ishmael (Ism¯a’¯ıl) 21 Islam 116, 170, 224 Jackendoff, Ray 203 Jainism 200 Jakobson, Roman 7, 32, 34 Jankowiak, W. R. 98 Japanese tradition 12, 21, 23, 25, 26, 40, 41, 103, 104, 106, 107, 112, 113, 121, 123, 124, 157, 160, 164, 167, 175, 191, 235 Jauss, Hans Robert 88 Jayadeva 156 Jealousy 79, 95, 114, 189, 240, 249 Jenkins, Lyle 252 Jerusalem Delivered 118 Jesus 15, 196, 197, 210, 211, 225, 226, 228, 235, 236 John of the Cross, Saint 170 Johnson, John William 127 Johnson, Mark 26 Johnson-Laird, Philip N. 76–80, 83, 89, 93 Joseph, George 246

294

Index

Joyce, James 67, 120, 121 Judaism 170 Judeo-Christian tradition 21, 196, 199, 217, 231 Juncture 12, 78, 91–3, 102, 103, 152–71, 223, 243, 246, 255, 264 Kabuki 25, 103, 107, 113, 133, 191, 217 Kafka, Franz 198 Kahneman, Daniel 58, 97 Kailasapathy, K. 28 Kalhan.a 195 K¯alid¯asa 21, 56, 105, 167, 220, 234 Kall¯ınos 161 K¯ama 93, 94, 128 Kan’ami 25, 106 Karn.a 130, 149, 208 Kaut.ilya 137 Kazin, Alfred 38 Keene, Donald 24, 41, 106, 113, 157, 160, 164 King, Bruce 10 King Lear 11, 118 Ki no Tsurayuki 164 Kintsch, Walter 43 Kiparsky, Paul 3, 7, 18, 21, 22, 27, 32, 34–6, 45 Kirson, Donald 66, 96 Klein, Stanley 61 Klemenz-Belgardt, Edith 29 Knight, W. F. Jackson 123 Knox, Bernard 190 Kokinshu¯ 157, 160, 164 Kondh poetry 195, 235 Korn, Alfons L. 38 Kouyat´e, Djeli Mamoudou 127, 226, 228 ¨ Kovecses, Zolt´an 10, 84–6 Ks.atriyadharma 137 Kuan Han-ch’ing 191, 211

Kuna tradition 21, 41, 42 Kunene, Daniel P. 21, 26, 38 Kuo yu¨ 117 Kurelu tradition 165 Kyoto kabuki 113 Labouvie-Vief, Gisela 1 Labov, William 91, 229, 230 Lakatos, Imre 11, 12, 14, 172, 173, 179 Lakoff, George 26, 59–61, 84–6, 184 Laks.m¯ı 130 Lal, P. 25 LaMar, Virginia A. 74 Larsen, Steen 66, 143 ´ János 66, 143 Lászlo, Laughlin, Philip xii Laver, Sue 29 Lawrence, D. H. 120 Layl¯a 24, 104, 107, 108, 147, 217, 220, 224, 234 LeDoux, Joseph 243, 256 Lerdahl, Fred 203 Lesbianism 20 Levinson, Jerrold 69 Lévi-Strauss, Claude 109 Levy, Reuben 162 Lewis, Geoffrey 158 Lexical context 131 Lexical entry 22, 55–7, 60, 61, 81–3, 95, 118, 129, 140, 141 Lexical identification 89 Lexical item 43, 55, 57, 239, 256 Lexical links 63 Lexical networks 57, 63, 103 Lexical processes access 22, 54–7, 61, 68, 81, 82, 103 activation 52, 53, 57, 59, 63–5, 67, 141–3, 147

Index cross-indexing 22, 55, 56 priming 56, 57, 61–5, 67–9, 71–5, 89, 104, 121, 141–3, 206, 227, 235, 237, 246 reordering 56 Lexical structure 54, 61, 140, 183, 239, 240 Lexicon 22, 46, 55, 56, 60, 87, 98, 103, 129 Li Ch’ing-chao 163 Lichtenstadter, Ilse 115, 161 Lichtenstein, Edward H. 7, 87 Lincoln, Kenneth 194 Ling Ching 163 Li Po (Li Bo) 163, 167 Literary universals website 8 Liu, James J. Y. 163, 164 Liu Hsieh 21 Liu Jung-en 21, 25 Locke, John L. 261 Loftus, Judith 61 Lord, Albert 28, 40 Love 11, 20, 23–6, 34, 47, 48, 53, 62–7, 71, 73, 79, 83, 84, 86, 87, 90, 93–8, 101–10, 115, 119, 120, 124, 129, 130, 133, 134, 136, 155, 156, 158–70, 173, 179, 180, 181, 182, 200, 205, 206, 208, 209, 217, 220, 222, 225, 227, 228, 230–2, 234–6, 240, 243, 244, 245, 247–51, 255, 259, 261 triangle 179–82, 200 Loyalty 113, 117, 199, 208, 216 Luk´acs, Georg 146 Luo Guanzhong 159 Lutz, Catherine A. 110, 251 Lyric poetry 2, 5, 12, 152–71, 174, 194, 195, 226 devotional 169–71, 174

