Narratology Beyond Literary Criticism: Mediality and Disciplinarity (Narratologia - Volume 6)

Narratology beyond Literary Criticism Narratologia Contributions to Narrative Theory/ Beiträge zur Erzähltheorie Edi...

Author: Jan Christoph Meister

105 downloads 1406 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Narratology beyond Literary Criticism

Narratologia Contributions to Narrative Theory/ Beiträge zur Erzähltheorie

Edited by/Herausgegeben von Fotis Jannidis, John Pier, Wolf Schmid Editorial Board/Wissenschaftlicher Beirat Catherine Emmott, Monika Fludernik Jose´ Angel Garcıá Landa, Peter Hühn, Manfred Jahn Andreas Kablitz, Uri Margolin, Matıás Martıńez Jan Christoph Meister, Ansgar Nünning Marie-Laure Ryan, Jean-Marie Schaeffer Michael Scheffel, Sabine Schlickers, Jörg Schönert

6

≥

Walter de Gruyter · Berlin · New York

Narratology beyond Literary Criticism Mediality, Disciplinarity

Edited by Jan Christoph Meister in collaboration with Tom Kindt and Wilhelm Schernus

≥

Walter de Gruyter · Berlin · New York

앝 Printed on acid-free paper which falls within the guidelines 앪 of the ANSI to ensure permanence and durability.

Library of Congress Cataloging-in-Publication Data Narratology beyond literary criticism : mediality, disciplinarity / [edited] by Jan Christoph Meister ; in collaboration with Tom Kindt and Wilhelm Schernus. p. cm. ⫺ (Narratologia ; 6) Papers presented at a symposium held at Hamburg, November, 2003. Includes index. ISBN 3-11-018352-8 (alk. paper) 1. Narration (Rhetoric) ⫺ Congresses. 2. Discourse analysis, Narrative ⫺ Congresses. I. Meister, Jan Christoph, 1955⫺ II. Kindt, Tom. III. Schernus, Wilhelm. IV. Series. PN212.N42 2005 8081.00114⫺dc22 2005006990

ISBN 3-11-018352-8 ISSN 1612-8427 Bibliographic information published by Die Deutsche Bibliothek Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at ⬍http://dnb.ddb.de⬎. 쑔 Copyright 2005 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin All rights reserved, including those of translation into foreign languages. No part of this book may be reproduced in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Cover design: Christopher Schneider, Berlin Printed in Germany

Table of Contents

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii

Introduction: Narratology beyond Literary Criticism: Mediality – Disciplinarity Jan Christoph Meister/Tom Kindt/Wilhelm Schernus

ix

On the Theoretical Foundations of Transmedial Narratology . . . . Marie-Laure Ryan The Double Dynamics of Focalization in W. G. Sebald’s The Rings of Saturn . . . . . . . . . . . . . . . . . . . . . . . . Silke Horstkotte

1

25

Levels of Play and Narration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Britta Neitzel

45

Narrative in Music: The Case of Beethoven’s “Tempest” Sonata . . Douglass Seaton

65

Metalepsis as a Transgeneric and Transmedial Phenomenon. A Case Study of the Possibilities of ‘Exporting’ Narratological Concepts . . 83 Werner Wolf Measuring Narrativity in Literary Texts . . . . . . . . . . . . . . . . . . . . 109 Vyacheslav Yevseyev Quantitative Methods in Narratology: A Corpus-Based Study of Motion Events in Stories . . . . . . . . . . . . 125 David Herman The Lydgate Storyworld . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Alan Palmer ‘Selective’ History. Why and How ‘History’ Depends on Readerly Narrativization, with the Wehrmacht Exhibition as an Example . . 173 Daniel Fulda

vi

Table of Contents

The Narratological Fabric of the Gospels . . . . . . . . . . . . . . . . . . . 195 Ute E. Eisen Narrative Discourse and Identities . . . . . . . . . . . . . . . . . . . . . . . . . 213 Michael Bamberg Psychotrauma, Narration in the Media, and the Literary Public—and the Difficulties of Becoming Interdisciplinary . . . . . . 239 Harald Weilnböck The Eye of the Beholder. Narratology as Seen by Social Psychology . . . . . . . . . . . . . . . . . . . 265 Wolfgang Kraus Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 Contact Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297

Acknowledgments

The present volume contains revised versions of twelve papers presented at the Second International Colloquium held by the Narratology Research Group, Hamburg University, in November 2003. They have been complemented by an article kindly contributed by David Herman. We appreciate the trust invested in us, the organizers and editors, by the Research Group. Furthermore, we would like to mention our library, the Staats- und Universitätsbibliothek Hamburg Carl von Ossietzky, which made available to us its conference room. This wonderful venue contributed in no small way to the conference’s success. Particular thanks are due to our colleague and co-organizer Malte Stein who unfortunately could not participate in the preparation of this volume. Alastair Matthews translated the introduction into English and also edited, with great competence and his usual attention to detail, some of the articles. Birte Lönneker and Stefanie Thiedig assisted us in correcting and formating the manuscripts as well as in the preparation of the index. Finally, our thanks go to the series editors of ‚Narratologia. Contributions to Narrative Theory’ for publishing our volume in their esteemed series. Jan Christoph Meister Tom Kindt Wilhelm Schernus

Narratology beyond Literary Criticism: Mediality – Disciplinarity

Introduction A glance at the current situation in literary criticism shows that narratology, pronounced dead twenty years ago, is remarkably alive and well. This fact has been noted repeatedly and with understandable self-satisfaction in the recent literature on research into narrative theory. Just how astonishing this rebirth is, however, becomes apparent only when we step back from literary criticism and the humanities to take a wider historical view of the developments in academic and theoretical circles that preceded it. The deeply symbolic year of 1968 marked the fall of the academic ancien régime. Partly in anticipation of this and partly in response to it, a number of new leading disciplines were raised to power in western Europe as sources of hope for the future. However much they may have differed from one another in political purpose (in theoretical circles or beyond), linguistics, political economy, psychoanalysis, and structuralist semiology—to name but a few of the superdisciplines of the time—clearly belonged to one and the same paradigm in terms of how they conceived of themselves: throughout, they sought to reveal universal, ahistorical regularities in human thought and action in their respective fields. In the mid-1970s, however, a new epoch began, in which many disciplines explicitly distanced themselves from the search for universal patterns. The problems of context and historical contingency, previously ignored, were placed at the heart of the new theoretical systems. Some of the oncesupreme approaches could not or would not adapt accordingly; they were either marginalized (e.g. orthodox Marxist political economy) or redefined themselves as disciplines concerned with empirical social data (e.g. linguistic discourse analysis). Others responded by embracing the post-structuralist and subsequent deconstructivist paradigms with open arms. French narratology was one of the fields in which signs of this reorientation could be seen at an early stage. Roland Barthes’s S/Z of 1970 is perhaps the best evidence of this. A fundamental methodological shift, however, did not take place until the mid-1980s. One reason for the delay was the fact that narratology had become international in the meantime.

x

Introduction

When Derrida and Kristeva were leading figures in its homeland, the discussion of narratology in the English-speaking countries was covering an anachronistic combination of theories. Prince disseminated the formalist programme with his Grammar of Stories of 1973, after which some theorists engaged in retrospective reception of the structuralist and formalist classics (see for example Culler’s Structuralist Poetics: Structuralism, Linguistics and the Study of Literature of 1975). Even as this programme was being discovered, however, other writers were considering how it might be reformulated and extended to cover new methods and media. Chatman’s Story and Discourse: Narrative Structure in Fiction and Film of 1978 is an example of this. It combines the most influential Anglo-American, Russian, and French approaches. The delaying effect of this methodological ambivalence becomes clear in the next part of our survey of Anglo-American publications. The first, highly influential American translation of Genette appeared in 1980 (Narrative Discourse). Prince published his definitive formalist classic, Narratology: The Form and Functioning of Narrative, in 1982. At the same time, a second American translation of Genette was published (Figures of Literary Discourse). Finally, Rimmon-Kenan’s Narrative Fiction: Contemporary Poetics of 1983 provided a concise overall picture of what is now known as classical narratology prior to the caesura marked by Peter Brooks’s Reading for the Plot: Design and Intention in Narrative of 1984. In this book, Brooks presents an approach marked by its post-structuralist orientation and the fact that it is argued in terms of the aesthetics of effect and reception. The paradigm shift finally becomes apparent in all its enormity from 1985 onwards when Elain Showalter (The New Feminist Criticism of 1985) and Susan Sniader Lanser (Toward a Feminist Narratology of 1986) begin the project of creating a feminist narratology. With uncompromising radicalism and forcefulness, this project denies the original concept of narratology as the science of narrative universals. This tradition, which has given birth to many of the later new narratologies, has survived to the present day. In the mid-1980s, narratology, which was now showing increasing American influence, began to undergo a process of change marked by two tendencies. First, the turn away from formalism and structuralism meant that the narrative theory used in the study of literature began to draw on a wide range of theoretical paradigms. Second, narratological theories themselves were increasingly formulated and received with a view to particular applications, which could perfectly acceptably lie beyond the study of literature. The study of narratology had become a cross-disciplinary undertaking. Another, unrelated development should also be mentioned here: completely independent of the tradition whose evolution we have traced above, a variety of non-literary disciplines have shown an increasing

Introduction

xi

interest in narrative representation since as long ago as the early 1960s (e.g. socio-linguistic studies in the tradition of Labov and Waletzky). The same is true of the study of folk narrative, which goes back even further into the past. The Narratology Research Group in Hamburg has held two conferences to date on the evolution and refinement of narratology as described above. The first—What Is Narratology?—took place in May 2002.1 Its objective was to reconstruct the development of narratology and discuss the resultant choices open to and consequences facing us when we build and apply text-based narratological theories. The second conference followed in November 2003. The present book is named after its theme: Narratology beyond Literary Criticism. The aim here was to extend our models of and reflection on the development of narratological methodologies in two ways. First, we intended to investigate how narratological methods might contribute to the analysis and interpretation of symbolic representations in media other than the traditional, text-based medium of literature. Second, we asked: what is the position of narratology relative to other disciplines and their methodologies—put simply, where does narratology belong, who does it belong to, and who demands what of it? Mediality The call for papers on this ambitious project, whose objective was to link practical application with metatheoretical methodological reflection, met with a wide international response. Over fifty proposals for papers were submitted, of which only twelve could be selected for presentation at the conference. As the table of contents in this book shows, they dealt with a remarkably wide range of topics, including narrative representation in music, film, therapeutic discourse, computer games, and comics. Deliberately disconnecting the concept of narrativity from textual media proved, it can be seen, an extremely profitable move, and the editors are indebted to the contributors from various disciplines beyond literary criticism. By venturing into a narratological lion’s den, they have made an important contribution to expanding our horizons. In the process, the conventional interpretations and use of several narratological concepts are inevitably undermined at times. Orthodox theorists of literary narrative might consider these concepts sacrosanct, but, in the context of cross-disciplinary dialogue, we should treat this disruption as an opportunity and stimulus 1

Tom Kindt /Hans-Harald Müller (eds.): What Is Narratology? Questions and Answers Regarding the Status of a Theory. (Narratologia. Contributions to Narrative Theory / Beiträge zur Erzähltheorie 1). Berlin / New York 2003.

xii

Introduction

to re-examine our own terminology rather than insisting on canonical definitions. Tolerance and a readiness to learn are desirable not least because the narratology of literary criticism itself contains a large number of conceptual imports and thus tends to extend rather than preserve terminological meanings anyway. A less positive picture presents itself when we consider the second theme of the conference. We had hoped to encourage methodological metareflection on the consequences of deliberately extending narratological methods and their field of application, but, apart from a small number of exceptions, neither the papers nor the discussion lived up to this expectation. This criticism is directed neither at the authors and speakers, nor at the audience, but rather at ourselves, the organizers and editors. We had hoped that it would be relatively easy to turn the debate about the object domain of narratology into one about narratological theory itself on a reflexive metalevel, but this was probably a deformation professionelle typical of literary critics. Such an intention would seem positively hubristic to natural scientists, and it must strike those involved in empirically or historically oriented human and social sciences as somewhat strange at best. In literary narratology, as in most forms of literary theory and criticism, theories are constructed primarily on the basis of reference to particular canonical texts rather than using corpus analysis. That is to say, they have a selective rather than empirical basis. And, when our methods are applied in the context of such selective material, progressing rapidly from analytical description to building theoretical models seems not only acceptable but positively desirable. Crossing so easily from the discourse of data to that of theory, however, cannot but seem unusual when we are working with empirical evidence, as is standard practice for many of the participants in the conference and contributors to this volume who have narratological but non-literary interests. These self-critical remarks can be turned to our benefit. Our attempt to combine discussion of the object domain of narratology with reflection on its methods met with little success. But this need not be a cause for despondency—future narratological discussion should instead attempt to rectify two shortcomings that this failure has revealed. First, it is clear that we must, by means of informed reflection, examine the methodologies and evolution of narratology from a firmly metatheoretical perspective. Second—and this becomes particularly apparent with each new cross-disciplinary application of narratological theorems—the central concept of narratology itself, narrativity, is still not sufficiently well defined. The first essay in this volume makes interesting suggestions in this area. It and other, conceptually varied approaches to providing a more precise definition of our central concept deserve discussion in a context dedicated to this theme.

Introduction

xiii

No narratologist would dispute the fact that narration is a cross-medial phenomenon. However, there have been few attempts to date to engage in a reflexive, metatheoretical discussion on how the phenomenon should be defined conceptually. This may be an important explanation for why present narratological theory contains suggestions for many divergent definitions, most of which are distinctly media-specific in nature (predominantly literary, text-based), and tends not to draw links between them. A brief glance at the numerous more traditional definitions, which derive the concept of narrativity from quasi-material properties of narrating texts, is enough to show the scale of disagreement about the definition of narrativity. Some supporters of this traditional kind of definition refer to the discours level of narratives (e.g. Stanzel or the Genette of the Nouveau discours), others find it more appropriate to elucidate the concept of narrativity using the elements of the histoire level of texts. Those in favour of the latter method are deeply divided over the question of whether the resultant definition should have the character of a formal description in the manner of Prince or be aesthetic and normative in nature, as in Lotman’s tradition. There is also a series of definitions based on the theory that a proper definition of narrativity must make reference to both the plot and mediation levels of texts (e.g. Chatman or the Genette of the Discours). In contrast to these attempts at phenomenological explanation, recent years have seen the proposal of various functional definitions. They operate by identifying the function of narrativity. Their concepts of this function are usually postulated rather than developed by means of argument and can take very different forms in each particular case. These definitions are based, among other things, on the ontogenetic, socio-historical, cognitive, or metaphysical effects of narratives, or how they affect the aesthetics of reception (e.g. Fludernik, Herman, Ryan, Sternberg, Lanser, or Ricœur). Disagreement in the controversy surrounding the concept of narrativity is not, however, limited to the question of what features of artefacts and their functions a definition should be based on. The kind of definition best suited to help us grasp the phenomenon is also disputed. Whereas some treatments attempt to state necessary and sufficient conditions of narrativity, others consider it more appropriate to be guided by the concept of family resemblance or the ideas of prototype theory (Herman). Against the background of this situation (the description of which does not make any claim to be complete), the following question arises: what kind and form of definition of narrativity would be theoretically compatible with a narratology that claims cross-medial relevance for its theory and methodology? Three minimal requirements that such a definition of narrativity must fulfil can be identified.

xiv

Introduction

(1) The definition must be intuitively adequate. A cross-medial definition of narrativity must be valid and meaningful for all media that contain representations and can be intuitively classed as narrating media. (2) The definition must be medium-neutral. A cross-medial definition of narrativity can only stipulate as necessary those criteria and categories that appear as features in all the forms of symbolic representation covered by (1). For example, if musical representations are intuitively classed as potentially narrative, condition (2) prohibits all definitions of narrativity that involve the category of character in the essentialist or mimetic sense of a human agent—this category is not convincingly marked in musical representations. (3) The definition must relate to symbolic material. A cross-medial definition of narrativity must not retreat to capturing narrativity in some linguistic expression (in the sense of Wittgenstein’s language game) lacking reference to material objects. It must be possible to verify the fulfilment or non-fulfilment of the criteria of narrativity as objective structural features of concrete tokens (texts, films, transcripts of speech, musical pieces, etc.) without the criteria becoming essentialist ones as a result. In view of these minimal requirements, one of the most promising chances of developing a cross-medial definition of narrativity would seem to lie, at least at first glance, in functional and logical approaches—in definitions that are, in one form or another, based on feature sets that capture the relationship between cognitive, aesthetic, or other effects typical of narrative on the one hand and the logical structure underlying the phenomenology of all possible narrative realizations on the other. Disciplinarity The various disciplines from beyond literary criticism have very different links with narrativity, which is itself, as we have said, still insufficiently defined and thus can be drawn on only as a pre-theoretical frame of reference. It is completely indisputable that historiography has a narrative character. In the case of music, on the other hand, there will be some disagreement (and such debate can only be profitable) as to whether musicologists can speak of narrativity in music in anything other than a purely metaphorical linguistic expression. Applied narratologies are now emerging across the entire spectrum of academic disciplines, and it would seem to be no more than a matter of time until we find ourselves faced with, say, a mathematical narratology or an inorganic narratology. This somewhat ridiculous prediction is intended to highlight a second and perfectly serious problem:

Introduction

xv

assuming that we do manage to develop a robust cross-medial concept of narrativity, what criteria would we then use to determine the status of the various approaches to and theories about the medium-specific forms of narrative? Are people necessarily narratologists, and are they necessarily doing narratology, if they are dealing with narrative phenomena in their particular fields? In other words, is narratology (4) a superdiscipline whose abstract object is narrativity in all its forms in the different media in which it occurs; (5) an interdisciplinary project in the context of which various individual disciplines can constantly inductively exchange and combine new object- and interest-specific definitions of narrativity; or (6) a formally defined procedure for elucidating its object, possessing interdisciplinary relevance and giving primacy to the category of narrativity in its theoretical system? These possibilities can only be compared with one another if there is agreement regarding the criteria with which they should be evaluated. We would argue that particular consideration should be given to whether the possibilities listed above (which need not be the only ones) are institutionally realistic, cognitively profitable, and theoretically and conceptually plausible. We are inclined to favour the last possibility (6); we find the prospect of a superdiscipline institutionally unlikely, and we feel that the project of interdisciplinary discourse favoured in the present climate displays too much arbitrariness and too little transparency. We do not deny that culturalist and contextualist definitions of our subject matter result in a massive expansion of the object domain of narratology and thus encourage interdisciplinary exchange. On the other hand, a serious methodological problem must also be pointed out, one that featured in the debate about the concept of intertextuality prior to contextualism and culturalism: the concept of context itself still lacks anything approaching a satisfactory theoretical definition. For this reason, the contextualization advocated so widely leads, sooner or later, to a methodological dilemma that threatens to undermine the scientific status of narratology itself. Thus, we suggest, the best future for narratology lies in retaining a critical awareness of the fact that it was originally intended to search for and study universals. The theoretical task facing contemporary narratology is not an easy one: it is easy to see that the phenomenon of narrativity occurs in different media; what we have to do is state what it actually is. It can be disputed whether such a statement should be couched in terms of cognitivism, aesthetics, the logic of representation, or perhaps a synthetic combination of some or all of these frameworks—but this does not change our conviction that a con-

xvi

Introduction

text-sensitive formalism with reformed methods gives narratology a better chance of doing what it is intended to do than does a contextualism with inadequately controlled methods. The six issues we have raised here under the headings of mediality and disciplinarity, and the arguments with which we have responded to them, have two purposes. First, we hope they will indicate the direction that should be taken by the theoretical reflection that we believe to be necessary in narratology. Second, and consequently, we have outlined an overarching set of questions against which the reader can consider the individual contributions gathered together in this book. These are the questions that gave rise to the idea of our conference on narratology beyond literary criticism in Hamburg. We thank all those who, regardless of differences in method and concepts, helped elucidate these questions, whether by writing papers or contributing to the discussion. When editing this volume, as when organizing the conference, we deliberately strove to resist the temptation of enforcing conceptual and methodological standardization. We do not suggest that the following essays speak the same language. But the reader will find nonetheless that they speak about the same thing—even if our picture of it is not yet as clear as it might be. Jan Christoph Meister Tom Kindt Wilhelm Schernus

Hamburg, Göttingen, September 2004

On the Theoretical Foundations of Transmedial Narratology Marie-Laure Ryan (Bellvue, Colorado) 1. Introduction Narratology, the formal study of narrative, has been conceived from its earliest days on as a project that transcends disciplines and media. In 1964, Claude Bremond wrote: [Story] is independent of the techniques that bear it along. It may be transposed from one to another medium without losing its essential properties: the subject of a story may serve as argument for a ballet, that of a novel can be transposed to stage or screen, one can recount in words a film to someone who has not seen it. These are words we read, images we see, gestures we decipher, but through them, it is a story that we follow; and it could be the same story (quoted from Chatman 1978:20).

With its almost exclusive concentration on verbal narrative, and within verbal, on literary fiction, the development of narratology has tended to ignore these statements. The purpose of the present volume is to correct this trend and to replace narratology on the transmedial and transdisciplinary track. But in the meantime, hopefully, our understanding of media has grown more sophisticated. We no longer believe that all media offer the same narrative resources and that all stories can be represented in media as different as literature, ballet, painting and music. Nor do we believe that the migration of a story from a medium to another does not present cognitive consequences. A core of meaning may travel across media, but its narrative potential will be filled out, actualized differently when it reaches a new medium. When it comes to narrative abilities, media are not equally gifted; some are born storytellers, others suffer from serious handicaps. The concept of narrative offers a common denominator that allows a better apprehension of the strengths and limitations in the representational power of individual media. Conversely, the study of the realization of narrative meaning in various media provides an opportunity for a critical reexamination and expansion of the analytical vocabulary of

2

Marie-Laure Ryan

narratology. The study of narrative across media is consequently beneficial to both media studies and narratology. This essay explores the theoretical foundations of transmedial narratology in both their negative and positive components. The negative component describes the positions that are incompatible with transmedial narratology, while the positive component explores the concepts of narrative and of medium that are compatible with the project.

2. Positions Hostile to Transmedial Narratology The main obstacle to the transmedial study of narrative is a position that comes from within narratology itself, namely what I call the language-based, or rather, speech-act approach to narrative. This position (represented by Prince 1987, Genette 1972, and Chatman 1978) defines narrative as an act of storytelling addressed by a narrator to a narratee, or as the recounting of a sequence of past events (Abbott 2002). In these definitions, the condition for being a narrative is the occurrence of the speech act of telling a story by an agent called a narrator. The semantic content of this speech act must be events that already occurred, either actually or in make-believe. This conception of narrative as a language-based phenomenon not only rejects the possibility of visual or musical forms of narrative, it also excludes texts with a language track that do not use an overt narrator, or texts that do not represent events retrospectively. For instance, Gerald Prince writes in his Dictionary of Narratology: ‘A dramatic performance representing many fascinating events does not constitute a narrative, since these events, rather than being recounted, occur directly on stage’ (1987:58). According to this view (which Prince abandons in the second edition of the Dictionary), the transmedial study of narrative is limited to the distinction between oral storytelling and written literary narrative. Some theorists endorse a milder form of this position that uses the speech-act based definition as a metaphorical model for the analysis of non-verbal texts. In contrast to the radical approach, the mild version accepts the possibility of visual or dramatic narratives, but only if these texts can be fitted into the verbal mold. This approach would analyze drama and movies as the utterance of a narratorial figure, even when the film or the play does not make use of voice-over narration. Its advocates include Metz 1974, Chatman 1990, Gaudreault and Jost 1990, and its opponents Bordwell and Thompson 1990. Another position incompatible with the study of narrative across media is the doctrine of radical media relativism. Though I cannot associate it with specific names, it is latent among critics influenced by Saussurian

Foundations of Transmedial Narratology

3

linguistics and deconstruction. Radical relativism regards media as selfcontained systems of signs, and their resources as incommensurable with the resources of other media. Just as two languages cannot convey the same semantic values under the doctrine of linguistic relativism, two different media cannot convey similar meanings or use similar devices under the doctrine of medial relativism. This view comes in a strong and a weak form. In the strong form, the signified cannot be separated from the signifier. Since a transmedial concept of narrative presupposes a distinction between narrative meaning and the signs that carry it, the strong interpretation kills in the egg the project of transmedial narratology. In its weaker form, medial relativism accepts common meanings but insists on the uniqueness of the expressive resources of each medium, thereby forcing the theorist to rebuild the analytical toolbox of narratology from scratch for every new medium. This approach ignores the fact that the conceptual tools of verbal narratology are often borrowed from other media: for instance, theme from music, perspective from painting, and camera-eye narration from the cinema. In some cases borrowing seems inevitable: for instance when a medium tries to imitate the effects of another medium, or when two media share a common channel. The alternative to medial relativism is to recognize that theoretical concepts can be either medium-specific, or applicable to several media. Examples of narratological concepts that apply across media are the distinction story / discourse, as well as the notions of character, event, fictional world and metalepsis (Wolf, this volume). On the other hand, montage is a technical concept native to film; but literary critics have borrowed it when language-based narrative began to imitate some of the techniques of the cinema.

3. Defining Narrative In the past ten years or so, the term ‘narrative’ has enjoyed a popularity that has seriously diluted its meaning. Jerome Bruner speaks of narratives of identity, Jean-François Lyotard of the ‘Grand Narratives’ of a capitalized History, Abbe Don of the narratives of interface in computer software, and everybody speaks of cultural narratives, meaning by this not a heritage of traditional stories but the collective values that define a culture, such as belief in free speech in Western societies, or latent stereotypes and prejudices, such as narratives of race, class and gender. The dissolution of ‘narrative’ into ‘belief’, ‘value’, ‘experience’, ‘interpretation’, or simply ‘content’ can only be prevented by a definition that stresses precise semantic features, such as action, temporality, causality and world-construction. A transmedial definition of narrative requires a broadening of the concept beyond the

4

Marie-Laure Ryan

verbal, but this broadening should be compensated by a semantic narrowing down, otherwise all texts of all media will end up as narratives.1 As I have already mentioned, the main problem facing the transmedial study of narrative is to find an alternative to the language-based definitions that are common fare in classical narratology. A useful point of departure is a very sketchy, but fairly uncontroversial definition proposed by H. Porter Abbott: narrative is ‘the representation of an event or a series of events’ (Abbott 2002:12). What makes this definition useful, despite its lack of precision, is that it relies on the medium-free concept of representation. It tells us that narrative is a signifier of a variable nature—the representation—that calls to mind a meaning—or signified—of a fixed nature, the ‘event or series of events’. But I would like to make the prefix re of representation optional, in order to accommodate two kinds of texts: narratives that represent events from a retrospective point of view—events that are already ‘in the book’ of history; and narratives that present events as they occur, as is the case with live broadcasts, computer games, and perhaps even drama and movies. If the transmedial identity of narrative lies on the side of the signified, this means that narrative is a certain type of mental image, or cognitive template which can be isolated from the stimuli that trigger its construction. I propose to define the cognitive template constitutive of narrative through the following features. 1. Narrative involves the construction of the mental image of a world populated with individuated agents (characters) and objects (spatial dimension). 2. This world must undergo not fully predictable changes of state that are caused by non-habitual physical events: either accidents (happenings) or deliberate actions by intelligent agents (temporal dimension). 3. In addition to being linked to physical states by causal relations, the physical events must be associated with mental states and events (goals, plans, emotions). This network of connections gives events coherence, motivation, closure, and intelligibility and turns them into a plot (logical, mental and formal dimension). 1

As Werner Wolf has argued (2003:181f), an example of a definition of narrative that seems too broad for transmedial narratology is Monika Fludernik’s (1996) suggestion that narrative is in essence an expression of human experience, and that plot is not a necessary dimension. I am certainly not questioning the importance given by this definition to human (or anthropomorphic) experience. But if ‘experientiality’ makes plot dispensable, any piece of music that arouses emotions (and don’t they all?), any abstract painting titled ‘War’, ‘Anguish’, or ‘Serenity’, and any lyrical poem would be every bit as narrative as a classical novel, and more narrative than a historical chronicle that focuses on external events.


5

This definition consists of hard and fast rules that specify minimal conditions. One of the conditions listed under 2 appears however more controversial that the others: is it necessary for a story to involve non-habitual events, or can it concern fully routine actions, such as the stereotyped scripts which constitute in part our general knowledge of the world? This dilemma points to an area where the issue of narrativity is particularly difficult to disentangle from the issue of what makes a story worth telling. Should a definition of narrative involve aesthetic principles, defined through preference rules, in addition to hard and fast conditions, or should it outline a bare wireframe common to all stories? Some theorists (Bruner 1991) argue that a story must have a point, and that principles of tellability are therefore an integral part of a definition of narrative. Others will reply that there are good and bad stories and that storytelling is not necessarily performed for the sake of entertainment. We may for instance produce narratives as a response to questions from a doctor, a prosecutor, a teacher, or a parent who wants explanations for our behavior, and these stories can satisfy the questioner without fulfilling the felicity conditions that pertain to what Pratt (1977) calls ‘narrative display texts’, such as conversational storytelling or the production of literary fiction. I believe that there is a continuum that runs from strict conditions of narrativity to conditions of tellability, and from tellability to full-fledged aesthetic principles, such as the Aristotelian principles that recommend exposition, conflict, climax, and a resolution brought up by a dramatic turn of events. A definition of narrativity cannot be limited to the great masterpieces of literature, and it must therefore draw a line, however contestable, somewhere between the two poles. This is why I haven’t included precise guidelines under 3 on matters such as coherence, conflict, and closure. Each of the conditions listed above implies the preceding one. We can therefore have non-narrative texts that fulfil (1) only (e.g. the static description of a world, as we find in ethnographic works or in Borges’ “Tlön, Uqbar, Orbis, Tertius”), mildly narrative texts that fulfil (1) and (2) (e.g. chronicles of random or unrelated actions, reports of dreams, or certain incoherent stories told by children)2, and fully narrative texts that fulfil all three condition. But we cannot have (3) without (2), unless the

2

For instance this story told by a four-year old boy, reported by Brian Sutton-Smith (1997: 161–62): ‘Once there was a dragon who went poo poo on a house and the house broke then when the house broke the people died and when the people died their bones came out and broke and got together again and turned into a skeleton and then the skeletons came along and scared the people out of the town and then when all the people got scared out of the town then skeleton babies were born and then everyone called it skeleton town […].’

6

Marie-Laure Ryan

actions of (2) are purely virtual (as is the case in novels that focus on the ruminations of Hamlet-like characters who are unable to take action), nor (2) without (1), since there cannot be changes of state, and consequently history, without concrete entities that undergo metamorphoses. The division of my definition of narrative into three components means that the degree of narrativity of a given text cannot be evaluated by simple algorithms, such as computing the proportion of static and active events.3 While all three conditions must be fulfilled for a text to be widely accepted as narrative, individual narrative texts will give variable emphasis to the features specified in the three conditions. With their detailed construction of an imaginary world, science fiction and fantasy over fulfil condition 1, and they often treat the plot as a mere discovery path across fictional space. The demand for action and changes of state that makes up the second condition is the dominant feature of thrillers and adventure stories, while the third condition, by insisting on causal connections and on the mental constructs that render actions intelligible, rules over detective stories and heavily plotted narratives, such as comedies of errors. Neglect of the third condition leads by contrast to the low narrativity of many postmodern novels. While these novels create a world, populate it with characters, and make something happen (though they often take liberties with condition 2), they do not allow the reader to reconstruct the network that motivates the actions of characters and binds the events into an intelligible and determinate sequence. Rather than locating narrativity in an act of telling, the present definition anchors it in two distinct realms. On one hand, narrative is a textual act of representation (or presentation)—a text that encodes a particular type of meaning. The definition remains unspecific as to what type of signifier is used to encode this meaning. On the other hand, narrative is a mental image built by the interpreter as a response to the text. But it does not take a representation proposed as narrative to trigger the cognitive template constitutive of narrativity: we may form narrative scripts as a response to life itself. For instance, if I observe a fight on the subway, I will construct in my mind the story of the fight, in order to tell it to my colleagues when I get to work. To account for the narrative potential of life, I propose to make a distinction between ‘being a narrative’, and ‘possessing narrativity’. The property of ‘being a narrative’ can be predicated of any semiotic object produced with the intent to evoke a narrative script in the mind of the audience. To be more precise, it is the receiver’s 3

This method of assessing narrativity is implicit to Gerald Prince’s discussion of the topic (1982, chapter 5). Also see the contributions by Yevseyev and Herman in the current volume.


7

recognition of this intent that leads to the judgment: this text is a narrative, though we can never be sure that sender and receiver have the same story in mind. ‘Having narrativity’, on the other hand, means being able to evoke such a script, whether or not the author of the text intended to do so, and whether or not there is an author (Ryan 2004:9). For instance, if a random pattern of cracks in a rock happened to resemble writing, and if the writing could be read as a story, the pattern of cracks would not be a narrative, but it would certainly possess narrativity by the present definition. The idea that life as well as texts can possess some degree of narrativity does not mean that narrative plots are the sort of thing that are found in the world, a view that Hayden White successfully put to rest. I am rather saying that the cognitive construct typical of narrativity can be a response to affordances and stimuli that originate within the world. Making stories out of life experience is neither picking up ready-made objects that somehow lie out there, nor constructing these objects out of any kind of materials, but finding a fit between what life offers, and the cognitive pattern that we try to fill in. Sometimes life provides adequate raw materials, sometimes it does not, and sometimes the materials are created by the imagination. The fullest realization of narrative occurs when we have a text that is both intended as narrative, and possesses sufficient narrativity to be construed as such. But the properties of being a narrative and having narrativity can be dissociated in a variety of ways. The standard case of dissociation occurs when the story is so poorly presented that the audience cannot reconstrue the proper script, even though it recognizes a narrative intent. In this case the text is a narrative of low narrativity. The opposite case is the already-mentioned case of a life situation that the cognitive subject interprets in narrative terms. The property of being a narrative is much more clear-cut than the property of having narrativity, since it depends on the sender’s intent, but it becomes fuzzy in a number of situations. For instance, what do you call a text that uses narrative scripts in an instrumental way? A sermon or a philosophical text typically resorts to parables and narrative examples on the micro-level, but the global purpose of the text is certainly not to tell a particular story. Computer games present a similar situation: they may use a narrative script to attract the player, but once the player is immersed in the game, he usually forgets whether he is fighting terrorists, car thieves, or evil aliens, or whether he is a terrorist himself. All that matters is winning the game through clever use of resources. The question ‘is it a narrative?’ is even more problematic when the text embodies the artistic intent to both arouse and frustrate narrative desire. Many postmodern texts present themselves as bits of pieces of a narrative image but prevent the reader from ever achieving

8

Marie-Laure Ryan

the reconstruction of a stable and complete narrative script. This explains why narrative theory has never been comfortable with either including or excluding postmodern literature (Ryan 2004:9–10). My endorsement of a cognitively rather than verbally based definition of narrative should not be taken as an unconditional adherence to a position that has recently taken cognitive science, the social sciences, and the humanities by storm. This position proclaims the fundamentally narrative nature of thought, and it equates our never-ending efforts to make sense of the world and of our lives with a process of ‘emplotting’ or ‘storification’. Without denying that storytelling (to oneself or to others) is an efficient way to make life and the world more intelligible, and that the formation of narrative scripts plays an important role in mental life, I believe that there are sense-making operations that do not take narrative form: capturing the laws of physics through an equation such as E=mc2 fails for instance the first and second conditions of my definition, since it produces a timeless law rather than a historical scenario involving individual entities. Sense-making can also result from the drawing of analogies and contrasts between phenomena, rather than from the chronological and causal ordering of individual events. The mental construct that I regard as constitutive of narrative admittedly puts into play cognitive processes that we also use in everyday life, such as focusing thought on certain objects cut out from the flux of perception, a process that also enables us to distinguish discrete states and events; inferring causal relations between these states and events; thinking of events as situated in time; and reconstructing the content of other peoples’ minds as an explanation of their behavior. We resort to these mental operations when we drive a nail with a hammer (acting on the basis of inferred causal relations), when we plan our schedules (temporal ordering), when we make grocery lists (focusing on certain items selected from the wide range of available products), and when we participate in social interaction, especially conversation (reading other people’s minds). Some cognitivists will say that these processes are inherently narrative, and that activating them is an act of storytelling. But this is a rather circular reasoning that leads to entirely predictable results, and that doesn’t explain anything: how could thought produce anything but stories, if it is in essence a narrative activity? I would like to suggest on the contrary that the activation of one or the other of these cognitive processes is not sufficient to produce narratives, because they can operate independently of each other, as my examples suggest. It is only when they all come together and form a reasonably stable mental image that they generate representations that fulfil all the conditions of my definition. Narratives are more than temporary drafts in the theater of the mind, more than transitory firings of neurons in the brain along individual


9

pathways; they are solidified, conscious representations produced by the convergence of many different mental processes that operate both within and outside stories.

4. Revisiting the Relations of Language and Narrative The adoption of a cognitive definition may uncouple narrative from verbal expression, but it does not entirely solve the thorny problem of the relation between language and narrative. It seems clear that of all semiotic codes language is the best suited to storytelling. Every narrative can be summarized in language, but very few can be retold through pictures exclusively. The narrative limitation of pure pictures stems from their inability to make propositions. As Sol Worth has argued, visual media lack the code, the grammar and the syntactic rules necessary to articulate specific meanings. A propositional act consists of picking a referent from a certain background and of attributing to it a property also selected from a horizon of possibilities. While language can easily zero in on objects and properties, pictures can only frame a general area that contains many shapes and features. Pictures may admittedly find ways around their lack of propositional ability to suggest specific properties (for instance through caricature), but there are certain types of statements that seem totally beyond their reach. As Worth argues, pictures cannot say ‘ain’t’. Nor, as Rimmon-Kenan observes (1989:162) can they convey possibility, conditionality, or counterfactuality. Werner Wolf is more specific on this issue: ‘pictures do not have at their disposal the means that verbal narratives have for thematizing the presence [= bringing to mind?] of actually absent or virtual realities of the represented world’ (Wolf 2003: 192). Moreover, being limited to the visible, they are unable to express abstract ideas, such as causality. Only language can make it explicit that the queen died of grief over the death of the king, or that the fox stole the cheese from the crow by fooling him into believing something that was not the case (Ryan 2004:10). The narrative limitations of music are even more blatant than those of pictures, because sound waves (or tones and rhythms) are not in themselves semiotic objects. Pure sounds can be used to evoke mental images, some of which may resemble stories, but they possess neither an iconic meaning, as do pictures, nor a context-independent, stable core of signification definable by lexical rules, as does language (Ryan 2004:11). All of these observations seem to support the conclusion that verbal language is the native tongue of narrative, its proper semiotic support. Without denying the unsurpassed narrative ability of language, I would

10

Marie-Laure Ryan

like to defend a more nuanced position. If we define narrative in cognitive terms, it is not a linguistic object but a mental image. While it may be true that only language can express the causal relations that hold narrative scripts together, this does not mean that a text needs to represent these relations explicitly to be interpreted as narrative. Take for example a sequence of seemingly disconnected cinematic images: First you see a man tossing and turning, unable to sleep. Then a mirror breaks. Then a telephone rings.