295

Macbeth 118, 132 ¨ 104 Ma Chih-yuan Mack, Maynard 154, 163, 169 Mah¯abh¯arata 111, 125, 130, 135, 149, 208, 237 Mah¯ad¯eviyakka 174 Mah¯ak¯avya 156 Mahapatra, Sitakant 195, 235 Maier, John 138 ¯ 24, 104, 107, 108, 147, Majnun 220, 224, 234 Malian tradition 94, 111, 127, 215 Mapping 66, 141, 142, 144 Manohra 25, 102 Marcus, Ruth Barcan 151 Markandaya, Kamala 198, 236 Marriage 86, 95, 102, 105, 107–10, 115, 164, 180–2, 191, 209, 222, 223, 225, 226, 230, 232, 234, 236, 244, 247, 250, 251, 259 Marthiya 161 Martinovitch, Nicholas N. 23 Martyrdom 136, 200 Marx, Karl 9, 16, 29, 208 Massey, Reginald 77 Masson, Jeffrey 161 Mathews, A. 64 Maximal relevance see Patterning Maximization of patterning see Patterning M¯ay¯a 51, 192 Mayan tradition 191 McClelland, J. 58 McLean, Mervyn 21, 26, 165, 194 McLeod, Peter 58 Memory 3, 20, 29, 35, 39–44, 51–7, 62–75, 82, 89, 91, 141–5, 206, 214, 229, 243, 246 buffers 57, 62

296

Index

Memory (cont.) long term 54, 56, 57, 62 personal 29, 54, 55, 65, 66, 68, 71, 73, 74, 89, 141, 143, 206, 214, 230 perspectival center in 141, 142, 144 rehearsal 39–43, 56, 57, 62, 243 working 39, 43, 44 Meriah 235 Merwin, W. S. 161 Meta-ethical issues 136 Micronarratives 65, 88, 118, 243 Military Romance, Chinese 116, 117 Miller, Barbara Stoler 21 Miller, Dale T. 58, 97 Miller, George 36 Miller, Peggy J. 67, 68 Milton, John 196, 202, 208, 217 Mind objectal view/reification of 145, 146, 151, 209 qualia in 240 subjective view of 145–9, 209, 212, 240 Miner, Earl 164 Mintz, Judith 67 M¯ır¯ab¯a¯ı 174 Mirth 2, 76, 90, 249 Miyake, Akira 43 Modeling of narrative structures 21, 120, 123 Models 14, 21, 35, 55, 59–61, 84, 87, 120, 123, 129, 204, 224, 225, 252 Modernism 2, 118, 120, 121, 198 Moks.a 54, 93, 128, 224 Monier-Williams, Monier 78 Monogatari 160 Mood 59, 63, 64, 66, 181

Moral conflict 133–9, 145, 147, 149–51, 199, 227, 252 Moreh, Shmuel 23 Morrison, George 115, 157 ¯ a) 21 Moses (Mus¯ Muh.ammad 116 Murnau, F. W. 210 Murphy, Roland E. 155 Muslim tradition 21, 115, 116, 162, 170, 200, 217 Mwindo Epic 112, 126, 130, 138, 139, 148, 224, 236, 237 Nahuatl tradition 191 Naim, C. M. 150 ¯ Namiki Sosuke 135 Naoise 181 N¯arada 134 Narayan, R. K 127 N¯ar¯ayan.a 94 Narrative Hypothesis 152–7, 160, 166, 169, 171, 194 Narratology 203 Narukami and the God Fudo 191 Nas¯ıb 162 ´ astra 24, 47, 76, 77 N¯a.tya S¯ Navaho tradition 194 Necessary and sufficient conditions 30, 42, 82–5, 87, 89, 92, 98, 102, 133, 136, 166, 186, 204 Nelson, Katherine 204, 205 Nepal 187 New Comedy 13, 24, 108, 133, 134, 205, 207, 208, 217, 234 ˜ ı wa Thiong’o 9, 10, 16 Ngug˜ Niebelungenlied 118 Night Chant 194, 195 Nisbett, Richard E. 35, 61, 241, 242 Niyatapti 78 Niz.a¯ m¯ı 24, 104, 108, 147, 217, 220 N¯o 25, 104, 106, 127, 157