The film theorists Bordwell and Thompson argue that these images can be constructed into a narrative sequence if the spectator supplies common agents and logical connections. For instance ‘The man can’t sleep because he’s had a fight with his boss, and in the morning is still so angry that he smashes the mirror while shaving; next, his telephone rings and he learns that his boss has called to apologize’ (Richardson 2000:170). The visual track may be unable to explicitate causal relations, but what matters in this case is the spectator’s ability to infer them from the text. Even purely verbal texts, which are perfectly able to say ‘the queen’s grief over the king’s death caused her to die’ usually dispense with such explanations (Ryan 2004:11). One of the major problems faced by transmedial narratology is to reconcile the centrality of language with the original contributions made by other media. Language all by itself can support a wider variety of narratives than any other single-track medium, not just because of its logical superiority but also because only words can represent language and thought. Dialogues and representations of mental life are indeed a major component of narrative. But this does not mean that media based on sensory channels cannot make unique contributions to the formation of narrative meaning. There are, quite simply, meanings that are better expressed visually or musically than verbally, and these meanings should not be declared a priori irrelevant to the narrative experience. The philosopher Kierkegaard claimed for instance that the desire of Don Juan—a desire that doesn’t want to die, and yet doesn’t want to renounce fulfilment—can only be expressed by the music of Mozart. Narrative is not just the wire frame plot captured by summaries, but the expansion of this wire frame into a total imaginative experience, into a spatio-temporal world to which we react intellectually, emotionally and sometimes aesthetically. Verbal narrative certainly offers the most balanced and diverse expansion of the wire frame, but other media can outdo language in their expansion of one of its particular dimensions. For instance, the sound track in a movie conveys certain moods and elicits emotional responses much more powerfully that


11

the image and the dialogue can do. I notice for instance that when I cry during movies, it is always in moments when music is playing; and I cry in movies much more easily than I do while reading. 5. Narrative Modes If narratology is to expand into a medium-free theory, the first step to be taken is to recognize other narrative modes than the standard way of evoking narrative scripts: telling somebody else that something happened. I do not take this term of mode in the traditional narratological sense defined by Genette 1972 (who uses it as a rather vague umbrella term for concepts such as frequency, direct and indirect discourse, perspective and focalization), but in a personal sense, to mean a distinct way to bring to mind the cognitive construct that defines narrativity. Let me illustrate this concept of modality through a list of concrete examples. This list, which I regard as open-ended, consists of binary pairs. In each case the left term can be regarded as the unmarked case, because the texts that present this feature will be much more widely accepted as narrative than the texts that implement the right-hand category (discussion of modes from Ryan 2004:13–15). External / Internal: In the external mode, narrative meaning is encoded in material signs; it is textualized. In the internal mode, it is does not involve a textualisation: we can tell ourselves stories in the privacy of our minds (cf. Jahn 2003). Diegetic / Mimetic: This distinction goes back to Plato’s Republic. A diegetic narration is the verbal storytelling act of a narrator. As the definition indicates diegetic narration presupposes language, either oral or written; it is therefore the mode typical of the novel and of oral storytelling. A mimetic narration is an act of showing, a ‘spectacle’. In forming a narrative interpretation the recipient works under the guidance of an authorial consciousness, but there is no narratorial figure. Mimetic narration is exemplified by all dramatic arts: movies, theater, dance, and the opera. But each of these two modes can intrude into a narration dominated by the other. The dialogues of a novel are islands of mimetic narration, since in direct quote the voice of the narrator disappears behind the voice of the characters; and conversely, the phenomenon of voiced-over narration in cinema reintroduces a diegetic element in a basically mimetic medium. Autonomous / Illustrative: In the autonomous mode, the text transmits a story that is new to the receiver; this means that the logical armature of

12

Marie-Laure Ryan

the story must be retrievable from the text. In the illustrative mode, the text retells and completes a story, relying on the receiver’s previous knowledge of the plot. The illustrative mode is typical of pictorial narratives, for instance of medieval paintings of Biblical scenes. Halfway between these two poles are texts that offer a new, significantly altered version of a familiar plot, such as a modern retelling of a classical myth. Receptive / Participatory: In the receptive mode the recipient plays no active role in the events presented by the narrative: he merely receives the account of a narrative action, imagining himself as an external witness. In the participatory mode the plot is not completely pre-scripted. The recipient becomes an active character in the story, and through her agency she contributes to the production of the plot. This mode has been practiced for quite a while in staged happenings, ‘improv’ theater, and scripted role-playing games (e.g. “Dungeons and Dragons”) but it has flourished with the advent of interactive digital media. In many computer games, for instance, the user is represented in the game world through an avatar. By solving problems in the real time of the game session she determines whether the life-story of this avatar will end in success or failure, or how long the avatar will live. Determinate / Indeterminate: In the determinate mode the text specifies a sufficient number of points on the narrative trajectory to project a reasonably definite script. In the indeterminate mode, only one or two points are specified, and it is up to the interpreter to imagine one (or more) of the virtual curves that traverse these coordinates. The indeterminate mode is typical of narrative paintings that tell original stories through the representation of what Lessing calls a pregnant moment. The pregnant moment opens a small temporal window that lets the spectator imagine what immediately preceded and what will immediately follow the represented scene. But a full-blown story normally covers an extended stretch of time, and every spectator will probably imagine the remote past and the remote future in a different way. Literal / Metaphorical: What constitutes a literal or metaphorical narration depends on the particular definition given to narrative. While literal narration fully satisfies the definition, the metaphorical brand uses only some of its features. The degree of metaphoricity of a narrative thus depends on how many features are retained, and on how important they are to the definition. The great advantage of recognizing a metaphorical mode is that it enables narratology to acknowledge many of the contemporary extensions of the term ‘narrative’ without sacrificing the precision of its core definition.


13

Here are some examples of what I consider metaphorical types of narrative. If we define narrative as the representation of a world populated by individuated characters, and if characters are intelligent agents, the following relaxations of the definitions should be regarded as metaphorical: scenarios about collective entities rather than individuals (e.g. the ‘Grand Narratives’ of Lyotard, as well as their heirs, the ‘narratives of class, gender and race’ of contemporary cultural studies);4 narratives about entities deprived of consciousness (e.g. Richard Dawkins’ exposition of biology as the story of ‘selfish genes’), and dramatizations that attribute agency to abstract objects. Here is an example proposed by the mathematician Ken Devlin: ‘Mathematicians deal with a collection of objects—numbers, triangles, groups, fields—and ask questions like, ›What is the relationship between objects x and y? If x does thus to y, what will y do back to x?‹ It’s got plot, it’s got characters, it’s got relationships…a bit of everything you can find in a soap opera.’5 In other words, mathematics is a story! This is certainly an extreme case of metaphorical narrative. If we want to stretch the metaphor to its limits, we can apply it to art forms deprived of semantic content, such as music and architecture. In the case of music, the metaphor can be invoked to analyze the structure of the work in terms of narrative effects or narrative functions. Narrative terminology is indeed common in music theory: relations between chords are described as exposition, complication and resolution. Given a specific exposition and complication, only certain chords will provide a satisfactory resolution. In this metaphorical interpretation, all music becomes narrative, while if we use an illustrative interpretation, narrativity is a feature that occurs in only some compositions—those that allude to a narrative through their title, such as the Don Quixote Suite, by Telemann, or the Apprentice Sorcerer, by Paul Dukas. In the case of architecture, a metaphorical interpretation would draw an analogy between the temporality of plot and the experience of walking through a building. In a narratively conceived architecture, the visitor’s discovery tour is plotted as a meaningful 4

5

Does this conception of metaphorical narrative apply to works of history—such as histories of France or of the Mediterranean? On the level of the title, it certainly does: France and the Mediterranean are collective entities deprived of consciousness. But if we look into these works, we notice that they do not narrate the destiny of an allegorical character named ‘France’ (or Marianne) or the Mediterranean, attributing to this character a mind of her own and particular goals. They rather tell the concrete stories of the leaders who shaped the destiny of large numbers of people, or of more anonymous individuals whose story typifies the living conditions of an entire social class. Therefore, while the project of producing the history of a country, region, or community is metaphorical on the macro-level, it is made of multiple literal narratives. If the project remains on the level of general statements about collective entities, it is not narrative. Quoted in The Denver Post, January 9, 2001:2.

14

Marie-Laure Ryan

succession of events. This occurs in Baroque churches, where the visitor’s tour is supposed to reenact the life of Christ.

6. What Are Media? The concept of medium is no less problematic than the concept of narrative. Just as structural linguists face the task of defining the units of the language system, transmedial narratologists must decide what counts as a distinct medium from a narrative point of view. This may seem a strange question: don’t we all instinctively know what media are? And yet, if we ask specialists of different disciplines to propose a list of media, we will receive a bewildering variety of answers. A sociologist or cultural critic will answer: TV, radio, cinema, the Internet. An art critic may list: music, painting, sculpture, literature, drama, the opera, photography, architecture. An artist’s list would begin with clay, bronze, oil, watercolor, fabrics, and it may end with exotic items used in so-called ‘mixed-media’ works, such as grasses, feathers and beer can tabs. An information theorist or historian of writing will think of sound waves, papyrus scrolls, codex books, and silicon chips. A philosopher of the phenomenologist school would divide media into visual, aural, verbal, and perhaps tactile, gustatory and olfactory. In media theory, as in other fields, what constitutes an object of investigation depends on the purpose of the investigator (Ryan 2004:16). Can one subsume all these conceptions of medium under a unified theory? If we look back at the answers of our imaginary informants, we can classify them into three broad categories: semiotic, material-technological, and cultural. The semiotic approach looks at the codes and sensory channels that support various media. It tends to distinguish three broad media families: verbal, visual and aural. It is only our habit of not ranking cuisine and perfume among media—probably because they do not transmit the proper kind of information—that prevents this list from including olfactory and gustatory categories. The groupings yielded by the semiotic approach broadly correspond to art types, namely literature, painting and music, but they extend beyond the aesthetic use of signs; language for instance has both literary and non-literary uses; pictures can be artistic or utilitarian. In its narratological application, the semiotic approach investigates the narrative potential and limitations of a given type of signs. Werner Wolf (2002) asks for instance to what extent sound and pictures are able to narrate, and he ranks the three major media families, or art types, in this decreasing order of story-telling ability: verbal, visual and musical. In a similar vein, Lessing’s ground-breaking study Laocoon investigates the


15

differing storytelling affordances of painting, a spatial medium, and poetry, an art form supported by the temporal medium of language. Left by itself, the semiotic approach yields only broad families. To bring further refinement to media theory, we must ask about the materials and the technologies that support the various semiotic types. Within each semiotic group of media, we must distinguish raw material supports, such as clay for pottery, stone for sculpture, the human body for dance, or the human vocal apparatus for music, from supports that are elaborated through technological means. It is further necessary to distinguish technologies of pure reproduction, such as sound recording or xerox copying, from technologies that create new media objects and open new expressive possibilities. Only the latter presents interest for transmedial narratology. Moreover, not all technologies that bring expressive diversity in a media family do so in a narratively significant way. In the sound category, for instance, diversity is created by the various musical instruments developed through the ages, but none of them has significantly increased the limited narrative potential of music. Much more consequential for narrative are the technologies that affect language-based and visual media. In the language category, these technologies correspond to the various ways to inscribe verbal signs (manuscript writing, print, and digital encoding),6 as well as to the various methods of encoding and transmitting spoken language (radio and telephone). In the visual domain, the most narratively significant technologies correspond to methods of capture, such as photography, film, and television. The digital encoding of images has also brought new expressive possibilities, but their narrative impact is questionable. The technological approach not only refines semiotic categories, it also cuts across them and reorganizes media into different families: media of long distance communication, media of the moving image, and above all, ‘old media’ versus ‘new media’. This label of new media may be used in a narrow sense to cover media or submedia that only perform through the computer (VR installations, video games, e-mail, Internet chat, hypertext), and in a wide sense, to describe media that use digital technology as mode of production, but end up being taken out of the computer (digital photography, digital recordings, and films with computer-generated scenes). Technology also regroups semiotic families into multiple-channel media (or ‘multi-media media’) that affect several senses. The classical example of an approach to transmedial narratology based on technological categories is the work of Walter Ong on the impact of the invention of writing on narrative form. 6

This list should also make room for the kinds of objects that support these various modes of inscription: papyrus scrolls, codex books, the computer screen, and space itself for VR technology.

16

Marie-Laure Ryan

The third important dimension of media is their cultural use. This dimension is not entirely predictable from semiotic type and technological support. In fact, some ways of disseminating information are regarded as distinct media from a cultural point of view, despite their lack of a distinct semiotic or technological identity. Newspapers, for instance, rely on the same semiotic channels and printing technology as books, but ‘the press’ is widely regarded by sociologists as a medium in its own right, on par with the other so-called ‘mass media’ of TV, film, radio and the Internet. By far the majority of media studies have been devoted to cultural use. These studies will ask for instance about the social impact of film violence, Internet pornography, television news reporting, or multi-user computer games. In a study of cultural use, consideration must also be given to the network of relations among media, a network commonly described through the metaphor of media ecology (Heise 2002). For instance, the cultural role of the cinema shifted after the invention of television, though the technology itself did not undergo significant changes. In the pre-television days, movie theaters showed a variety of features: news reels, documentaries, cartoons, and a feature film. They combined reality-based and fiction films in a continuously running show. This diversity and continuous running has been taken over by television, and nowadays movies specialize mainly in the fictional, with a distinct preference for the fantastic, while TV favors reality (or reality-effects), in the form of news, documentaries, and fictional representations of everyday life. It has even turned the real into a spectacle in the increasingly popular genre of ‘reality TV’. The semiotic, the technological and the cultural dimension all affect the narrative potential of media, but they do so to variable extents. Take for instance the case of television, compared to film and to the theater. On the semiotic level TV is not distinct from these other two media: all three involve a visual, a verbal and an (optional) sound channel, and all three enact narrative in the mimetic mode. This mode restricts the use of language to dialogue and subtitles, and severely limits the representations of counterfactual or purely mental events that are such an important part of the novel. In contrast to the theater, which relies mainly on what I have called raw material supports (props, costumes, and the bodies of actors),7 TV and film have a distinctive technological component that allows the capture of moving images by a moveable recording device with variable focal length. This technology opened possibilities unknown to the theater, such as variations in point of view and quick changes of location. While 7

The theater does however present some technological dimensions: lighting systems, the machinery that allows quick changes of stage setting, or the systems that permitted the ‘Deus ex Machina’ effect in the Baroque age.


17

the high definition of the image and large screen of the cinema led to an emphasis on setting, travel, and physical action (an emphasis reinforced by the recent development of computer-generated special effects), TV remained limited by the relatively poor quality of its image and by the small size of its screen, features that prevented immersion and prompted McLuhan to call TV a cool medium. These properties predisposed TV to the traditional themes of the theater, namely domestic drama focusing on human interaction, rather than to the visually stunning subject matters favored by the more spectacle-oriented cinema. Thanks to the capability of the technology to broadcast signals twenty-four hours a day, seven days a week, these themes of human interaction stretched into longrunning serials, rather than being condensed into the two-hour action typical of movies and the theater. Because serial narratives require lots of plot to be written in a relatively short time, they employ large teams of scriptwriters, and they favor multiple parallel plot lines with intersecting courses (Thompson 2003:40; all further quotes from the same source). The best settings for long-running TV narratives are therefore public spaces traversed by many characters, such as hospitals, police stations, law firms and prisons (57). In a further contrast to the cinema, TV technology allows the instantaneous transmission of images, a property that led to unique narrative forms, such as the reality show or the live sports broadcast. Many of the distinctive properties of TV narrative however are not due to its technological support exclusively, but to the cultural use of the medium. The habit of financing TV channels with advertising led to a segmented structure in which narrative shows alternate with commercial spots. Moreover, the demands of efficient programming put severe limits on the length of narrative episodes: half hour to one hour for serials, two hours for the broadcast of films made for the large screen. These culturallyimposed conditions put special demands on the script writers. Because the commercial breaks must come at logical stopping places (17), the episodes of one hour TV serials tend to be structured into four acts, each with its own crisis and climax (42). TV serials thus present the epic structure of a loose succession of relatively autonomous episodes on the macro-level, but they adhere to an Aristotelian dramaturgy, with well-defined story arcs, on the micro-level. Because spectators may turn on the TV set at any time during a show—in contrast to film and theater spectators, who must attend the performance from beginning to end—frequent recaps of the action are necessary, and they must be smoothly integrated into the dialogue (63). These narrative features are culturally conditioned, since they facilitate the integration of TV into daily life, but this integration is itself made possible by the continuous stream of information sent to the home by the supporting technology.

18

Marie-Laure Ryan

7. Defining Media from the Perspective of Transmedial Narratology For students of narrative, what counts as a medium is a category that truly makes a difference as to what stories can be evoked or told, how they are presented, why they are communicated, and how they are experienced. Narrative differences may concern three different semiotic domains: semantics, syntax, and pragmatics. In narrative theory, semantics becomes the study of plot, or story; syntax becomes the study of discourse, or narrative techniques; and pragmatics becomes the study of the uses of narrative. On the semantic level, different media favor different variations of the basic cognitive template: for instance, film prefers dramatic narratives shaped according to the Freytag triangle, TV prefers episodic narratives with multiple plot lines, and computer games prefer quest narratives with one plot line, but several autonomous episodes corresponding to the levels to be passed. On the discourse level, media may produce different ways to present stories, which will necessitate different interpretive strategies on the part of the user. For instance, narrative is broken up into distinct frames in comic strips, while it is represented by an image that seems to evolve continuously in film, at least until the next camera take. On the pragmatic level, finally, different media may offer different modes of user involvement and different ‘things to do’ with narrative. An example of these new things to do is the the posting of private diaries on the Internet, a phenomena known as ‘blogging’. With digital media, it is now possible to share narratives of personal experience with millions of strangers. In summary, a medium will be considered narratively relevant if it makes an impact on at least one of these domains. This approach implies a standard of comparison: to say for instance that ‘radio is a distinct narrative medium’ means that radio as a medium offers different narrative possibilities than television, film, or oral conversation (Ryan 2004:18). ‘Mediality’ (or mediumhood) is thus a relational rather than an absolute property. Moreover, the relevance of a given technology for narrative may be subject to change, as people put it to new uses. A case in point is the global positioning system (GPS), a technology that allows people to identify with great precision the latitude and longitude of their current location. Narrative applications of GPS may appear today as unlikely as did the literary use of number-crunching machines fifty years ago, but both ideas have come to pass: computers have notoriously become ‘poetry machines’, and GPS technology has been used to anchor stories in real-world locations. A case in point is 34 North 118 West, a project developed by Jeff Knowlton, Naomi Spellman and Jeremy Hight. In this project, users walk around a district of Los Angeles with a lap-top computer equipped with GPS technology. When they enter certain loca-


19

tions, the GPS sends signals that prompt the playing of stories through earphones. Walking around the area thus becomes a treasure hunt for the hidden stories that capture the spirit of the place. I won’t go as far as saying that GPS technology has become a narrative medium, because its role is (so far?) limited to monitoring the transmission of ready-made texts, but its use in 34 North 118 West certainly allows a new thing to be done with narratives, namely linking them to their real-world referent. This example suggests that transmedial narratology should not only pay attention to established and productive narrative media, but also to any kind of technology that affects the narrative experience. For a type of information support to qualify as a narrative medium, it must not only make a difference in the areas of story, discourse or pragmatics, it should also present a unique combination of features. The following list presents an overview of the kind of media features that affect the experience of narrative, besides the semiotic, technological and cultural factors already discussed: – Spatio-temporal extension. Media fall into three broad categories: purely temporal ones, supported by language or music exclusively; purely spatial media, such as painting and photography; and spatio-temporal media, such as the cinema, dance, image-language combinations, and digital texts. One could however argue that oral storytelling and print narrative involve a visual, and consequently spatial component; this would leave only long-distance oral communication such as radio and telephone as a language-supported example of the purely temporal category. – Kinetic properties. A spatio-temporal medium can be static (i.e. combinations of still pictures and text) or dynamic (moving pictures, or media relying on the human body as means of expression, such as dance or the theater). By contrast, all purely temporal media are dynamic, and all purely spatial media are static. – Number of channels. In the spatial category there is only one channel, unless one considers that sculpture and architecture have a tactile dimension. In the temporal category we have either one-channel media (language or music), or combinations of the two temporal media, language and music, in songs and sung forms of poetry. Most media of the spatio-temporal category have multiple channels, but mime and silent pictures use only visual data. Combinations include: image-language (in illustrated books); image-music (in silent films); or image-musiclanguage (in film, the opera, and digital media). – Priority of sensory channels. Thus the opera should be considered distinct from a theater production that makes use of music, even though

20

Marie-Laure Ryan

the two media include the same sensory dimensions and semiotic codes, because the opera gives the sound channel higher priority than the theater. Another issue to resolve for the definition of medium from the perspective of transmedial narratology is the delimitation of medium with respect to genre. Both medium and genre exercise constraints on what kinds of stories can be told, but while genre is defined by more or less freely adopted conventions chosen for both personal and cultural reasons, medium imposes its possibilities and limitations on the user. It is true that we choose both the genre and the medium we work in. But we select media for their affordances, and we work around their limitations, trying to overcome them or to make them irrelevant. For instance, painters introduced perspective to add a third dimension to the flat canvas. Genre by contrast deliberately uses limitations to optimise expression, to channel expectations, and to facilitate communication: for instance, tragedy must be about the downfall of a hero and use the mimetic mode of narrativity; and sonnets must consist of fourteen lines, organized in two quatrains and two tercets with a certain rhyming pattern. These conventions are imposed as what Jurij Lotman has called a second-order semiotic system on the primary mode of signification. Genre conventions are genuine rules specified by humans, whereas the constraints and possibilities offered by media are dictated by their material substance and mode of encoding (Ryan 2004:19). It is not always easy to distinguish genre from medium; but I would like to suggest a criterion based on the old question: what comes first—the chicken or the egg? Let the text be the chicken and genres and media be the egg. With genres, the chicken comes first. Genres originate in innovative texts that create a desire to duplicate their generative formula. With media on the other hand the egg comes first, since a text can only come into existence when the material support for its signs and the technology for their transmission are already in place. Insofar as media are sets of affordances, or possibilities, new media give birth to new forms of text and to new forms of narrative, which in turn may be codified into genres. Far from being narrowly conditioned by the properties of their supporting medium, narratives have developed various ways to handle these properties. First, narrative may go with its medium and take full advantages of its affordances. This would be the case of a computer game in which players manipulate an avatar and create its destiny through their actions. Computer games are so dependent on the interactive nature of their digital environment that they cannot be taken out of the computer. Second, narrative can ignore the idiosyncrasies of the medium and use it purely as a transmission channel. This is what happened when Stephen


21

King posted one of his novels on the Internet. The text was meant to be printed, and it took no advantage whatsoever of the artistic resources of a digital support. Third, narrative can actively fight some of the properties of the medium for expressive purposes. For instance, a print narrative with multiple branches subverts the linear reading protocols typical of novels distributed in book form, and it anticipates the possibilities of electronic textuality. Here we can say that the text yearns for another medium. And fourth, a text may expose latent properties in its medium, properties that expand its expressive potential beyond current practice. This happened when postmodern print novels began playing with graphic layout, and made us aware of the spatiality of the print medium, a spatiality that is forgotten when print is considered to be nothing more than the translation of spoken language. The diversity of games that narrative can play with the resources of its medium is one of the many reasons that make the intersection of narratology and media studies, an area still largely unexplored, into a productive field of investigation. References Abbott 2002 Abbott, H. Porter: The Cambridge Introduction to Narrative. Cambridge 2002. Aristotle 1996 Aristotle: Poetics. Translation and introduction by Malcolm Heath. London 1996. Barthes 1977 Barthes, Roland: “Introduction to the Structural Analysis of Narrative.” In: Barthes, Roland: Image-Music-Text. Translated by Stephen Heath. New York 1977. Bordwell / Thompson 1990 Bordwell, David / Thompson, Kristin: Film Art. 3rd ed. New York 1990. Bremond 1973 Bremond, Claude: Logique du récit. Paris 1973. Bruner 1991 Bruner, Jerome. “The Narrative Construction of Reality.” In: Critical Inquiry 18.1, 1991:1–22. Chatman 1978 Chatman, Seymour: Story and Discourse. Narrative Structure in Fiction and Film. Ithaca, NY 1978. Chatman 1990 Chatman, Seymour: Coming to Terms. The Rhetoric of Narrative in Fiction and Film. Ithaca, NY 1990. Dawkins 1976 Dawkins, Richard: The Selfish Gene. New York 1976.

22

Marie-Laure Ryan

Don 1990 Don, Abbe: “Narrative and the Interface.” In: Laurel, Brenda (ed.): The Art of Computer Interface. Redding, Mass. 1990:383–391 Fludernik 1996 Fludernik, Monika: Towards a ‘Natural’ Narratology. London 1996. Gaudreault / Jost 1990 Gaudreault, André / Jost, François: Le récit cinématographique. Paris 1990. Genette 1972 Genette, Gérard: Figure III. Paris 1972. Genette 1988 Genette, Gérard: Narrative Discourse Revisited. Translated by Jane Lewin. Ithaca, NY 1988. Heise 2002 Heise, Ursula K.: “Unnatural Ecologies: The Metaphor of the Environment in Media Theory.” In: Configurations 10.1, 2002:149–168. Jahn 2003 Jahn, Manfred: “‘Awake! Open Your Eyes!’ The Cognitive Logic of External and Internal Stories.” In: Herman, David: Narrative Theory and the Cognitive Sciences. Stanford 2003:195–213. Knowlton / Spellman / Hight (no date) Knowlton, Jeff / Spellman, Naomi / Hight, Jeremy: 34 North 118 West. [Accessed 17 July 2004] Lessing 1984 Lessing, Gotthold Ephraim: Laocoon: An Essay on the Limits of Painting and Poetry. Translation and introduction by Edward Allen McCormick. Baltimore 1984. Lyotard 1984 Lyotard, Jean-François: The Postmodern Condition: A Report on Knowledge. Translated by Geoff Bennington and Brian Massumi. Minneapolis 1984. McLuhan / McLuhan / Zingrobe 1996 McLuhan, Marshall / McLuhan, Eric / Zingrobe, Frank: Essential McLuhan. New York 1996 Metz 1974 Metz, Christian: Film Language. A Semiotics of the Cinema. Translated by Michael Taylor. New York 1974. Ong 1982 Ong, Walter J.: Orality and Literacy. The Technologizing of the Word. London 1982. Pratt 1977 Pratt, Mary Louise: Toward a Speech Act Theory of Literary Discourse. Bloomington 1977. Prince 1982 Prince, Gerald: Narratology: The Forms and Functions of Narrative. The Hague 1982. Prince 1987 Prince, Gerald: Dictionary of Narratology. Lincoln, Neb. 1987. Richardson 2000 Richardson, Brian: “Recent Concepts of Narrative and the Narratives of Narrative Theory.” In: Style 34.2, 2000:168–175.


23

Rimmon-Kenan 1989 Rimmon-Kenan, Shlomith: “How the Model Neglects the Medium: Linguistics, Language, and the Crisis of Narratology.” In: The Journal of Narrrative Technique 19.1, 1989:157–166. Ryan 2004 Ryan, Marie-Laure: “Introduction.” In: Ryan, Marie-Laure (ed.): Narrative Across Media: The Languages of Storytelling. Lincoln, Neb. 2004:1–40. Sutton-Smith 1997 Sutton-Smith, Brian: The Ambiguity of Play. Cambridge, Mass. 1997. Thompson 2003 Thompson, Kristin: Storytelling in Film and Television. Cambridge, Mass. 2003. White 1981 White, Hayden: “The Value of Narrativity in the Representation of History.” In: Mitchell, W.J.T. (ed.): On Narrative. Chicago 1981:1–23. Wolf 2002 Wolf, Werner: “Das Problem der Narrativität in Literatur, bildender Kunst und Musik: Ein Beitrag zu einer intermedialen Erzähltheorie.” In: Nünning, Vera / Nünning, Ansgar (eds.): Erzähltheorie transgenerisch, intermedial, interdisziplinär. Trier 2002:23–104. Wolf 2003 “Narrative and Narrativity: A Narratological Reconceptualization and its Applicability to the Visual Arts.” In: Word and Image 19.3, 2003: 180–197. Worth 1981 Worth, Sol: “Pictures Can’t Say Ain’t.” In: Gross, Larry (ed.): Studies in Visual Communication. Philadelphia 1981.

The Double Dynamics of Focalization in W. G. Sebald’s The Rings of Saturn Silke Horstkotte (Leipzig) It is frequently observed that the twentieth century has been a century of images.1 The omnipresence of images in newspapers and magazines, advertisements, film, and video, and on TV, computers, and the Internet has led scholars such as Barbara Stafford (1994) and Martin Jay (1996) to conclude that Western culture is currently undergoing a shift in which the visual medium, traditionally thought to be playing a secondary role as the illustration of text, is emerging as the dominant paradigm of thought. Associated with the new omnipresence of images are new ways of thinking about images and new modes of cultural analysis, which are no longer predicated on the arbitrary and shifting sign relations associated with textuality, but rather on phenomena such as spectatorship and the gaze. This ‘iconic’ or ‘pictorial turn’ has been hailed as a major epistemological shift by Gottfried Boehm (1994) and W.J.T. Mitchell (1994:11–34). Mitchell’s and Boehm’s argument that images do not represent isolated intruders into a textual universe, but rather constitute a new paradigm of thought, may be extended by inquiring into the conception of culture implied by the iconic turn. What happens if we treat culture not as a text, but as a repository of images? Such an iconic conceptualization of culture would require a modified methodology of cultural analysis which would account more closely for phenomena of looking and spectatorship than the traditional methods of textual analysis have done. Furthermore, an iconic methodology of cultural analysis might yield a more interactive concept of the relation between cultural artifacts and their recipients, as the model of a communicative vision developed by Mieke Bal indicates (1996:255–288). Drawing on Norman Bryson’s distinction between the gaze (a mode of vision that denies its historical and embodied nature and thus objectifies the contemplated object) and the glance (an active, engaged and embodied vision), Bal argues for a mode of vision that grants agency to both par1

The research for this article was supported by a fellowship within the Postdoc program of the German Academic Exchange Service (DAAD).

26

Silke Horstkotte

ties—the embodied, looking subject, and the contemplated (cultural) object (cf. Bryson 1983, 1988, 1994). Such a mode of looking would have to involve the subject within the field of vision, for ‘the awareness of one’s own engagement in the act of looking entails the recognition that what one sees is a representation, not an objective reality’ (Bal 1996:264). On the other hand, granting agency to visual objects implies that images ‘can be construed as theoretical objects that ›theorize‹ cultural history’, and that images actively participate in cultural dialogue (Bal 1999:5–7). What could an iconic methodology of cultural analysis look like? While wholly agreeing that cultural theory needs to treat images more seriously, I do not find it particularly useful to simply replace the textual paradigm with a visual or iconic one, thereby exchanging one set of critical buzzwords for another. Such an exchange runs the danger of artificially separating the iconic from the textual paradigm. Yet the interaction between words and visual images is fraught with ambiguities, complexities and contradictions. The ways in which images function in Western culture are inextricably linked with verbal and textual structures and ways of thinking, just as textual analysis cannot do without an intimate knowledge of how images work. In the following analysis, I shall therefore draw on an existing model of textual analysis that can be modified to account for the interaction of textual culture with the new iconic paradigm: the model of narratology. Narratology offers the means to analyze visual as well as verbal works of art and is therefore eminently suited to an inquiry into the relation between visual and textual culture. The kind of narratology I am envisioning here would thus be an intermedial narratology which deals with the analysis of artifacts that explicitly combine verbal and visual discourses. The iconic paradigm need not be restricted to visual artifacts. In fact, literature is one paradigmatic medium in which the exchange between words and images is negotiated. The last fifteen years have seen the publication of an increasing number of textual artifacts that explicitly incorporate or evoke visual material. My research focuses on one such body of texts: literary narratives that use photographs—imaginary or narrated as well as graphically reproduced ones—in order to depict the Second World War and the memory processes involved in its cultural transmission. These are the works of W.G. Sebald, especially The Rings of Saturn (2002b; 1998), The Emigrants (1993; 1996) and Austerlitz (2001; 2002a), as well as László Márton’s Die schattige Hauptstraße (2003); Monika Maron’s Pawels Briefe (2001); Marcel Beyer, Spione (2000); Rachel Seiffert, The Dark Room (2001); Ulla Hahn, Unscharfe Bilder (2003); Stefan Wackwitz, Ein unsichtbares Land (2003); and the recent re-publication of Peter Henisch’s Die kleine Figur meines Vaters (2003), among others. These texts clearly demonstrate the impossibility of strictly separating verbal and visual as-

Double Dynamics of Focalization

27

pects, while at the same time highlighting the representational nature of images. All of them are characterized by highly sophisticated interactions between images and words, or pictorial and verbal discourses, in which images are not cast in the role of subordinate illustrations to the dominant verbal text, but are equal partners of verbality within an intermedial fabric. I have chosen to use the term ‘discourse’, rather than the more common ‘medium’, because images and words are not just different kinds of representation, they also imply separate cultural values and distinct codes, conventions and channels, as well as different sensory, cognitive and affective modes. Hence, this corpus of novels is symptomatic of a type of intermedial discourse that can no longer be adequately described from within the textual paradigm of culture, but which requires a modified, more visually oriented set of interpretive tools. One author who has played a key role for the new literary use of images is the late W. G. Sebald, whose travelogue The Rings of Saturn constitutes my object of analysis in this article. Typically for Sebald’s works, The Rings of Saturn include many graphically reproduced images and photographs which stand in an uneasy, often ambiguous relation with the surrounding verbal discourse. Sebald’s text raises numerous questions, all of which ultimately concern the complex relation of visual and verbal discourses. An analysis of The Rings of Saturn will therefore have to use concepts which—like their object—combine verbal and iconic paradigms. I am thinking here particularly of narratological theories of focalization. In the analysis of linguistic narrative, the concept of focalization is used to refer to the story represented, as opposed to the concept of narration, referring to the representation of a story. Thus, as Gérard Genette and especially Mieke Bal have pointed out, every linguistic narrative contains a focalizor ‘distinguished in function, not identity, from the narrator’ (1997: 163; cf. Genette 1972). Older terms for focalization are ‘perspective’ or ‘point of view’, but these terms are problematic because they do not clearly distinguish between the levels of story (focalization) and text (narration) and can therefore apply to narration as well as to focalization. Through its basis in the notion of perspective, focalization is associated with matters of vision; it is therefore a concept bridging textuality and visuality, and can be fruitfully employed to analyze artifacts that combine the visual and the verbal. Mieke Bal has pointed out that, unlike the older terms ‘point of view’ and ‘perspective’, focalization entails asking ‘not primarily, where the words come from and who speaks them, but what is being proposed for us to believe or see before us’ (1997:224). Thus, focalization refers to the ‘vision through which the elements [of a story] are presented’ rather than to ‘the identity of the voice that is verbalizing that vision’ (Bal 1997:143). Thus, Focalization can serve to conceptualize

28

Silke Horstkotte

the visual aspects inherent in all narrative. On the other hand, it may be used to develop novel approaches toward the interpretation of non-narrative visual artifacts, as Bal has convincingly argued (1997:161–170). Still, what I propose here is not a transmedial narratology in the sense of a set of universal, media-independent tools of interpretation, but rather a series of strategic combinations of models and concepts which differ for each object of analysis. In the following case study, Bryson’s distinction between the gaze and the glance is combined with a concept of focalization that has its roots in Bal’s theory of narratology, which is itself already strongly oriented toward the visual. Since my object of analysis is an intermedial narrative, it is no coincidence that my analytic model draws on authors who are themselves working in the area between visual and textual studies. Focalization is a useful tool towards a communicative conception of vision for several reasons. Firstly, focalization emphasizes the visual nature of reception processes, including the reception of literary works. Secondly, since focalization inevitably colors the reader’s perception of the narrative, it implies a dynamic relation between the artifact (‘what is being proposed’) and its recipients (‘for us to believe or see before us’), or—conversely—between analysis and practice. Furthermore, unlike the terms ‘points of view’ or ‘perspective’, which only indicate a structural position, ‘focalization’ can be used both as a verb (to focalize) and as a subjective noun (focalizor). Thus, the concept of focalization can be used to conceptualize the element of agency inherent in the reception process. Furthermore, focalization stresses the practice and processuality of narrative instead of conceptualizing narrative as a finished product. As an indirect proposal which colors the narrative and thus strongly influences the reader’s perception, focalization is also that aspect of the narrative which most closely ties in with the novel’s reception. Within the dynamic interaction between cultural object and contemplating subject, which can be understood as a relation of framing (Culler 1988; Bal 2002:133–173), a second dynamic can be situated. This concerns the relation between verbal and visual discourses, which mingle to produce the intermedial rhetoric which characterizes The Rings of Saturn. That second, intratextual dynamic may be understood as a performative arrangement, in which the verbal and the visual link up to form a topographic structure—the printed surface of the intermedial text. However, this arrangement is never a final product, since it results from a performative process which occurs at the reception level. Hence, the concept of the performative, as it is used here, implies a dynamics which occurs in time, while at the same time drawing on the spatial positioning of images and verbal text in the printed book. In this sense, performativity is not only a result of agency, but also an aesthetic effect.


29

The dynamic model appears to harbor more analytic potential than static concepts of image-text relations such as illustration or ekphrasis, which are premised on the secondary and subservient role of pictorial discourse (cf. Boehm 1995; Krieger 1992). Through the concept of focalization, the exhibitory strategies of Sebald’s intermedial ‘image-’ or ‘iconotext’ may be conceived of as active gestures, rather than fixed combinations of text and image (Mitchell 1994; Wagner 1995, 1996). Within the theory of intermediality developed by Peter Wagner, ‘iconotext’ refers to ‘an artefact in which the verbal and the visual signs mingle to produce rhetoric that depends on the co-presence of words and images’ (Wagner 1996:16). Thus, an iconotext does not simply combine two separate media, but is a truly intermedial fabric in which the two media are inseparably bound up with each other, even if they never blur or merge entirely. Instead of looking for distinct ‘messages’ dictated by each of the two media, the analysis of iconotexts would therefore have to ask what sort of functional features textual and pictorial discourse employ within the communication of cultural object and contemplating subject. In The Rings of Saturn, the respective functions of photography and verbal discourse are intimately bound up with the model of cultural memory developed by the anonymous homodiegetic narrator. Sebald’s book is a somewhat disjointed account, in ten chapters, of the narrator’s walking tour through Suffolk, on which the narrator is confronted ‘with the traces of destruction, reaching far back into the past, that were evident even in that remote place’ (RS 3 / 11).2 The narrator visits many ‘relics of an earlier age’ (RS 90 / 112) and meets numerous refugees, diasporic and exilic characters. As in the novels of Thomas Bernhard, one of the key influences in Sebald’s writing, the narrative agent is mainly a reflecting consciousness, rather than acting person. The narration takes place in retrospect and is shot through with references to the narrator’s present situation, and with recollections from his earlier past, as well as embedded, intradiegetic narratives detailing the history of the visited buildings and landmarks. A large number of these intradiegetic narratives concern the natural disasters that have plagued the region since times immemorial—towns swallowed by the sea (RS 155 / 187), the destruction of the environment—and the ensuing economic decline, which prompts the narrator to imagine himself ‘amidst the remains of our own civilization after its extinction in some future catastrophe’ (RS 237 / 282). This short quotation already indicates that the perspective of the narrator colors not only the extradiegetic narrative, but also the intradiegetic historical digressions. Thus, the internal focalizor of 2

RS = The Rings of Saturn. Page numbers refer to the English translation of The Rings of Saturn first, and to the German original second.

30

Silke Horstkotte

the represented stories is embedded within the function of the narrator, the agent responsible for the stories’ representation. The dominance of reflective and recollective parts of the narrative can be illustrated through a tentative statistic. Counting the sub-headings in the book’s table of contents, intradiegetic references to the destruction of the region account for about 12 % of the narrative, only slightly less than the story of the walking tour itself (with around 15 %). Approximately 18 % of the narrative is concerned with references to forgotten genocides and other mass disasters of history: the slaughter of 700,000 Serbs, Jews and Bosnians by the Croatian Ustasha (RS 97f / 119–121); the mass suicide of the Taipingis in 1853 (RS 141 / 169); the Chinese famine of the 1870s (RS 150 / 181); and the Irish Troubles (RS 214f / 255f), to name but a few. By far the largest part of the narrative—roughly 40%—consists of intertextual and other cultural references which have no direct connection with the narrator’s journey: references to Thomas Browne, especially to his essay on urn burial (RS 19–26 / 29–39); to the fiction of Jorge Luis Borges (RS 67–71 / 87–91); the biography of Joseph Conrad (RS 104–122 / 127–149); and to Rembrandt’s Anatomy Lesson of Dr. Tulp (RS 12–17 / 22–27), among other things. These references establish the main concern of the narrator: the construction of a texture of allusions to cultural history, to paintings (Rembrandt, Ruysdael) and to literary works (by Browne, Borges, Joseph Conrad, Goethe, Benjamin, Hölderlin and others), as well as references to Sebald’s other work, especially to The Emigrants (RS 86 / 106). The verbal narrative of The Rings of Saturn functions as a repository of cultural memory, and the narrator as its curator. His thoughts when reading the old log in the Sailors’ Reading room can be taken as programmatic in this respect: ‘Every time I decipher one of these entries I am astounded that a trail that has long since vanished from the air or the water remains visible here on the paper’ (RS 93 / 116). Like the old log book, The Rings of Saturn preserve the traces of a long-vanished past in visible form. This visible preservation of the past, in which the reproduced photographs act both as historical relics and as memory triggers, is premised on the idea of archaeological strata as a model for memory, which the narrator develops in interaction with Thomas Browne’s essay on urn burial (1658). According to this model, the past is not ‘another country’, but is co-present with the here and now: it is simply buried at a deeper layer. The Suffolk landscape which the narrator traverses is full of such material, visible memories: for instance, the ruined or half-ruined manor houses; cemeteries; or the wells of the vanished town Dunwich, which are still visible at low tide at the bottom of the sea (RS 155–157 / 187–189). The concept of memory developed by the narrator is topographical rather than temporal in orientation and relies on memorial objects as


31

triggers for individual recollection. In the introduction to a volume of essays about material memory, Marius Kwint has convincingly argued that ‘objects form records: analogues to living memory, storing information beyond individual experience. Entering us through the senses, they become history, like the fragments that speak to the palaeontologist or geologist’ (Kwint 1999:2). But it is not only single objects that store, trigger, and shape the cultural memory The Rings of Saturn seek to preserve. Taken together, the individual objects constitute a topography of memory which is physically present in the Suffolk landscape, and which is also mirrored in the spatial arrangement of photography and verbal discourse within the printed text of the novel. By ‘spatial arrangement’, I do not refer to Joseph Frank’s ‘spatial form’ (1963), which dealt with the space-logic in modernist literature, but to the actual, visible arrangement of text and image in the printed book—the second part of the double dynamic described above, which forms the basis of the novel’s performative interaction with its readers and spectators. The images in Rings of Saturn should not be confused with actual photographs, since the latter are characterized by social uses, visual and material qualities quite distinct from those of photographic reproductions. Reproductions of the kind Sebald employs are copied from material photographs in a process called ‘screening’, which leads to substantial changes in the visual appearance of the images. For instance, screening accounts for the bad quality of many images in The Rings of Saturn (e.g. the photographs on pp. 97 / 120 and 103 / 125). Furthermore, the quality of the reproductions differs considerably between the various available editions of Sebald’s book. However, despite the significant differences between material and reproduced photographs, the images in The Rings of Saturn elicit many of the responses and affective reactions that are characteristic for interactions with actual photos. Nevertheless, it is important to keep in mind that the vision presented in the photos, i.e. their focalization, may be significantly influenced by changes in the photos’ reproduction. Moreover, while the reception of material photographs is premised on the mechanism of the ‘reality effect’, Sebald’s imaginative integration of photographic reproductions with verbal discourses foregrounds the representational nature of images.3 With the exception of Bal’s work, narratological theory has by and large neglected visual, spatial, and topographical structures. Werner Wolf expresses a broad consensus among narratologists when he describes 3

On the ideology of the ‘reality effect’, cf. Barthes (1982). Van Alphen (1997) has discussed the implications of this term for photographs, particularly in the context of discourses of memory and trauma. Cf. also my article on Sebald’s The Emigrants (2002).