Index Nondispositional objects 253, 255, 258, 260, 262, 263 Nontechnical correlations 20, 25, 44 Nosferatu 210 Nouveau roman 118, 120 Nwoga, Donatus 21 Nyanga tradition 112, 126, 130 Oatley, Keith xii, 4, 46, 51, 63, 66, 67, 71, 76–83, 89, 90, 93–5, 204, 243, 259 Obedience 199 Obligatory techniques 21, 22 Obtrusiveness 36, 37, 40 O’Connor, Cary 66, 96 O’Connor-Visser, E. A. M. E. 191 Odyssey 38, 120, 236 Oedipus Complex 181 Oedipus plays 104, 128, 190, 211, 228, 237 Ohnuki-Tierney, Emiko 178 Oiemono plays 113 ¯ Okabou Ojobolo 113 Olthof, Tjeert 187 Omens 21 Ong, Walter J., S.J. 28, 40, 262 Opland, Jeff 28 Optional techniques 21, 22 Orality 12, 14, 20, 26, 28, 40, 43, 98, 162, 174–6, 194 Orature 28, 175, 188 Oresteia 104, 128 Orians, Gordon 262 Ortony, Andrew 58, 79, 227, 228 Ostracism 228 Out-group 141, 144 Owen, Stephen 170 Ozidi Saga 113 Padgett, Ron 153 Pagels, Elaine 211 Pandit, B. N. 51

297

Pandit, Lalita 9, 73 P¯an.ini 45 Panksepp, Jaak 253, 254, 256, 259, 260 Paradigms 5, 6, 20, 82, 94, 95, 103, 115, 118, 120, 122, 127, 135, 146, 148, 154, 160, 191, 211, 213, 216, 238, 244, 245 Parallelism 21, 32, 123 Parrott, W. G. 257 Particularism 8, 9, 10, 16, 110 Pathos 47, 50, 72, 124, 164, 215 Patriarchy 9, 249–51 Patterning, maximization of 33–40, 43, 44, 63 see also Relevance Pavel, Thomas G. 203, 204 Penelope 120 Pennington, Bruce 43 Perkins, George 152 Persian tradition 24, 107, 108, 114, 115, 129, 157, 162, 217 Personalization 67, 68, 97, 209, 212, 219, 221, 228, 229, 238, 262 Philippi, Donald L. 21, 28, 175–8 Pindar 161 Pinker, Steven 4, 8, 254, 262 Pity 144, 182 Plague, The 199 Plantinga, Carl 8 Plot circularity 21, 33, 35, 37, 218 Plunkett, Kim 58 Poetic line 20, 37–44 Poetics 45, 46, 48, 77, 79, 168 Poetry 3, 5, 6, 12, 20–3, 26–8, 32–42, 45, 46, 48, 110, 118, 120, 121, 127, 147, 152–71, 174–82, 185, 188, 194, 195, 202, 218, 222, 225, 226 Point of view 112, 139, 140, 176 Polynesian tradition 26, 165

298

Popper, Karl 11 Postmodernism 2, 99, 120, 121 Praise poem see Eulogy Prefocus 64, 65 Priam 123, 124, 130, 149 Priming see Lexical processes Prince, Gerald 203 Probes 142, 144 Project 223–5 Protection 254, 258 Proto-emotions 253–64 Prototypes 6, 11, 12, 15, 30, 42, 57–62, 65, 66, 80–98, 100–5, 110, 111, 113, 114, 116–22, 128–33, 136, 137, 139–41, 147, 150–5, 160, 166, 169–71, 178, 181–6, 188, 191, 194–6, 198–200, 202–6, 208, 209, 212, 214, 215, 218, 219, 221–4, 226–31, 236–40, 242–8, 250–2, 255, 257–61, 264 Psalms, Book of 155, 169 Pseudouniversals 9 Psychoanalysis 1, 109 Puenzo, Luis 199 Pukui, Mary Kawena 38 Puram poems 161 Pursuits 222–4, 254, 258 Qas.¯ıdah 161, 162 Quetzalcoatl 191–3 Quint, David 123 Racine, Jean 38 Racism 9, 16, 234 R¯ama 111, 126, 130, 132–5, 139, 140, 156, 157, 180, 208, 224–6, 236, 237 Ramanujan, A. K. 156