32

Silke Horstkotte

narrating, the basis on which narrativity rests, as the representation of rudiments of an imaginable world, in which at least two different actions or states are centered on the same anthropomorphic characters, standing in a potentially meaningful, but not a necessary relation to each other (Wolf 2002:51). Even though Wolf points out that the basic units of a narrative have to be connected through more than mere chronology—i.e. that the connection between two events has to be motivated—chronology remains at the basis of his argument. Wolf concedes that images receive temporality through their reception process, which occurs in time (Wolf 2002:54), yet he assigns a low degree of narrativity to images, arguing that different media participate in narrativity to varying degrees. He then posits a ‘scale of narrativity’ characterized by the successively decreasing ratio of narremes (the constituent factors of narrativity inherent in a work of art, especially in a work of literature) to a narrativisation practiced by the recipient, which is characteristic of the reception of images (Wolf 2002:95f). The argument that poetry is temporal, and the graphic arts spatial, has a venerable tradition, dating back to Lessing’s Laocoon and beyond. However, W.J.T. Mitchell has persuasively argued that ‘works of art, like all other objects of human experience, are structures in space-time, and that the interesting problem is to comprehend a particular spatial-temporal construction, not to label it as temporal or spatial’ (Mitchell 1986:103). From a narratological point of view, whether literary texts can be described as spatial depends on which level of the narrative structure we focus on: the level of text, story, or fabula (Bal 1997).4 An analysis of Sebald’s iconotext has to concentrate on the layer of text, since this is the structural level occupied with representation, on which the photographs are situated and where visuality comes into play. It is also the surface on which the performative interaction with the reader / spectator occurs, while story and fabula are not part of the printed text, but constitute constructions of the reader. Temporality is an important element of the fabula (duration) and an integral aspect of the story (sequential ordering), but plays a subordinate role on the level of text, where it is restricted to the reception, which occurs in time. Space, on the other hand, is of supreme importance to the analysis of text, since it is on the level of text that the spatial arrangement of images and verbal discourse is situated.5

4

5

Bal’s model is more useful for my analysis than Genette’s (1980), since she conceptualizes the layer of representation as text (aesthetic product), rather than narration (enunciating act). Moreover, Bal treats focalization, which is my main interest in this article, more concisely than Genette does. On spatial structures in postmodern literature cf. also Bal (2003).


33

The spatial arrangement of The Rings of Saturn is inspired by Thomas Browne’s “The Garden of Cyrus” (1658), which the narrator paraphrases at some length. In this essay, Browne described the ‘patterns which recur in the seemingly infinite diversity of forms’ (RS 19 / 31), paying special attention to the so-called ‘quincunx’, a pattern Browne saw repeated throughout animate and inanimate nature and which Sebald has reproduced in The Rings of Saturn (RS 20 / 31). Not only does the narrator set out to find these hidden patterns in the Suffolk landscape, but the novel as a whole constitutes such a spatial pattern through the ways in which photography and verbal text are arranged on the printed page. The arrangement of verbal discourse and photography transports a variety of meanings which have to be inferred by the reader / spectator, since the photographs are rarely explicitly mentioned in the verbal discourse. In Bal’s terms, the photos initiate a communication with the reader / spectator which situates the reader in a field of vision. The visual presence of the photographs compels the reader to actively engage with the novel’s topographic arrangement and to formulate an interpretation that integrates the verbal and visual discourses into a meaningful whole. In the reception process, the photos function as focalizing agents which color the reader’s reception of the verbal discourse and which therefore enter into a dialogue with the focalization of the verbal narrative. For instance, the photograph reproduced on p. 155 / 187 can be understood to represent the perspective of the narrator, which he describes immediately below the image: ‘It was as if I had been walking for hours before the tiled roofs of houses and the crest of a wooded hill gradually became defined.’ In narratological terms, the internal focalizor of this passage is embedded in the function of the homodiegetic narrator. Intriguingly, the photographic reproduction shows a wooded hill above a small scattering of houses in the background, suggesting that the focalizor of the photograph and that of the verbal narrative are identical. However, the verbal text does not make any explicit reference to the photograph. Hence, the relation between the pictorial and verbal parts of the iconotext, and thus the identification of the verbal and visual focalizors, rests entirely on the spatial proximity of image and verbal description. Combined with assumptions about the evidentiary character of photography, the photo’s positioning may be taken to indicate that the photograph functions as a visual illustration of the verbal text. Yet such a relation is nowhere made explicit in the verbal discourse and therefore constitutes an interpretive act of the reader / spectator. As a result, we must carefully distinguish between the visible iconotext of The Rings of Saturn, and the image-text relations constituted by the reader, which belong on the level of story.

34

Silke Horstkotte

Like the verbal discourse, the photographs reproduced in The Rings of Saturn are colored by focalization. Thus, we are dealing here with a double process of focalization, in which the performativity of the two focalizing agents of Rings of Saturn—the internal focalization embedded in the homodiegetic narrator, which is present in the creation of a fabric of allusions to literary and cultural history and in the interpretations the narrator gives to this history, and the complementary focalization of the photographs, which directs the ways in which the reader / spectator understands the verbal discourse—is inextricably linked with the interpretative agency of the reader / spectator. The distinction between focalizor and narrator is crucial in narrative. In visual art, however, it is not always possible to clearly discriminate between a narrative agent and the represented perspective. In an iconotext like The Rings of Saturn, this facilitates the identification of the visual focalizors with the focalizor of the verbal narrative, and, ultimately, with the narrator; especially since in photography, the photographer—the agent of the vision represented in the photo—cannot be represented within the photograph. In this, photos crucially differ from paintings, which can contain internal focalizors in the shape of characters looking into the represented space. This tendency to identify the two focalizors adds to the general difficulty of interpreting photographic images. Mitchell has convincingly argued that images are generally read differently from verbal texts. While text is read like a graduated thermometer, where every mark has a fixed value, a picture is normally ‘read’ in something like the way we read an ungraduated thermometer. Every mark, every modification, every curve or swelling of a line, every modification of texture or color is loaded with semantic potential. … The image is syntactically and semantically dense in that no mark may be isolated as a unique, distinctive character (like a letter of an alphabet), nor can it be assigned a unique reference or ‘compliant’ (Mitchell 1986:67).

At the same time, pictures acquire meaning to a large extent through the verbal discourses framing them, and through their usage in actual interaction. Photographs in particular are necessarily incomplete and context-dependant. Victor Burgin (1986:58) has therefore concluded that photography is not a purely visual medium but ‘a complex of exchanges between the verbal and the visual.’ In The Rings of Saturn, as I have indicated, these exchanges are premised on the spatial arrangement of verbal and visual discourses. Like the semantic potential of the image itself, the arrangement of visual and verbal discourses is also dense and polyvalent. In the passage preceding the reproduction of Roger Casement’s signature (RS 134 / 162), the spatial positioning of pictorial and verbal discourses strongly influences the semantic


35

operation of the iconotext. Casement’s biography represents one of the many intradiegetic narratives intruding into the narrator’s account of his walking tour. Intriguingly, the narrator’s interest in Casement stems from a documentary film about Casement and Joseph Conrad which he missed almost completely when he fell asleep in front of the TV. The narrator then sets out to reconstruct this unseen film through a narrative account of Casement’s and Conrad’s lives. Roger Casement, a member of the Anglo-Irish ascendancy, met Joseph Conrad in the Congo, where Casement served as a British consul. He first became notorious for a report he wrote about the exploitation of the black population of the Congo under Belgian colonial rule—another historical catastrophe that the narrator tries to wrest from the threshold of cultural amnesia by including it in his narrative. Later, Casement became an advocate for Irish Home Rule; the Irish Troubles, too, are a recurring issue throughout The Rings of Saturn. In 1915, Casement traveled to Germany on a secret mission, determined to secure German help for the planned Dublin Easter Rising. However, his attempt remained unsuccessful, and Casement was returned to Ireland. At the time of the Easter Rising, he was already imprisoned in the Tower of London awaiting trial. Casement’s fate is intimately bound up with questions of mediation, and of authenticity and forgery, making it paradigmatic for the function of verbal and visual discourses in The Rings of Saturn. For Casement’s trial ultimately did not concern his participation in the planning of the Easter Rising, but rested on the contested authenticity of his diary: In order to preempt any petitions for pardon that might have been made by persons of influence, excerpts from what was known as the Black diary, a kind of chronicle of the accused’s homosexual relations found when Casement’s home was searched, were forwarded to the King of England, the President of the United States, and the Pope. The authenticity of this Black Diary, kept until recently under lock and key at the Public Records Office in Kew, was long considered highly debatable … But since the release to general scrutiny of the diaries in early 1994 there has no longer been any question that they are in Casement’s own hand (RS 131–34 / 159–62).

This passage of the verbal narrative is interrupted by the two-page reproduction of a handwritten diary page, presumably from Casement’s diary—since The Rings of Saturn do not include a list of illustrations, the reader / spectator cannot be entirely certain of this. This diary page is highly suggestive for my line of inquiry because it subverts the opposition between visual and verbal discourses and thus clearly indicates that images are just as much part of the text as the verbal narrative. After all, the diary page is both a visual reproduction and a verbal text. As an autographic source, it also represents a more embodied form of writing than the conventional

36

Silke Horstkotte

typesetting of the verbal narrative.6 Such an embodied writing has consequences for the reception process. While the modern norm of silent reading has come to be regarded as a purely cognitive act, the deciphering of handwriting is much more closely associated with vision and looking, hence, with activities that involve not only the mind, but also the body of the recipient. The reproduction of Casement’s handwriting therefore draws attention to the visual and embodied nature of all reception processes. Equally suggestive as the diary page itself is its relationship to the reproduction of Casement’s signature one page later (RS 134 / 162). First of all, the signature seems to confirm that the Black Diary was indeed written by Casement himself. Immediately following the reproduced diary pages, the signature may be understood as a signature to the diary—even though there is a passage of intervening verbal discourse between the two reproductions, as well as a long time gap in the dates (the diary is dated 1903, the signature, 1916). Secondly, the presumed relation between Casement’s diary and his signature raises questions about notions of authorship and authority, as well as about the relation between the autographic and the allographic, which concern the connection between the two facsimile reproductions and the verbal narrative. The establishment of such a connection rests entirely on the specific positioning of the verbal and iconic texts. The crucial importance of this spatial arrangement becomes obvious when one compares the German and English editions of The Rings of Saturn. In the German original, Casement’s signature is inserted into the verbal text directly after the last sentence quoted above. In the English translation, however, the signature has been moved to the end of the paragraph, which also concludes the entire chapter. The changed positioning of the signature is significant because it invites a different reading of the preceding narrative. By placing the signature at the end of the chapter, it comes to resemble the signature under a letter or document, thereby bestowing a high degree of authority upon the preceding account of Casement’s life, his trial, sentencing and execution: the authority of the writer’s signature. Thus, the placement of the signature in the English translation might be taken to suggest that Casement himself has authorized the preceding account of his life. The differences between the German original and the English translation point to a fundamental problem when analyzing iconotexts: it is not always clear whether the arrangement of verbal and visual discourses is part of the text (and thus controlled by the narrator), or if it touches

6

The aesthetic implications of handwriting have been discussed by Sonja Neef (2000).


37

upon the paratextual, which lies outside the scope of the narrator, since it concerns the material production of the book, upon which even the writer may have limited influence. Obviously, this is especially problematic in the case of the translation. One result of this ambiguity concerning the positioning of images is that the borders of the iconotext dissolve, since there is no overarching authority guaranteeing its unity of meaning. However, if the unity of the text is put under erasure, the text itself can no longer be clearly distinguished from its dynamic interaction with the reader / spectator. The third semantic field implied by Casement’s signature (both through its positioning and through the inclusion of the date) is that of a tombstone or epitaph—thereby highlighting, once again, the narrative’s function as a material repository of memory. And finally, since the last sentence of the chapter mentions the exhumation of Casement’s remains in 1965, the positioning of the signature establishes a strong connection between Casement’s body of writing and his mortal body. Thus, the inclusion of photographs stresses the spatial and visual dimensions of the text, which tend to be overlooked in narratological analyses, while simultaneously foregrounding the interpretive agency of the reader / spectator. Specifically, the photographic reproduction of handwritten sources draws attention to the embodied nature of writing, which engages the whole body of the reader / spectator by freeing it from the constraints of a purely cognitive conception of reading. Although spatial structures are most obvious in the arrangement of visual and verbal discourses, they also play a prominent role within the verbal narrative. The narrator’s journey does not adhere to the structural model of a sequence of events in time, but is shaped as a topography of places visited. Even the various historical digressions are connected through analogy and association rather than through temporal causality. The analogy between the Belgian colonial rule in the Congo and the Irish struggle for Home Rule, for instance, is predicated on an associative logic which subordinates chronology to a memorial topography constructed by the narrator, and which is therefore closely related to the intermedial arrangement of the iconotext. If The Rings of Saturn is a text about journeys and space, it is also a meta-topographic text that exposes the technique of iconotextual arranging as the structural principle of its narrative. The narrator’s dream about the maze at Somerleyton which he believes ‘had been created solely for me’ and which ‘represented a cross-section of my brain’ can be read as a self-conscious reference to the spatial arrangement of the iconotext (RS 173 / 206). Topographical thinking is also responsible for the narratorial strategy of ‘going the wrong way’ (RS 84 / 105), and for the recurring re-

38

Silke Horstkotte

presentations of exterritorial and fantastic space. One such fantastic place is Somerleyton palace, which not only demonstrates the spatial structure of time, but at the same time highlights the unreality of space itself, since it seems to open up into several incommensurable dimensions at once: There are indeed moments, as one passes through the rooms … at Somerleyton, when one is not quite sure whether one is in a country house in Suffolk or some kind of no-man’s-land, on the shores of the Arctic Ocean or in the heart of the dark continent. Nor can one readily say which decade or century it is, for many ages are superimposed here and coexist (RS 36 / 49).

Here, both time and space appear to be composed of numerous layers of reality, and indeed, the picture of Somerleyton reproduced in the novel looks less like an English country house than like an utterly fantastical Kubla Khan’s palace (RS 34 / 47). Following Tzvetan Todorov’s famous definition of the fantastic (1973), the superimposition of many times and spaces may be understood as a fantastic structure. Stanley Cavell has summed up Todorov’s argument as follows: the fantastic is linked to ‘such ideas as that of the imaginary journey, especially in quest of the self; and such ideas as that of being on some boundary or threshold, as between the impossible and the possible; and ideas of the confrontation of otherness; and of some adverse relation to the modern scientific sensibility’ (Cavell 1988:183). The narrative of The Rings of Saturn confirms this pattern in a number of ways. The foregrounding of topographic structures and the association with a wide range of historical events lends a strongly imaginary quality to the narrator’s journey; indeed, the quotation comparing the maze at Somerleyton to the narrator’s own brain reveals his journey as a ‘quest of the self’. Boundaries and thresholds play a central part in this strange internal journey, as does the confrontation of otherness. Many of the places the narrator visits serve as thresholds into another, or indeed into numerous other, dimensions. The neglected and dust-covered Ashbury House in Ireland, which the narrator had visited on a former occasion, is another example of fantastic space. The sad state of the Ashbury mansion is the result of one of the many historical catastrophes touched upon in The Rings of Saturn: the Irish civil war, which, while freeing the country from British colonial rule, also dealt the deathblow to an entire social class. The Ashburys are some of the last members of the former Anglo-Irish ascendancy still living in their Big House. This gives them a ghostly and strangely exterritorial appearance, and turns their half-ruined mansion into a fantastic threshold that also threatens to dissolve the narrator’s consciousness: ‘Whenever I rested on that bed over the next few days, my consciousness began to dissolve at the edges, so that at times I could hardly have said how I had


39

got there or indeed where I was’ (RS 210 / 250). The imaginary dissolution of time and space then leads the narrator to fantasize about yet another terrible site of destruction: ‘what I saw in my mind’s eye was … a battlefield somewhere in Lombardy over which the vultures circled, and, all around, a country laid waste by war’ (ibid.). However, the intrusion of exterritorial states is not restricted to the narrator’s consciousness; for instance, Mrs. Ashbury also mentions the ‘feeling’, after her marriage, that she was ‘completely out of this world’ in Ashbury House (RS 214 / 255). Moreover, the Ashbury house is linked not only with exterritorial, but also with supernatural modes of making sense of experience: because nobody wants (or can afford) to buy the mansion, the surviving Ashburys remain ‘tied to it, like damned souls to their place’ (RS 219f / 262). The motive of exterritorial and fantastic space introduces a mythical dimension into the narrative—a historical depth that is laden with symbolism. The Suffolk landscape acquires depth not only because it is composed of archaeological strata, but also through this fantastic double-structure, which rests almost entirely on the subjective perception of the narrator. The juxtaposition of disasters from different regions and eras of history indicates that the narrator searches not only for specific historical relics, but also for ‘fundamental patterns of human behaviour’ (RS 159 / 191)—the cultural equivalent of Browne’s quincunx. The topography of The Rings of Saturn is a memorial and imaginative construction, rather than a realistic representation of space. The fantastic dimension of the narrative dissolves not only the borders between various historical periods and geographies, but also the thresholds of time and space. Traveling emerges as a movement which challenges the distinction between space and time. The narrator’s journey is marked by fantastic disjunctions of time: he uses a guidebook ‘published shortly after the turn of the [twentieth] century’ (RS 42 / 57) and encounters people ‘dressed in the style of the Thirties’ (RS 43 / 57f). Frequently, the narrator even has the impression that the people he meets belong to a different time and space altogether, as in this description of Lowestoft beach: ‘It is as if the last stragglers of some nomadic people had settled there, at the outermost limit of the earth, in expectation of the miracle longed for since time immemorial, the miracle which would justify all their erstwhile privations and wanderings’ (RS 51f / 68). This is no more than a figment of the narrator’s imagination, for in the next sentence, he concedes: ‘In reality, however, these men camping out under the heavens have not traversed faraway lands and deserts to reach this strand. Rather, they are from the immediate neighbourhood …’ (RS 52 / 68). Nevertheless, the first image is so striking that the two explanations remain side by side—a distinctly fantastic structure: this undecidability, which hovers ambiguously between