Index R¯am¯ayan.a 21, 24, 102, 104, 111, 126, 127, 130, 132, 134, 135, 162, 170, 180, 208, 226, 228 Rasa 46, 47, 50, 51, 53, 54, 62, 65, 67, 74, 79, 81 Rasadhvani 50–2, 57, 62–4, 71, 74 Rati 93, 94 R¯av¯an.a 111, 126, 170, 236 Realism 29 Reception of literature 39, 74, 92, 185 Relevance 33–5, 37, 39, 40, 62, 63 see also Patterning Religion 30, 54, 105, 120, 128, 129, 131, 134, 136, 141, 144, 152, 158, 162, 169, 170, 171, 179, 186, 193, 197, 198, 207, 208, 224, 234, 262 Remindings 91, 143 Remorse 125, 128, 132, 138, 149, 150, 179, 193, 213, 214, 222, 232, 233 Reparation 131, 138, 145, 150, 189, 191, 193, 199 Repentance 181 Research programs 7, 8, 10–16, 30–2, 37, 41–3, 44, 54, 79, 95, 122, 152, 172, 173, 185, 200, 201, 238 Resurrection 24, 102, 197, 228, 232 Revenge 128, 177–9 Rhyme 21, 22, 33, 36, 43, 56–8, 163 Richardson, Alan 4, 8 Richman, Paula 111, 135 Ritual 23, 86, 128, 181, 187–9, 192–7, 232, 235–7 Rival 101, 107–9, 119, 120, 178–82, 191, 216, 217, 231, 232, 234, 235, 237

Index Roadarmel, Gordon 156 Robbe-Grillet, Alain 99, 120 Rogers, Joel 249 Rolls, Edmund T. 58 Romanticism 2, 208 Romantic union 93, 94, 103, 106, 108, 121, 122, 128, 129, 131, 134, 157, 182, 222–5, 244, 245, 250 Romeo and Juliet 71, 103, 104, 108, 109 Ross, Lee 61, 241, 242 Rostám 21, 114, 115, 148, 207, 216, 235 Rubin, David C. 20, 40, 41, 43 Rumelhart, David E. 58, 203, 204 Rureke, Shé-kárisi 237 Russell, James 85, 86 Russell, Ralph 134, 162 Sabi 168 S¯adh¯aran.adharma 137 Said, Edward 16 ´ Saktibhadra 111 Sakuemon 107 ´ Sakuntal¯ a 105, 156, 167, 224 Sallekhana 200 Salthouse, Timothy 43 Samurai 107, 121 Sandars, N. K. 21, 38, 138, 149 Sanskrit tradition 2, 13, 21, 24, 28, 45–9, 64, 71, 76–83, 89, 90, 93, 94, 102, 105, 111, 128, 129, 131, 156, 161, 167, 204, 208, 225, 228, 256, 262 ´ antarasa 54 S¯ Sappho¯ 161 Sarbin, Theodore R. 153 Sartre, Jean-Paul 200 Sarup, Lakshman 25, 102 Satan 114, 208, 217, 228, 231, 235

299

Satanism 208, 213, 217 Sat¯ı 195 Saver, Jeffrey L. 4 Savoring 39, 53 Scapegoat 190, 195, 210, 211, 235, 251 Schachter, Stanley 241, 242 Schacter, Daniel L. 62, 142, 144 Schank, Roger C. 86, 91, 204 Schemas 21–3, 25, 28, 35, 36, 54, 57, 58, 60–2, 66, 84–6, 129, 133, 142, 184 Schwartz, Judith 66, 96 Scripts 84, 85, 153, 204 Secondary principles 32, 37, 44 Seidman, Robert J. 120 Seilman, Uffe 66, 143 Self-concept 29, 140, 141 Self-denial 199 Self-perception theory 240 Senex iratus 205 Sexuality 29, 30, 93, 109, 140, 141, 160, 161, 165, 174, 182, 192, 196, 222, 223, 244, 248, 249, 255–9, 262 Shah, Priti 43 Shˆahnˆame 114, 117, 134, 207, 216 Shakespeare, William 7, 11, 25, 71, 72, 74, 102, 108, 109, 118, 134, 225 Shanks, David R. 61 Shaver, Phillip 66, 80, 96, 97 Sherzer, Joel 21, 28, 41 Sh¯ı’ah 115 Shibatani, Masayoshi 175 Shilappadikaram (Cilappatik¯aram) 111 Shirane, Haruo 167, 168 ¯ u¯ 157 Shuish Sia tradition 194 Sidney, Philip 156 Sindb¯ad 25, 108