40

Silke Horstkotte

rationalistic and supernatural modes of making sense of experience, is a key issue in Todorov’s definition of the fantastic. The undecidability of the fantastic is conspicuous in the narrator’s encounter with what initially appears to be a couple making love on the beach: ‘A couple lay down there, in the bottom of the pit, as I thought: a man stretched full length over another body of which nothing was visible but the legs, spread and angled’ (RS 68 / 88). As so often in The Rings of Saturn, this realistic impression quickly gives way to the narrator’s fantastic and uncanny associations: ‘In the startled moment when that image went through me, which lasted an eternity, it seemed as if the man’s feet twitched like those of one just hanged’ (ibid.). Even stranger than the association with a hanged man is the following comparison with a two-headed sea monster, which soon takes on a reality of its own in the eyes of the narrator. ‘Misshapen, like some great mollusc washed ashore, they lay there, to all appearances a single being, a many-limbed, two-headed monster that had drifted in from far out at sea …’ (ibid.). Characteristically for The Rings of Saturn, the narrator’s highly subjective and imaginative vision is connected with a catastrophic view of (natural) history, for this great beast is perceived to be ‘the last of a prodigious species, its life ebbing from it with each breath expired through its nostrils.’ At first, the misshapen beast is just a simile (‘like some great mollusc … they lay there’). As the narrative progresses, however, the rhetorical trope turns into a subjective reality, and the narrator becomes confused about what he has or has not seen. He seriously considers the possibility that he really saw a sea monster on the beach: ‘Before I reached the town, the first drops of rain were falling. I turned to look back down the deserted stretch I had come by, and could no longer have said whether I had really seen the pale sea monster at the foot of the Covehithe cliffs or whether I had imagined it’ (RS 69). Intriguingly, the initial sight of a couple making love on the beach has now been completely displaced by the fantastic vision of a sea monster. This progressive move away from rationalistic and towards supernatural and imaginative ways of making sense of experience is then linked to Borges’s story “Tlön, Uqbar, Orbis Tertius”, an important intertext for The Rings of Saturn, and to Borges’ narrator’s attempt to create secondary and tertiary worlds. Not only does the narrator of Rings of Saturn here explicitly align himself with a highly fantastic mode of storytelling, but also with a meta-fictional form of narrative that corresponds closely with his own strategies. The narrator’s walk on the Covehithe beach reveals focalization as a performative process that closely implicates the reader. The vision of the internal focalizor constantly evolves: from the description of a couple on the beach, through the comparison with a great sea beast, which is then


41

turned into a real sight, towards an utterly fantastic perspective in which both possibilities become equally strange. This alienating change of perspective highlights the strongly visual qualities of the verbal narrative. At the same time, it leads to the recognition that what the reader sees is a representation, not a reality. The meta-fictional and openly artificial aspects of the verbal narrative reflect on the status of the reproduced images, which resist conventional ways of seeing. There is nothing ‘authentic’ or ‘evidentiary’ about the photographs in The Rings of Saturn, with their unfocused, often grainy appearance, their maddeningly uncertain provenance and their highly suggestive, yet entirely implicit relationship with the verbal discourse. Instead of illustrating the verbal text and visually confirming its message, the photographs enable the reader / spectator to adopt a fantastic double vision that closely corresponds to the focalization embedded in the homodiegetic narrator. In this fantastic mode of seeing, time and space open up into multiple dimensions, words and images merge, and nothing is what it seems. The fantastic mode of storytelling, with the resulting, topographically structured iconotext, enables a new concept of focalization. This is a focalization which does not rigidly control the reader’s gaze, but allows it to remain mobile and to roam freely across the topographic terrain. The double focalization inherent in the combination of verbal and visual discourses always offers the reader several alternative choices how to interpret the relation between the different parts of the iconotext. At the same time, this topographic and fantastic principle enters into the verbal narrative itself, as in the passage on the Covehithe beach. The topography of the iconotext thus emerges as a field in which the reader / spectator can choose to follow different paths. With the mobile gaze, this mode of reception involves the whole body of the reader / spectator, not just his or her cognitive capacities. Characteristic of W.G. Sebald’s works, the topographical model developed in The Rings of Saturn offers an alternative to the problematic view that conceptualizes visual art in static opposition to verbal narrative and relegates the visual domain to the illustration of verbal texts. Instead, the Rings establish a communicative relation between embodied subject and contemplated object that is predicated on the interactive vision of the ‘glance’.

42

Silke Horstkotte

References Alphen 1997 Alphen, Ernst van: Caught by History: Holocaust Effects in Contemporary Art, Literature, and Theory. Stanford, CA 1997. Bal 1996 Bal, Mieke: Double Exposures: The Subject of Cultural Analysis. London / New York 1996. Bal 1997 Bal, Mieke: Narratology. Introduction to the Theory of Narrative. 2. ed. Toronto 1997. Bal 1999 Bal, Mieke: Quoting Caravaggio: Contemporary Art, Preposterous History. Chicago 1999. Bal 2002 Bal, Mieke: Travelling Concepts in the Humanities: A Rough Guide. Toronto 2002. Bal 2003 Bal, Mieke: “Critique of Voice: The Open Score of Her Face.” In: Pderi, N. (ed.): Travelling Concepts III: Memory, Narrative, Image. Amsterdam 2003:91–114. Barthes 1982 Barthes, Roland: “L’effet de réel.” In: Barthes, Roland: Litterature et réalité. Paris 1982. Beyer 2000 Beyer, Marcel: Spione. Köln 2000. Boehm 1994 Boehm, Gottfried: “Die Wiederkehr der Bilder.” In: Boehm, G. (ed.): Was ist ein Bild? München 1994:11–38. Boehm 1995 Boehm, Gottfried: “Bildbeschreibung. Über die Grenzen von Bild und Sprache.” In: Boehm, G. / Pfotenhauer, H. (eds.): Beschreibungskunst – Kunstbeschreibung. Ekphrasis von der Antike bis zur Gegenwart. München 1995: 23–40. Browne 1658 Browne, Sir Thomas: Hydriotaphia, urne-burial: or, A discourse of the sepulchrall urnes lately found in Norfolk = Together with The garden of Cyrus, or, The quincuniall, lozenge, or net-work plantations of the ancients, artificially, naturally, mystically considered/with sundry observations. London 1685. Facsim. ed, The Noel Douglas replicas. London 1927. Bryson 1983 Bryson, Norman: Vision and Painting. The Logic of the Gaze. London 1983. Bryson 1988 Bryson, Norman: “The Gaze in the Expanded Field.” In: Forster, H. (ed.): Vision and Visuality. San Francisco 1988:87–114. Bryson 1994 Bryson, Norman: “Art in Context.” In: Bal, M. / Boer, I. (eds.): The Point of Theory: Practices of Cultural Analysis. New York 1994:66–78.


43

Burgin 1986 Burgin, Victor: The End of Art Theory: Criticism and Post-Modernity. Atlantic Highlights, NJ 1986. Cavell, Stanley 1988 Cavell, Stanley: “The Fantastic of Philosophy.” In: Cavell, S. (ed.): In Quest of the Ordinary. Lines of Skepticism and Romanticism. Chicago 1988:181–188. Culler 1988 Culler, Jonathan: Framing the Sign: Criticism and Its Institutions. Norman, Okl. 1988. Frank 1963 Frank, Joseph: “Spatial Form in Modern Literature.” In: Frank, J.: The Widening Gyre: Crisis and Mastery in Modern Literature. New Brunswick, NJ 1945:3–62. Genette 1972 Genette, Gérard: Figures III. Paris 1972. Genette 1980 Genette, Gérard: Narrative Discourse: An Essay in Method. Translated by J.E. Lewin. Ithaca, NY 1980. Hahn 2003 Hahn, Ulla: Unscharfe Bilder. München 2003. Henisch 2003 Henisch, Peter: Die kleine Figur meines Vaters. Salzburg / Wien 2003. Horstkotte 2002 Horstkotte, Silke: “Pictorial and Verbal Discourse in W.G. Sebald’s The Emigrants.” In: Iowa Journal of Cultural Studies (2), 2002:33–50. Jay 1996 Jay, Martin: “Vision in Context: Reflections and Refractions.” In: Brennan, T./Jay, M. (eds.): Vision in Context: Historical and Contemporary Perspectives on Sight. New York 1996:1–12. Krieger 1992 Krieger, Murray: Ekphrasis. The Illusion of the Natural Sign. Baltimore 1992. Kwint, Marius 1999 Kwint, Marius: “Introduction: The Physical Past.” In: Kwint, M. / Breward, C. / Aynsley, J. (eds.): Material Memories. Oxford / New York 1999:1–16. Maron 2001 Maron, Monika: Pawels Briefe. Frankfurt / Main 2001. Márton 2003 Márton, Lászlo: Die schattige Hauptstraße. Wien 2003. Mitchell 1986 Mitchell, W.J.T.: Iconology. Image, Text, Ideology. Chicago 1986. Mitchell 1994 Mitchell, W.J.T.: Picture Theory: Essays on Verbal and Visual Representation. Chicago 1994. Neef 2000 Neef, Sonja: Kalligramme. Zur Medialität einer Schrift, anhand von Paul van Ostaijens “De feesten van angst en pijn”. Amsterdam 2000. Sebald 1993 Sebald, W.G.: Die Ausgewanderten. Vier lange Geschichten. Frankfurt / Main 1993.

44

Silke Horstkotte

Sebald 1996 Sebald, W.G.: The Emigrants. Translated by M. Hulse. London 1996. Sebald 1998 Sebald, W.G.: The Rings of Saturn. Translated by M. Hulse. London 1998. Sebald 2001 Sebald, W.G.: Austerlitz. München 2001. Sebald 2002a Sebald, W.G.: Austerlitz. Translated by A. Bell. London 2002. [Original edition: London 2001.] Sebald 2002b Sebald, W.G.: Die Ringe des Saturn. Eine englische Wallfahrt. 5. ed. Frankfurt / Main 2002. [Original edition: Frankfurt / Main 1995.] Seiffert 2001 Seiffert, Rachel: The Dark Room. London 2001. Stafford 1994 Stafford, Barbara: Artful Science, Enlightenment Entertainment and the Eclipse of Visual Education. Cambridge, MA 1994. Todorov 1973 Todorov, Tzvetan: The Fantastic: A Structural Approach to a Literary Genre. Translated by R. Howard. Ithaca, NY 1973. Wackwitz 2003 Wackwitz, Stefan: Ein unsichtbares Land. Familienroman. Frankfurt / Main 2003. Wagner 1995 Wagner, Peter: Reading Iconotexts: From Swift to the French Revolution. London 2003. Wagner 1996 Wagner, Peter: “Introduction: Ekphrasis, Iconotexts, and Intermediality – the State(s) of the Art(s).” In: Wagner, P. (ed.): Icons – Texts – Iconotext. Essays on Ekphrasis and Intermediality. Berlin / New York 1996:1–40. Wolf 2002 Wolf, Werner: “Das Problem der Narrativität in Literatur, bildender Kunst und Musik: Ein Beitrag zu einer intermedialen Erzähltheorie.” In: Nünning, V. / Nünning, A. (eds.): Erzähltheorie transgenerisch, intermedial, interdisziplinär. Trier 2002:23–104.

Levels of Play and Narration Britta Neitzel (Tampere)

1. Introduction Writing an article for this collection of conference papers puts me in an unfamiliar position. Most of the audiences that I have addressed in recent years have had a reasonable or better understanding of the concepts of story, narrative, and narration, but little knowledge of narratology, being familiar with computer games instead. I now find myself in the opposite situation—I can be certain that my readers know a lot about narratology, but I am not sure how familiar they are with computer games. When reading the abstracts for the papers presented at the conference, I was lulled into a false sense of security by the secondary literature cited in them. False, because I am not really a narratologist at all (I have always associated narratology with literary theory) but a student of media studies who uses narratology to analyse certain aspects of computer games. I would also describe myself as someone interested in the uses of narratology beyond literary theory and in the outer extremes of the range of phenomena covered by the concepts of narrative and story. Accordingly, this contribution uses narratology to analyze some aspects of computer games. Departing from a position in media studies that is interested in inter- or transmedial structures it will focus on the scope of a narratology applied outside the borders of literary studies, and my contribution will moreover address the question of the borders of narratives and stories themselves. The best way to observe and examine borders and borderline cases of narrative phenomena is to compare narratives with phenomena that do not normally have anything in common with them. Games are one such example. Both narratives and games have a long cultural heritage and a central function in the dynamics of culture (see, for example, Barthes 1983 and Huizinga 1950). Narration is a widespread and long-established framework for organizing the world in a certain way. Narratives recount actions. When telling a story, a narrator or author (I return to this distinction below) selects certain events that take place in the (or a) world and

46

Britta Neitzel

arranges them in ‘a chronological cause and effect chain’, as Bordwell (1985: 49) describes the story.1 When this act of organizing the world leads to a particular ending the narration makes sense (see Luhmann 1985). Play and games are based on another concept. To play is to perform certain actions to which are ascribed the attributes of freedom, ambivalence, relative unity, and a special temporal structure and relationship to reality, all of which differentiate them from other sequences of actions (see, for example, Scheuerl 1990, Huizinga 1950, and Caillois 1967). Most of these attributes can be assigned to play because of the features that separate them from the (rest of the) world. These features can take the form of spatial and temporal restrictions; rules that permit or prohibit certain activities; and, in some cases, fictive worlds shared by the players. What games do, we might say, is provide the conditions under which play is possible.2 Playing games, therefore, does not give order to the (or a) world; instead, it constructs new or different worlds in which actions can be performed. Authoring a game means putting in place the rules or conditions of such a world in which play can take place. Playing in itself does not lead anywhere (it does not, for example, progress towards an end); instead, it is self-sufficient and can take place again and again in endless cycles of repetition. Narratives on the one hand and play and games on the other present us with opposing concepts. The former concerns the recounting of actions, the latter the actions themselves. They also appear to have different chronological structures. Stories are based on linear chronology, whereas play has a circular structure.3 Computer games are, as the term suggests, games. Unlike other games, however, they depict the actions that are performed in them. They have a layer that presents the actions, and because of this they could potentially have narrative qualities. There is no doubt that applying narratological methods to computer games presents narratology with a challenge; to it this article responds. I begin by discussing some general aspects of the way (computer) games change narrative structures in the first section before considering some narratological models that can be applied to computer games despite the obvious differences between games and narratives. 1

2

3

Bordwell uses the term ‘fabula’ as introduced and used by the Russian Formalists. But ‘fabula’ and ‘story’ are synonymous in the respect I use them here. For a comparison of narratological terms see Neitzel (2000:59–75). Theorists who support an idealistic concept of play would question this statement because they consider play to be independent of games. Consideration of the debate within play theory concerning such topics is beyond the scope of this article. This is something of an oversimplification with respect to stories and play alike—it is possible for narratives to deviate from chronological order, and play can be directed in a certain direction by setting goals to be attained.

Levels of Play and Narration

47

In doing so my aim is not to provide an overall narratological analysis of computer games, but rather to expose some of the methodological points which I consider crucial to a narratological analysis of computer games. In the second section of this article, I use examples from real computer games to examine how narratological models can be applied to computer games despite the obvious differences between games and narratives. Thus, my overall aim will be to highlight some of the crucial areas where the narratological analysis of computer games ceases to be straightforward, the application of narratological methods should arguably be reconsidered, and other methodologies may well be more appropriate. But I also hope to show that, if we proceed carefully, some narratological terms and models can be applied to computer games in fruitful manner. 2. Computer Games and Narrative Levels Computer games affect each of the three levels that narratologists have identified in literary narratives. They are recapitulated in brief below. (1) The level of the discourse (discours) is ‘the narrative statement, the oral or written discourse that undertakes to tell of an event or a series of events’ (Genette 1980:25). Since most computer games now employ graphics, the discourse level has changed from one of textual to visual presentation.4 Such a change affects the temporal relations between story and discourse and has been widely discussed in film studies (see, for example, Bordwell 1985, Chatman 1978, Gunning 1991, Jost 1989, and Gaudreault 1984; 1989). (2) The level of narration (narration) is ‘the event that consists of someone recounting something: the act of narrating taken in itself’ (Genette 1980:25). This level is affected by the way in which the player takes part in a game. In computer games, the recipient (that is: the player) is presented not only with an interpretive task, as in the case of literary texts, but also with exploratory and configurational tasks (Aarseth 2001). As stated above, the author does not lay down a fixed sequence of events but provides in the program a range of possibilities on the basis of which certain events are selected by the player. (3) The participation of the player also affects the level of the story (histoire), which is ‘the succession of events, real or fictitious, that are the subjects of this discourse, and to their several relations of linking, opposition, repetition, etc.’ (Genette 1980:25), because the actual sequence of events is determined during the course of the game. 4

Text adventures (e.g. Zork) are examples of the use of textual presentation.

48

Britta Neitzel

In the remainder of this article, I ignore computer games that self-evidently lack a story (e.g. most sports games, flight simulators, and abstract games like Tetris). I concentrate instead on games that appear to be narrative because they include a protagonist who has to overcome certain obstacles. Even here, however, I concentrate not on the level of the story and the potentially different story structures but primarily on the discourse level and how it relates to the other levels. It is on the discourse level that the differences between narratives and games are most apparent. 3. The Level of Discourse in Non-Literal Narratives The visual layer which most modern computer games employ has important implications for the discourse level. It raises once more the distinction between showing and telling, which as been long been discussed since it was first introduced by Plato (1991). In the Politeia (Republic), Plato says of the dramatic mode (the mode of showing) that the narrator is concealed so that it seems that events are happening in the here and now. In the narrative mode, on the other hand, it is possible to identify a narrator who recounts events. We encounter this distinction and its associated problems again when dealing with film. A film shows pictures in which a spectator can see things happen. There is no overt narrator. Gaudreault (1984) introduced the term ‘monstration’ to reflect the impression that films nonetheless have the ability to narrate. The monstrator records the images of a film. In contrast to drama, on which earlier discussions were based, the events presented by a film do not happen while the film is being shown. They are recorded beforehand. Thus, according to Gaudreault, narration is also present in films. The narrator arranges the pictures after the monstrator has taken them. The editing performed by the narrator gives a film chronological freedom: it allows time-shift effects, the rearrangement of chronology, and omissions to occur. Faced with the same problem, Gunning (1991:19–21) developed another model to describe filmic narration. He introduced the concept of the narrator system, which has three levels. (1) The pro-filmic level contains everything that is exposed to the camera to be filmed. It includes such things as actors, lighting, set design, and the choice of setting. This is the level where basic decisions are made: ‘as narrative discourse the pro-filmic embodies a series of choices and reveals a narrative intention behind the choices’ (Gunning 1991:19).


49

(2) The level of the enframed image involves the composition of the image, the camerawork, and the resultant shots. (3) The editing level is where the shots are selected and arranged. Treating the narrating entity in a film as a system is sensible for two reasons. First, it accounts for the technical components of a film rather than referring to an anthropomorphic narrator who speaks from a particular point of view with a particular distance and voice. Second, it allows us to include all the technical levels of film in our analytic model. Even though the above system was developed for describing filmic narration, it also allows us to identify the differences between the kinds of narration found in different kinds of non-verbal media. A drama, for example, relies primarily on the first of Gunning’s levels: the content is acted out by the actors. Neither the second nor the third level is explicitly present, but they are in a sense subsumed under the first one: the frame of the stage has an effect analogous to that of enframing, and montage / editing can be performed by dividing the play into scenes and acts. 4. The Level of Discourse in Computer Games Like animated films, computer games do not use photographic images and because of that they lack a pro-filmic level.5 Scenes, actors, lighting are computer generated. The program creates a world in which certain characters can perform certain actions. It also determines the physical laws of this world as well as the point of view from which it can be perceived. Also the editing level is seldom found in computer games6: the diegetic actions take place while the game is being played and thus simultaneously with the non-diegetic actions performed by the player. Additionally, the diegetic actions and events are displayed in real time. Nothing happens before the game is played. What Tomashevsky (1965) would call the plot, the actions as they appear in the text, cannot be recorded before the game is played. In contrast to films, there is no ‘time lapse’ (Gaudreault) between what happens and the presentation of it. Even if the program defines a world, its laws, and a range of possible actions before the game is played, 5

6

In this respect, computer games resemble most verbal narratives. In both cases, the events that take place have no real antecedents (something that is the concern of semiotics rather than this article). Some computer game genres are an exception. Classic graphic adventures represent discrete rooms rather than continuous space. Time shifts can occur when moving between these rooms. Furthermore, cut scenes (prearranged scenes used mainly to link two levels and provide narrative information) use editing, although it should be noted that in such scenes the player is a spectator without the ability to play.