300

Index

Singha, Rina 77 Sioux tradition 194 S¯ırat ‘Antar 115, 116 Sissoko Kabine 114 S¯ıt¯a 104, 126, 130, 133–5, 156, 180, 225, 236 Sivan, Santosh 198 Social constructionism 257 Sohr´ab 21, 114, 115, 134, 148, 216, 235 Sommers, Christina 149 Song of Roland 118 Son-Jara, Epic of 94, 111, 127, 162, 173, 215, 226, 228 Sophocl¯es 104, 128, 190 Sorrow 2, 47, 52, 53, 62, 71, 72, 74, 76, 83, 86–98, 102–6, 110, 115, 121, 123, 124, 133, 149, 156, 157, 163, 186, 200, 222, 228, 234, 238, 254, 255 ¯ Sosuke 135 Soyinka, Wole 10, 198 Spatiality 228–30 Spenser, Edmund 156 Sperber, D. 33, 34 Spiritualization 106, 108, 180, 224, 225, 227, 230, 237, 238 Spolsky, Ellen 4 Stereotypes 60, 77, 137, 159 Stimulation 260, 261 Story grammars 203, 204 Story of the Stone 106, 159 Stratagem of Interlocking Rings, A 236 Subvocalization 42, 43 ¯ S´ udra 135, 139, 140 ¯ S´ udraka 105 Suffering 53, 60, 70–2, 105, 106, 108, 117, 123–8, 133, 137, 138, 142, 145, 147–51, 155, 163, 177, 179, 181, 189, 190, 193, 197–200, 208,

211, 213, 214, 216, 217, 222, 225, 227, 228, 230, 231, 233, 235, 237, 251 Sufis 170 Suggestion see Dhvani, Rasadhvani Suicide 24, 102, 103, 106, 107, 123–6, 130, 237 Sundiata see Son-Jara Surprise 76, 77, 263 Sympathy 109, 123, 134, 135, 147, 211, 212 Ta’ziyeh 115 Tamil poetry 28, 111, 161 Tan, Ed S. 7 Tasso, Torquato 118 Techniques, repertoire of 20–2, 25–7, 32, 35, 37, 44, 148, 158 Telemachus 120 Telic structure 204, 205, 227, 229 Tellability 88 Telos 224, 225 Tempest, The 118 Temporality 228–30 Terence 134, 234 Terwogt, Mark 187 Thagard, Paul 35 Thai tradition 25, 102 Thousand and One Nights, Tales from the 108 Threat 24, 67, 104, 108, 110–20, 136, 147, 173, 177, 178, 181, 185, 213, 231, 236, 254, 255, 257, 258, 260 Threat/defense sequence 111, 112, 114, 177, 178, 231 Three Kingdoms 117, 158, 159 Thresholds, attentional 37 Tiffin, Helen 9 Tikopia tradition 21, 26, 165, 194

Index Tiresias 191 Titanic 119 Todorov, Tzvetan 13 Tomkins, S. S. 243 Tooby, John 4, 263 Tosa Diary 160 Tradition 6, 9, 11–26, 32, 33, 36–9, 42, 44, 46, 50, 77, 78, 98–103, 105, 107–9, 118, 122, 127, 134, 154, 156, 158, 161, 165, 168, 170, 174–7, 179, 181, 185, 211, 217, 245, 251 Trafton, J. Gregory 61 Tragedy 11, 24, 91, 103, 104, 109, 112, 115, 119, 121, 163, 211, 216 Tragi-comedy 11–15, 19, 20, 24, 65, 95, 98–105, 107–10, 112, 113, 116–24, 129, 131, 132, 134, 135, 137, 145, 147, 148, 150, 151, 153, 154, 161–3, 166–70, 173, 174, 176, 177, 179, 181, 182, 185, 186, 189, 190, 198–200, 204, 206–8, 212, 224, 226, 228, 232, 233, 236, 251, 259, 264 heroic 11–15, 65, 98–104, 109–24, 128–40, 145–57, 161–6, 169–73, 176–82, 185–92, 195, 199, 200, 202, 206–16, 221, 224–31, 234–7, 251, 259, 264 romantic 11–20, 24, 65, 87, 95–122, 129–36, 153–74, 178–82, 185–9, 191, 195, 198–200, 205–9, 215, 220, 222–36, 245, 251, 258, 259, 264 sacrificial 15, 154, 181, 185–201, 210, 211, 215, 219, 220, 225–37, 259, 264

301

Transcendentalization 224, 225 Trask, Willard R. 165 Trauma of heroism 131, 135, 137–9, 142, 144, 145, 148, 151 Trobriand Islands 165 Troubadours, French 153 Tu Fu (Du Fu) 163 Turiel, E. 143 Turkish tradition 116, 127, 225 Turner, Frederick 40, 42 Turner, Mark 3, 4, 7 Tutuola, Amos 188, 198 Tversky, Amos 58 Tyrtaios 161 Tz’u 2, 164 Universalism 7–10, 16, 78 Universals 2, 4, 6–23, 25–32, 37–45, 71, 76, 77, 79, 80, 91, 95, 98–101, 110, 118, 121, 122, 129, 136, 137, 144, 152, 153, 172–7, 182–6, 190, 198, 200, 204, 244, 245, 249–53, 256, 257, 259, 260, 263, 264 absolute 19, 20, 22, 26–8, 30, 43, 100, 153, 186 biological 252–63 cultural 9, 10, 15, 250–2, 258–60, 263 historical 28, 29 implicational 4, 27, 28, 44 indexical 29, 30, 176 statistical 19, 20, 27, 32, 38, 44, 100 typological 28, 43, 44 Usurpation 11, 110–21, 138, 177–9, 182, 191, 192, 199, 208, 209, 227, 230–2, 234, 236 Uts¯aha 93, 94

302

Index

V¯alm¯ıki 126, 132, 134, 135, 162, 226 van Buitenen, J. A. B. 156 Varley, H. Paul 127, 147 Ved¯antism 51, 224 Vibh¯ava 82 Victim perspective 128 Victor perspective 128 Villainy 133, 135, 144–7, 209 Virgil 123, 202 Vis.n.u 111, 130 Voight, Ellen Bryant 153 Vy¯asa, Krishna-Dwaipayana 125, 149 Walcott, Derek 10, 49, 50 Waley, Arthur 25, 106, 164 Wang Wei 155 Wason, P. C. 58 Watanabe, Hitoshi 178 Watson, Burton 117 White, Hayden 13, 116 Wierzbicka, Anna 82, 83, 86, 92, 256, 257 Williams, Kimberly 67 Williams, Mark 121

Wilson, D. 33, 34 Wilson, Edward O. 4 Winter’s Tale, The 25, 102 Wittgenstein, Ludwig 82, 240 Wonder 15, 77, 90, 141, 262, 263 Woodbury, Anthony 28 Woolner, A. C. 25, 102 Wright, Louis B. 74 Xhosa tradition 28 Yarshater, Ehsan 114 Yeats, William Butler 197, 198 Yirkalla tradition 26, 38, 165 Yoruba tradition 50, 188, 190, 196, 197, 210, 217 Young, Kay 4 ¨ drama 104, 211, 236 Yuan Yuasa, Nobuyuki 167 Yudhis.t.hira 125, 237 Zeami Motokiyo 24, 106, 107, 127, 157, 237 Zhang Longxi 10, 159 Zhirmunsky, Victor 116, 195 ˜ tradition 194 Zuni

Other books in the series (continued from front of book) Metaphor and Emotion: Language, Culture, and Body in Human Feeling ¨ Zolt´an Kovecses Feeling and Thinking: The Role of Affect in Social Cognition Joseph P. Forgas The Causes and Consequences of Feelings Leonard Berkowitz Emotions and Beliefs: How Feelings Inﬂuence Thoughts Nico H. Frijda, Antony S. R. Manstead, and Sacha Bem Identity and Emotion: Development Through Self-Organization Harke A. Bosma and E. Saskia Kunnen Speaking from the Heart: Gender and the Social Meaning of Emotion Stephanie A. Shields The Hidden Genius of Emotion: Lifespan Transformation of Personality Carol Magai and Jeannette Haviland-Jones

Colm Hogan The Mind and Its Stories PDF - PDFCOFFEE.COM (2024)