50

Britta Neitzel

actions that might have narrative qualities, do not take place before they are displayed on the monitor. Story, and discourse, including plot, appear simultaneously with the act of playing. The three levels of narrative can be easily distinguished in verbal narration, but in computer games they influence one another and are less clearly separated. To separate these levels, the story of a computer game must be conceptualized as a construction of a recipient,7 either after the game is played or before it is played. Thereafter, the three levels can be separated: The diegetic actions in the game world (the story actions), the audiovisual presentation of these actions (the plot actions) on discourse level and the production of these actions by the non-diegetic activities of a player. In verbal narration, the sequences of story and narration are always linked by a temporal relation. According to Genette (1980:215–217), we can distinguish ‘subsequent narration’ (an event occurs in the story before it appears in the narration), ‘prior narration’ (e.g. prophecies), ‘simultaneous narration’, and ‘interpolated narration’. The most frequent form is subsequent narration: something happens first and is narrated afterwards. Since in computer games the realization of the events in the story depends on the act of playing (or narration), narration in computer games is not a matter of simultaneous narration as found, for example, in live television broadcasts. We are dealing instead with interpolated narration, in which ‘the story and the narrating can become entangled in such a way that the latter has an effect on the former’ (Genette 1980:155). Computer games differ from novels with interpolated narration, however, in that the real actions that produce a computer game and leave traces in the discourse take place simultaneously with the discourse and plot. It should now be clear how important it is to consider whether playing computer games is comparable with storytelling or is better understood as a set of actions performed under certain conditions, as is the case with other types of games. Computer games differ from other games, however, in that if the sequence of actions can be interpreted as a story, that story is one that happens now but not here, for the events occur somewhere else (i.e. in the virtual world of the game). Moreover, players know that the ending and all the possible routes through the game have been determined in advance by the program; all that is required of them is to find the ending during the game. Actions in computer games depend on the player’s activities. So, unlike films or literary works, the narrativity of computer games cannot be analysed without considering the process of play itself. The process of 7

The concept of the story as a construction of the recipient was introduced by Todorov (1966) and is also employed by Bordwell (1985).


51

playing a game is necessary if a narrative is to emerge. The player authors the narrative of a game (this does not, however, mean that the player is the author of the game as a whole). Two questions are raised by the player’s contribution to the possible narrativity of computer games and the simultaneity of the three levels in them: how can the actions of the story be separated from the discourse, and how can authorship be described in the case of computer games? 5. Story, Plot and Narration in Computer Games The first of the above questions can be approached using the text adventure game Zork (1972) as an example. As the term ‘text adventure’ suggests, Zork employs text as its only form of representation. It begins by displaying the following statement: West of House You are standing in an open field west of a white house, with a boarded front door. There is a small mailbox there.

To play the game, the player has to enter commands. For example: >go to house

When a command is executed, the program responds to the player immediately. For example: >go to house It’s here! >open door the door cannot be opened >go east The door is boarded and you can’t remove the boards. >go north North of House You are facing the north side of a white house. There is no door here, and all the windows are boarded up. To the north a narrow path winds through the trees. >go east Behind House You are behind the white house. A path leads into the forest to the east. In one corner of the house there is a small window which is slightly ajar. >open window With great effort, you open the window far enough to allow entry.

52

Britta Neitzel

>enter Kitchen Score: 10 You are in the kitchen of the white house. A table seems to have been used recently for the preparation of food. A passage leads to the west and a dark staircase can be seen leading upward. A dark chimney leads down and to the east is a small window which is open. On the table is an elongated brown sack, smelling of hot peppers. A bottle is sitting on the table. The glass bottle contains: A quantity of water. >go west Living Room You are in the living room. There is a doorway to the east, a wooden door with strange gothic lettering to the west, which appears to be nailed shut, a trophy case, and a large oriental rug in the center of the room. Above the trophy case hangs an elvish sword of great antiquity. A battery-powered brass lantern is on the trophy case.

After picking up the lantern, collecting the rope and knife in the attic, and taking the garlic inside the brown sack, the player takes the sword, removes the rug, and goes underground through a concealed trapdoor. The troll encountered en route can be killed by typing ‘kill troll with sword’. After this explicit performative act has been carried out, the player is able to enter a confusing maze of rooms.8 When playing the game, the player chooses certain routes between the rooms, with the result that a network of rooms and routes between them comes into being. In the course of the game, the ‘you’ addressed by the program follows certain routes, examines objects, and fights potential enemies. In this way, the world of Zork is examined and a sequence of actions that can be called a plot unfolds. The program itself, on the other hand, contains object definitions, databases, and algorithms that do not present a story but define a range of possibilities and enable potential sequences of events. In a computer game, the distinction between the events of the story and the events of the narrating is not a chronological one but one between potentiality and actuality. To describe the different levels of which computer games consist, I propose distinguishing between the commands typed by a player (e.g. ‘go west’) as fictional acts of narration and the act of typing as the real process of narration. The imagined moves made by the ‘you’ constitute the events of the story.9 8

9

Zork has 191 different rooms and 211 manipulable objects. The dictionary of Zork contains 908 words, with which 71 actions can be described (see Lebling/Blank/Anderson 1979). As can be seen from the example, in Zork there are very few descriptions of events; descriptions of states are far more frequent. Each command and its result, that is a description of the next room, are displayed on the screen. Movements and other actions are not displayed or described; they exist only in the player’s imagination.


53

The process of selecting from among the possible routes and objects can, like a process of verbal narration, be understood as a process of making sense. According to Luhmann, sense is a ‘selective relationship between system and environment’ (1985:34), which means that a selection and actualisation is made of that, given as a possibility. The non-selected possibilities do not disappear completely but remain in the background as non-actualised possibilities. In computer games, this background is the program. If the term ‘narration’ is to be applied to computer games, its definition must be modified so as to read as follows: the act of narration is the selection and arrangement of events by the player. In a text adventure like Zork, it is relatively easy to separate the commands given by the player from the room descriptions. This becomes more difficult when we consider the visual presentation used in graphic adventures. In the evolution of the latter type of games, the commands of the player have coalesced with the movements of the avatar;10 it has become less and less common for the commands to be an observable part of the discourse level. In the game Blade Runner (Virgin Interactive, Westwood 1997), for example, the player points at a place on the screen to which the avatar then goes.

Figure 1: Blade Runner

10

An avatar is a character controlled by a player in a game.

54

Britta Neitzel

Silent Hill (Konami 1999) has a mixed command structure. While the avatar moves though space in the game in real time, the actions of collecting objects still show a clear command structure. This overt structure dissolves more and more in action adventures such as the Tomb Raider series of games (Core, Eidos 1996–2003).

Figure 2: Tomb Raider 1

Figure 3: Half Life


55

The command structure is completely hidden in action games. Third-person and first-person shooters not only depict moving and shooting in real time; the avatar also picks up objects by moving over them.11 But while the elements of the plot and the elements of the narration fuse more and more, it is still possible to distinguish between elements of the story and elements of the discourse. The reason for this is that the games specify a possible complete story. In Tomb Raider, for example, Lara Croft is given the task of finding an ancient artefact called the Scion, which she has to obtain for her client, Jacqueline Nadla. Lara goes to Peru, where she searches for and finds the Scion. She realizes that Nadla is evil and defeats her with the Scion. Lara then sails into the sunset on her opponent’s yacht. This story can be actualised12 by the walking, running, and jumping of the avatar, leading to the creation of a plot. As Certeau puts it, ‘the act of walking is for the urban system what the utterance (the speech act) is for language or for formulated statements’ (Certeau 1980:189). Metaphorically speaking, then, computer games use movements to tell stories. There are several reasons for adopting this view: (1) The actions of movement are different each time the game is played. This means that they involve the plot rather than the story. (2) Games do not list movement actions in their score tables. Movements affect only the time taken to complete a level. This is a measure of performance and represents narrative rather than story time. (3) Textual commands such as ‘go west’ and their visual equivalents are performative symbolic acts. They designate actions, but they are not the actions themselves. In narratological terms, computer games transform the actions of the (real) act of narration—in the case of the computer game: pressing buttons—into plot actions. In terms of play theory: a game is played. This becomes clear by a short glance at other types of games. When playing, players perform actions that are not taken seriously; they are merely pretending. This is most apparent when an actor pretends to be somebody else in a role-playing situation. The same applies to play-fighting—the people involved are only pretending to fight. People who play must always be aware of the

11

12

Many computer games involve exploring an unknown space and/or solving a task in an unfriendly environment. It follows that making movements in order to explore the environment, shooting enemies, and collecting objects that might help solve problems are very common actions. Of course a player must not follow this story-line but can do other things with the avatar, e. g. practicing swimming or jumping skills.

56

Britta Neitzel

fact that they are playing. If an actor playing Hamlet believed himself to be Hamlet, he would be schizophrenic. Thus, play is always a combination of two systems, the world of play and the real world, which is referred to and negated in the process. Play is contradiction: a playful action is and simultaneously is not a certain action. Or as Gregory Bateson (1955:41) describes play: ‘›These actions, in which we now engage, do not denote what would be denoted by those actions which these actions denote.‹ A playful nip denotes the bite, but it does not denote what would be denoted by the bite.’ These two levels are separated in computer games: outside the computer, a player in the real world presses buttons or keys while the screen displays actions in a fictional world. So, as far as play is concerned, pressing buttons does not mean pressing buttons but, for example, shooting monsters. Play and narrative, then, have in common that they deal with levels of reality and fictionality, but they do so in different ways: while narratives separate the two levels, play fuses them.

6. Narrative Situations In computer games, the real actions of narration are transformed into fictive plot actions. And these fictive plot actions are displayed in a certain way in a discourse. Narratology can be very helpful when it comes to describing the result. Narrative situations are one example of how narratological models can be applied to computer games. My description of narrative situations is based on that of Genette (1980), and employs a distinction made by him in his discussion of narrative mood. He shows that the categories of perspective and voice should not be confused. According to Genette, perspective or point of view describes the visual angle from which the story is seen in the discourse, and provides an answer to the question of who sees. Voice, on the other hand, concerns the narrator and the relation between narrator and story, thus providing an answer to the question of who speaks (cf. Genette 1980:162). Adapting the separation of the two questions to the analysis of computer games is unproblematic as far as the category of point of view is concerned: it can be employed without difficulty. Our understanding of voice, however, needs to be modified. I propose reformulating the question behind it as ‘who acts?’ instead of ‘who speaks?’, for the events in a computer game are not told but acted out, as described in the preceding section of this article. Accordingly, using point of view as a terminological model, the category of voice can be redesignated as that of point of action.


57

6.1 Point of View Perspective and point of view, or focalization, as Genette calls this category of mood, concern the extent of the information that the discourse reveals about the narrated world. Genette defines three kinds of focalization: (1) zero focalization, or a non-focalized story, is found when a story has an omniscient narrator; (2) internal focalization occurs when a story is told from the point of view of a character; and (3) external focalization occurs when the point of view concentrates on the actions of a character without anything about the feelings of that character being revealed. Apart from cases in which characters speak (uncommon even today), all knowledge in computer games is acquired through the things that can be seen in them (i.e. through visual perspective defined by spatial factors affecting the field of vision). This perspective affects the player’s distance from the virtual world. There are three kinds of perspective on the virtual world; they correspond roughly to Genette’s triad of focalization types. (1) Subjective perspective (internal focalization) is dominant in firstperson shooters, in which an avatar is not displayed and there is simply a hand or weapon at the bottom of the screen instead (see figure 3). The space that should contain the body to which the hand belongs as well as the eyes belonging to that body is empty. The image on the screen represents the field of vision of an avatar that players have to imagine. As part of this imagination they provide the avatar with their own body and seeing eyes. The diegesis is therefore not self-contained but extends beyond the monitor to the player. (2) Action adventures (e.g. Tomb Raider) visually display an avatar capable of seeing. For example, Lara Croft looks inside the caves of Qualopek before she enters them. The avatar is followed by a virtual camera and is usually seen from behind in the game (see figure 4). The player can also switch to ‘look mode’, in which the avatar can look in various directions. This point of view is linked to the movements of the avatar; it is not a substitute for the viewpoint, rather a case of viewing-with. In reference to Mitry (1998: 218), I should like to call it semi-subjective perspective. It is equivalent to external focalization in Genette’s framework. (3) The third kind of perspective in computer games, the objective point of view (zero focalization), is the oldest and most varied. The objective point of view presents from the outside a space in which actions take place. The virtual world cannot be viewed from any point inside it. This kind of point of view was used in the early computer games of the 1970s (see fig. 5).

58

Figure 4: Lara Croft

Figure 5: Space Invaders

Britta Neitzel


59

Graphic adventures, too, employ the objective point of view. They present an avatar in the space in which it performs its actions. The avatar does not have a perspective of its own on things. An especially clear indication that this is the representation of a space-for-action, is the use of ‘takes’, which sizes correspond exactly with the sector of the virtual world that can be manipulated.13 Likewise, simulations and strategy games employ the objective point of view in showing from above the space in which actions take place. Recently, games that allow the point of view to change have become increasingly common. The point of view in computer games influences the feeling of distance that the player has to the virtual world. While in games employing the subjective point of view the diegesis seems to be open, the semi-subjective and objective point of view provide a greater self-containment. 6.2 Point of Action The point of action describes the position where actions can originate and the way in which they can be performed in the fictional world. This position can be located inside or outside the diegesis, allowing us to distinguish between intradiegetic and extradiegetic points of action accordingly. It is clear that an intradiegetic point of action is present when a game depicts an avatar that is inside the diegesis and initiates the action. An intradiegetic point of action is also created when a role in the diegesis is assigned to the player, even if this role is not represented by an avatar. This is the case in Civilization (Microprose, Activision 1996), for example, in which the player chooses a ruler character who controls the lives of his subjects. An extradiegetic point of action is present when the player does not adopt such a role, as is the case in Myst (Broderbund, Cyan, 1993) and Tetris (1987). If, instead of focusing on the origins of actions, we look at their destinations, or goals, we can make a distinction between centred and distributed points of action. Distributed points of action allow an action to be initiated in multiple locations in the virtual world, as in simulations and strategy games, for example. Team sport games also operate on this principle. Centred actions can only be executed in a single location. This terminological addition allows us to differentiate between strategy games and action games: while the point of action is intradiegetic in both genres, it is distributed in one and centred in the other. A third distinction covered by the concept of point of action is that between direct and indirect actions. Depending on how the commands 13

Here the origin of graphic adventures in text adventures becomes clear. In Zork, for example, we have seen that it is always a particular part of the world that is described.

60

Britta Neitzel

are carried out, the structure of commands can or cannot be discerned on the computer screen. In Tomb Raider, for example, every keypress instantly leads to Lara performing an action, whereas in graphic adventures the player clicks on certain objects or characters so that the avatar then approaches them to examine or address them. As mentioned above, some adventures, such as Grim Fandango (Funsoft, Lucas Arts 1998) and Silent Hill (Konami 1999) combine indirect and direct points of action, whereas newer games tend to employ the direct mode. Indirect points of action are also found in simulations and strategy games. In some games, an indirect point of action is included in the diegesis: in Command and Conquer (Virgin Interactive, Westwood 1996), for example, the troops acknowledge their orders by saying ‘yes sir!’14 7. Authorship In this final section, I suggest a narratological approach to describing the authorship of narrative computer games. In computer games, the player determines the order of a concrete sequence of events in the course of playing the game. The player is responsible for creating the plot. Only the plot is affected by the exploratory and configurational functions of the player that are essential to all games. The player chooses from the possibilities that the program makes available and transfers them to the process of play, and so places a text-space in a time. By constantly inputting information throughout the game, the player prompts the computer to process the sequence of events. It is only in the course of play that possible plots are realized, individual objects combined, and actions produced and arranged in a chronological order. The program itself does not contain a chronological sequence of events; rather, it specifies possible sequential and causal relations by means of algorithms, object definitions, and databases. Any potential narrative only comes into being during the game and not before, for only then are chains of actions formed. A game can only be considered a narrative when it is actually played. Hence, the process of playing and the actions of the player are integral parts of a game and any possible narrative associated with it. This leads to a concept of shared authorship. Because players carry out the narrative 14

In more precise terms, the difference between direct and indirect points of action can be understood as one of modality, for it describes primarily the relative distance between story (when present) and discourse. An indirect point of action increases the distance and thus facilitates the process of distinguishing between them.


61

actions, they could be described as co-authors, for it is ultimately they who determine what takes place. However, such a description would give too much importance to the role of the player, for it implies that the player has the same importance as the programmer. Instead, I propose borrowing the term ‘implied author’ from literary theory in order to describe the function of the player as an author implied by the game. The term ‘implied author’, originally introduced by Wayne C. Booth, refers to an authority created by a text that is superior to the narrator and spans the entire text. Analogously, the implied author in a computer game is a necessary condition for the game to be a series of events, to be a narrative. It is implied by the game. Using the term ‘implied author’ also has the advantage of reflecting how many players view their involvement in computer games: they see themselves as the source of the symbolic actions that take place. The implied author is not solely responsible for the production of the game narrative. Another entity, the implied creator, as I shall call it, has primacy over the implied author. The implied creator is responsible for forming the virtual world, responsible that is, for the setting, the characters, and what happens in that world. The properties determined by the implied creator also include the characteristics of avatars (e.g. their appearance and how they perform actions) and the restrictions and obstacles that constrain actions in the game. The implied creator makes the initial selection from the mass of conceivable possibilities by making a subset of them available as choices in the program. The implied author then selects from them in a second process of selection. Summarizing the above, we can outline a system, resembling Gunning’s filmic narrator system, of the four levels of representation identifiable in computer games: (1) the world design (the rules of the virtual world), for which the implied creator is responsible; (2) the visual design (the presentation of the virtual world), for which the implied creator is responsible; (3) the plot design (the arrangement of action), for which the implied author is responsible; and (4) the chronological design (the temporal relation between story and discourse), for which the implied creator is responsible. When we consider these levels of representation, it does not seem feasible to apply to computer games the model of narrating entities that has established itself in the study of literature. The four levels described above are much more closely related to the system of cinematic discourse developed by Gunning. Unlike films, however, computer games have something that

62

Britta Neitzel

introduces an anthropomorphic element into the system: the implied author represented by an avatar. 8. Conclusion Using narratological methods to describe computer games provides insights into their structure. It also illustrates the limits of narratological methods and contributes to the description of differences between different media and between the wider cultural concepts of narrative and the game. Finally, it shows us that some features of computer games defy narratological description. References Aarseth 1997 Aarseth, Espen J.: Cybertext. Perspectives on Ergodic Literature. Baltimore / London 1997. Barthes 1983 Barthes, Roland: “Introduction to the Structural Analysis of Narratives.” In: Barthes, Roland: Selected Writings. Oxford 1983. Bateson 1955 Bateson, Gregory: “A Theory of Play and Fantasy. A Report on Theoretical Aspects of the Project for Study of the Role of the Paradoxes of Abstraction in Communication.” In: Psychiatric Research Reports 2, 1955:39–51. Bordwell 1985 Bordwell, David: Narration in the Fiction Film. London 1985. Caillois 1967 Caillois, Roger: Le jeux et les hommes. Le masque et le vertige. Paris 1967. Certeau 1988 Certeau, Michel de: Die Kunst des Handelns. Berlin 1988. Chatman 1978 Chatman, Seymour: Story and Discourse. Narrative Structure in Fiction and Film. Ithaca / London 1978. Chatman 1990 Chatman, Seymour: Coming to Terms. The Rhetoric of Narrative in Fiction and Film. Ithaca / London 1990. Gaudreault 1984 Gaudreault, André: “Narration et monstration au cinema.” In: Hors Cadre 2, 1984:87–98. Gaudreault 1989 Gaudreault, André: Du littéraire au filmique. Système du récit. Paris 1989. Genette 1980 Genette, Gérard: Narrative Discourse. An Essay in Method. Oxford 1980.


63

Gunning 1991 Gunning, Tom: D.W. Griffith and the Origins of American Narrative Film. The Early Years at Biograph. Urbana / Chicago 1991. Huizinga 1950 Huizinga, Johan: Homo Ludens. A Study of the Play Element in Culture. Boston 1950. Jost 1989 Jost, François: L’œil – caméra. Entre film et roman. Lyon 1989. Lebling / Blank / Anderson 1979 Lebling, David P. / Blank, Marc S. / Anderson, Timothy A.: “Zork: A Computerized Fantasy Simulation Game.” In: IEEE Computer, 12.4, 1979:51–59, see also: (02.03.04). Luhmann 1985 Luhmann, Niklas: “Sinn als Grundbegriff der Soziologie.” In: Habermas, Jürgen / Luhmann, Niklas: Theorie der Gesellschaft oder Sozialtechnologie – was leistet die Systemforschung? Frankfurt / Main 1985:25–100. Mitry 1998 Mitry, Jean: The Aesthetics and Psychology of the Cinema. London 1998. Neitzel 2000 Neitzel, Britta: Gespielte Geschichten. Struktur- und prozessanalytische Untersuchungen der Narrativität von Videospielen, Diss., Weimar 2000. http: // www.db-thueringen.de / servlets / DerivateServlet / Derivate-2063 / Dissertation.html. Plato 1991 Plato: Der Staat. Stuttgart 1991. Scheuerl 1990 Scheuerl, Hans (ed.): Das Spiel. 11th edition, Weinheim / Basel 1990. Todorov 1966 Todorov, Tzvetan: “Les catégories du récit littéraire.” In: Communications 7, Vol. 8, 1966:125–151. Tomashevsky 1965 Tomashevsky, Boris: “Thematics.” In: Lemon, Lee T. / Reis, Marion J. (eds.): Russian Formalist Criticism. Four Essays. Lincoln / London 1965:61–95.

Computer Games Blade Runner Developer: Virgin Interactive. Publisher: Westwood 1997. Civilization Developer: Microprose, Publisher: Activision 1996 Command and Conquer Developer: Virgin Interactive. Publisher: Westwood 1996. Grim Fandango Developer: Funsoft. Publisher: Lucas Arts 1998. Half Life Developer: Valve. Publisher: Sierra 1998. Myst Developer: Broderbund. Publisher: Cyan 1993.

64

Britta Neitzel

Tetris Developer: Alexey Pazhtnov 1985. Silent Hill Developer and publisher: Konami 1999. Space Invaders Developer: Taito. Publisher: Midway 1978. Tomb Raider Developer: Core. Publisher: Eidos 1996–2003. Zork Developer and publisher: Infocom 1972.

Narrative in Music: The Case of Beethoven’s “Tempest” Sonata* Douglass Seaton (Tallahassee) 1. Introduction Critical theory in music has recently begun to consider the possibility of musical narrative. Doubts have been raised about musical narrativity on the basis of music’s supposedly abstract material, the impossibility of a past tense in music, repetition in musical forms, and other grounds.1 Nevertheless, narrativity can be demonstrated even in purely instrumental musical works, and indeed it constitutes a defining factor for musical Romanticism. * 1

See the appendix to this article including a sound file at For a brief survey of this literature, see Maus (2001). Maus concludes by saying that ‘The exploration of instrumental music as narrative remains a tantalizing, confusing, problematic area of inquiry’ (642). In an earlier article Maus reviews a variety of different approaches to this issue by musicologists who have attempted narratological investigations (1997). The three main objections to a narratological approach to music are represented by Jean-Jacques Nattiez, Carolyn Abbate, and Peter Kivy. Nattiez (1990) is really talking about what I call plot, and he presumes that in order to be narrative music must represent a story either in the real world or in a verbal medium. Although his argument rests on the assumption that music is essentially abstract and therefore necessarily non-referential, one need not find this problematic in terms of musical plots. Musical events are real events—in an important sense more real than the events referred to in a fictitious story—and they raise and sometimes frustrate real expectations in the listener. A brief consideration of musical plot appears in my discussion below, and its applicability will be demonstrated with the sonata to be considered in this essay. Carolyn Abbate (1989; expanded 1991) discusses the problem of narrativity in music on the basis of music’s having no past tense. Her point is that literary narrative typically depends on the past tense in order to distinguish the (present) telling of a story from the (past) action that it reports. Peter Kivy (2002:chap.8) objects strenuously to the possibility of musical narrativity on the grounds that he finds it impossible to reconcile the inherent use of repetitions in common musical forms with the absurdity of such repetitions in a story. Two objections might easily be raised to this line of reasoning: (1) we need not—indeed, should not—carry the presuppositions of narrative in one medium into the understanding of

66

Douglass Seaton

A major weakness in previous discussions of musical narrative has been a too close dependence on the assumption that narrative per se belongs to literature. In fact, the idea of narrative in music does not depend on the imitation of literature, any more than the reverse.2 The present discussion will argue that, entirely apart from literary content, a musical work can satisfy the two requirements of narrative that transcend artistic medium: plot and voice. Plot is established in instrumental music of the eighteenth and nineteenth centuries by the definition of character through rhythmic / melodic identities and the articulation of action through harmonic patterns of stability, rising tension, climax, resolution, and dénouement. Voice—the narrative persona—is created in various ways, sometimes embodied in musical forms but often, too, in verbal cues, in musical behaviors surrounding the performance and hearing of a work, or in reception. This essay will first review the means and techniques by which music of the eighteenth and nineteenth centuries was able to enact plot, and it will suggest how listeners have encountered the narrative voice in musical works. In order to demonstrate how these principles work in practice, it will then consider a work that has troubled listeners and commentators for two hundred years, Beethoven’s 1802 Piano Sonata in D Minor, op. 31 no. 2, commonly known as the “The Tempest”.

2

narrative in another, and (2) neither time nor event order is necessarily unidirectional in verbal narrative, so that it is already obvious that narrative need not be bound by any assumption that events and time operate in plots as they do in direct life experience. The fact that narrativity in music has not necessarily been set aside by these objections is represented by William Kinderman (1992:esp. 141–44), who explicitly takes issue with Abbate’s and Nattiez’s arguments, though he does not offer a clear definition of narrativity. He is particularly concerned with effects that allow the music of Beethoven’s Piano Sonata op. 110 to ‘transcend a linear temporal succession’, and thus may also imply an answer to Kivy’s worry about repetition. Kinderman also insists, and rightly so, on an understanding of musical narrative unencumbered by the peculiar constraints of any literary medium. Werner Wolf (2002) provides a very helpful discussion of these and other issues by a scholar from outside the field of musicology. Among other problems of musical narrative that Wolf addresses is music’s non-referentiality, which he ultimately does not find sufficient to prevent musical works from being regarded as narrative (77–83). Abbate betrays the limits of her perspective when she writes that ‘literary theories of narrative suggest ways in which music cannot narrate’ (1989:228). My point here is that narrative in one medium need not employ the means or even necessarily adopt the assumptions of another. The present essay will insist on an understanding of narrative that is conscientiously not constrained by literary theories.

67

Narrative in Music

2. Methodological Principles We must first turn then to the issue of plot expressed in music.3 In simple terms, we can take as a definition that we have a plot in any action in which characters become involved in a sequence of events in which tension arises and is resolved. Now, as it happens, in the eighteenth century musicians began to employ a new way of describing musical ideas—rhythmic / melodic units—no longer according to their ‘affect’, as in the past, but according to ‘character’. Furthermore, the musical theme of a piece was no longer referred to as ‘subject’ but as ‘motive’, suggesting that it represented some sort of motivation, whether conscious or instinctive. Thus, a musical passage might be identified as ‘military’ or ‘amorous’, ‘aristocratic’ or ‘rustic’, and even ‘masculine’ or ‘feminine’; of course, such characterizations are not all mutually exclusive, so that a theme might be both aristocratic and amorous, or a musical idea might first appear as lyrical and return as heroic. At the same time, Enlightenment composers began to adopt an approach to long-range harmonic contour in substantial movements of musical works, which produced the principle that became known (long after the fact) as ‘sonata form’ (see Diagram 1). Conventional designation

Plot contour

Harmonic level

Musical material

Typical thematic character

Introduction (optional)

Prologue

Leading to tonic

O – opening

Anticipatory

Part I – Exposition Section 1

Stable situation

Tonic

P – theme(s) associated with principal key

Affirmative, forceful

Increasing tension

Modulatory

T – theme(s) associated with harmonic transition

Active

Contrasting situation

Dominant (or relative major)

S – theme(s) associated with secondary key K – theme(s) associated with closing

Contrasting to P, often lyrical Affirmative, often “stock” character

Rising action to climax

Unstable

Development of previous material; new material (N) may be introduced

Unpredictable

Section 2

Part II Section 3 – Development

Section 4 – Reprise

Resolution

Arrival of tonic

P

As in exposition

Dénouement

Tonic

(T), S, K

As in exposition, except that all characters are in stable position

Diagram 1: Sonata form as plot 3

The understanding that eighteenth- and nineteenth-century music depends upon the concept of plot is heavily indebted to Charles Rosen’s magisterial study The Classic Style: Haydn, Mozart, Beethoven. Rosen does not speak of plot but of drama—i.e. an action presented without an evident narrator—but the point is the same. Rosen formulates the difference between the musical presuppositions of the Baroque and the Enlightenment very clearly in his statement that ‘Dramatic sentiment was replaced by dramatic action’ (1971:43).

68

Douglass Seaton

Fundamentally, the principle is fairly straightforward. A movement must begin by establishing a home key or tonic. The music then modulates, rising into another key (usually the key of the fifth scale degree or dominant, or in minor keys the third scale degree or relative major), which creates a heightening of tension. This setting up of the movement’s fundamental opposition is often referred to as ‘exposition’. A more or less extended period of harmonic instability, sometimes called ‘development’ (although development in a general sense may occur at any point in the form) eventually leads to a moment of climactic anticipation. The return of the tonic provides harmonic resolution, restoring stability. Crucial to this musical plan is the coordination between the different musical themes, with their respective characters, and the harmonic events. The result of this concinnity is that the themes either motivate or experience the harmonic progress of the music; in other words, they are legitimately ‘agents’ or ‘patients’ in the work. Standard functions, each with its own character, include the following: Opening themes, the point of which is to create anticipation at the beginning of the action. These tend to be fragmentary and open-ended, perhaps even unsettled in character. They are frequently external to the harmonic plan and thus not involved in the plot. Principal themes, which establish the tonic key of the movement both at the beginning and at the moment of resolution. Themes in this position are likely to be strongly rhythmic and bold in character. Transition themes, which destabilize the key and move toward the contrast key. We expect such themes to be lively and active. Secondary themes, which is to say, themes associated with the secondary key, which serve to stabilize the secondary key and to contrast with the character of the principal theme. Often, secondary themes are lyrical and long-breathed. Closing themes, which confirm the secondary key. Closing themes tend to be formulaic and predictable in character. Of course, one of the fascinating things about listening to any given piece is the interplay between the characters and their roles within the overall plan. Generally, however, the form tends to assign a controlling, active role to the principal theme or character, because it has the responsibility of establishing the main key at the beginning and marking the return of that key at the resolution. The rest of the themes (or characters) are eventually brought from their original harmonic level into the main key, so that especially the character associated with the secondary key seems to assume a more passive role. The way in which all of this plays out in

Narrative in Music

69

any given work will, of course, be specific to the piece, just as any plot in any artistic medium is specific, and that is what holds our interest. It is not at all necessary, I hasten to emphasize, to translate pieces of music into verbal or pictorial stories. This has, of course, been done many times, sometimes more and sometimes less amusingly. The point here is that the musical characters and musical shaping of tension in a work are at the same time entirely specific and refer to nothing else. Naturally, we communicate about musical works in words. We can do this in great detail by using the technical terms of music analysis, but that approach inevitably sacrifices the immediacy of the character and the experiential contour of tensions and resolutions that we perceive directly in hearing the music. We can also describe music in metaphors or assign names to the characters (identifying a heroic theme as Promethean, for example), but that either brings too much external baggage to the character or fails because the metaphor or name means something different to anyone with whom we share such fantasies.4 For this reason I do not intend here to tell a story based on a Beethoven sonata but rather simply to let the music express its own character and enact its own plot. So plot is not the exclusive property of literature or even of verbal media; it belongs equally to music, and not merely as a kind of borrowing from literature into music that requires us to translate musical characters and actions into quasi-literary ones. Indeed, plot itself might properly be regarded as the quintessential Enlightenment way of conceiving experience, manifesting itself either in literature or in music—or in any other form—a principle that transcends medium. The sonata form meets all the criteria for plot, and some other musical forms may do so as well. But I have not so far made the case for musical narrative. Merely enacting a plot will make a musical work dramatic, corresponding to a play enacted on the stage, not narrative, like a novel. And, indeed, most musical movements of the eighteenth century are better regarded as dramatic than as narrative.5 When we move toward the nineteenth century, the issue of narrative begins to come into play. And the reason for this is obvious: to the Romantic imagination it became important to establish the subjective mind or spirit from which an artistic manifestation emerged. As Susan McClary (1997:24) has rightly observed: ‘during the nineteenth century virtually 4

5

John Neubauer (1997) discusses the ‘emplotment’ of musical works by various writers, representing various degrees of such metaphorical story telling. Fred Everett Maus (1991) raises the important issue here of the difference between action and discourse, but he also concludes that it is difficult to give any identity to a narrator in music. In the end, Maus is inclined to accept the concept of drama in music but to be skeptical about narrative.

70

Douglass Seaton

all cultural enterprises in Europe aspired to the condition of narrative.’6 Thus we need to deal with the issue of voice in music. At the outset we should acknowledge that the establishment or discovery of narrativity in music does not rely on conventional principles as conveniently as does plot. In addition, as mentioned earlier, the narrative voice may be identified audibly in the musical work itself or externally. Music-internal evidence for the narrative persona may be of several types, and I shall list here only three: A particular idiom that governs the work, essentially a recognizable rhetorical manner that can be attributed to a certain kind of speaker. Establishment of the narrative persona’s identity by quotation of or allusion within the piece to a specific repertoire of music. Intrusion into the piece’s plot of some music that clearly does not participate in the action, often a passage that serves as commentary. This is the kind of instance that we will observe here. Extramusical identification of the narrative persona also operates in a number of ways, sometimes under the composer’s control but also determined by historical information or reception history: ‘Paratextual’ verbal cues, such as the attachment of titles or programs to musical works. Conventions of performance, which, for the nineteenth century, would include the appearance of the conductor on the podium in front of the orchestra or the solo recital pianist seated in profile on the stage. Posterity’s identification of the persona with the composer, at least as the composer is known or understood by the conventional wisdom generated by biography (and sometimes by autobiography). We shall also have the opportunity here to observe an instance of this. Regardless where and how it takes place, however, the perception of the persona, a voice that we identify as speaking through the music, renders the musical experience Romantic. The listener responds to her or his awareness of this voice by expecting and seeking to know the persona as the subjectivity behind the musical expression. 6

McClary is, to be sure, more interested in the issue of a totalizing master plot than in the subjectivity that concerns us here; nevertheless, the validity of her statement extends to narrative voice, as well as to plot. This is the motivation for the impulse toward narrativization that has led to many arbitrary and farfetched programs applied to music by romanticizing commentators. On this subject, see Wolf (2002:83–94). The real existence of this impulse toward narrativization also provides, however, the justification for musicology’s seeking and studying the narrativizing aspects of the actual historical experience of music listeners in the nineteenth century.

71

Narrative in Music

3. Demonstration (I)—establishing narrative by music-internal analysis With the preceding discussion to serve as methodological background, we can now turn to Beethoven’s “Tempest” Sonata, which turns out to be a particularly interesting and instructive example. As has been discussed by many previous analysts, the sonata’s first movement employs the standard plot resources commonly identified as sonata form, but it also departs from the norm to create an analytically vexed situation in which the functions of character within the plot remain uncertain.7 Moreover, beyond its peculiar thematic and harmonic plot outline, the movement introduces a striking passage of recitative, a musical idiom ‘extra-generic’ to the piano sonata (see Diagram 2). Conventional designation

Plot contour

Harmonic level

Musical material

Thematic character Measure Numbers

Introduction? Exposition?

Unstable

Leading to D minor

O?/P?

Anticipatory

1–20

Part I – Exposition (continued?) Section 1

Immediately increasing tension

Only momentarily in D minor, than modulatory

T?/P?

Bold

21–41

Contrasting

A minor

Section 2

Part II Rising action Section 3 – Development to climax

Section 4 – Reprise

S

Mercurial

42–55

K

Stern

55–92

Unstable

Development of O?/P? Anticipatory T?/P? Bold K Stern

93–98 99–121 122–142

Resolution?

Arrival of D minor, immediately destabilized

O?/P? Recitative

143, 149–154 144–48, 155–171

Dénouement

D minor

Anticipatory Declamatory

S

Mercurial

172–185

K

Stern

185–225

Diagram 2: Form of first movement of Beethoven, Piano Sonata in D Minor, op. 31, no.2 “The Tempest”

The most provocative feature of the sonata’s first movement is its failure to produce a proper theme in the principal key. The opening passage is not set off as a mere introduction (as in Beethoven’s “Pathétique” Sonata, op. 13, for example), but in style it does resemble an introduction more than a first theme (or character) of a sonata plot. Confusing the situation still

7

For a helpful discussion of the “Tempest” Sonata, see Timothy Jones (1999:103–108). Jones largely adopts the analysis of the movement by Janet Schmalfeldt (1995:37–71). A detailed description of the compositional process of op. 31 no. 2 is provided by Barry Cooper (1990:177–96).

72

Douglass Seaton

further, a new idea begins at the close of this passage, clearly in the tonic key, D minor, with a motive surging boldly out of the bass and answered by a turning figure in a high register, supported by a vibrating accompaniment. This, however, also turns out not to constitute a proper principal theme but to move directly into the process of modulation. Its material does not serve to stabilize the key of D minor but only to destabilize the harmony with each new gesture. Thus the sonata begins in ambiguity—indeed, it has no proper beginning. After a clearly introductory passage there follows immediately a clearly transitional one. What ought to constitute a defining feature of any sonata movement, the stable theme in the principal key, simply does not appear at all.8 Once past this first uncertainty, the movement progresses more or less conventionally for a while. There comes a reasonably stable theme in the secondary key (in this case A minor) and a close in that key. The second part of the form opens with a developmental passage featuring, as we would certainly expect, the tonally unstable transitional material of the first part. Now the listener who approaches this movement with the expectations of the conventions of a sonata plot anticipates that its ambiguity will be cleared up at the moment of plot resolution (the reprise), when one or the other of the elements from the beginning must appear in some stable form to take on the normal and necessary function of reestablishing the stability of the music’s home key. If what at first appeared as introductory gestures really constituted the principal-key thematic material, then it ought to reemerge as the defining material for the restoration of the tonic. On the other hand, if the material that led so quickly into harmonic instability now comes to be understood as the beginning of the sonata form proper, then it ought to reappear in a tonally stable form here, now that no modulation is necessary. And so we may expect the return of the key of D minor to resolve the problem of clarifying the roles of the dramatis personae that the movement presented at its outset.

8

This ambiguity has been debated repeatedly in the history of analyses of this movement. For example, Adolf Bernhard Marx (1902 [1859]:142–144) manifests no doubt that the first twenty measures are introductory and that the principal theme begins in m. 21. Hugo Leichtentritt (1948:162–163) concurs. On the other hand, Ludwig Misch (1950:42–55) adamantly insists that the first measures constitute the principal theme, and the passage beginning in m. 21 is transition. Misch’s analysis simplifies the structural problem in the movement by disregarding the introductory character of the opening. He likewise describes the interrupting recitative that appears at mm. 144–158 and 155–171 as an ‘intensification’, misrepresenting its character as well.

Narrative in Music

73

But it does not. The opening passage does return, but it leads directly into a passage that comes from another world entirely, a recitative. This leads far afield, to the key of F minor, which in turn side-slips to another distant key, F-sharp minor. Then the music skips abruptly ahead to the secondary idea, now finally resolved into D minor. The rising bass motive never returns at all. As Janet Schmalfeldt puts it (1995:70), ‘the beginning of the recapitulation has the effect of obfuscating rather than clarifying the design at the beginning of the exposition.’ In short, then, the question that we have had throughout the course of the movement remains open, even when we might have thought it must be answered.9 And what of that recitative? Here we encounter one of those remarkable moments that shift the level of the discourse.10 Recitative does not belong to the world of the action of an instrumental sonata but to the genre of opera; it is a specifically vocal style, ‘extra-generic’ to the piano sonata.11 As such, it cannot properly constitute a part of the sonata form at all. Rather, it takes the part of an outside voice, indicating or highlighting precisely the crucial moment in the plot enacted by the rest of the music.12 It functions to draw our attention to the fact that, though we may have formulated a certain kind of question about the music, we will not

9

10

11

12

See Carl Dahlhaus (1980). In general, Dahlhaus’s concern, in contrast to my own here, is to demonstrate the coherence of the form rather than its essential disruption. The “Tempest” Sonata recurs frequently in Dahlhaus’s writings as a work of special interest. Dahlhaus hits the nail on the head when he points out that the ‘paradoxical relationship to the traditional model is actually part of the formal idea of the work’ (1991:169). In his survey of music in the nineteenth century this sonata is represented as symbolic of one side of the fundamental dichotomy of early Romantic musical aesthetics (1989: 13–15). Robert Hatten (1991) offers an excellent discussion of levels of discourse in music, with special reference to Beethoven. Lawrence Kramer (1998) also regards the recitative as external to the action—parergon—a position that corresponds to some extent with mine in the present discussion. He finds in this movement, and in the sonata taken as a whole, the legitimation of ‘the human subject in its own interiority.’ He does not pursue very closely the significance of the specific position of the vocal intrusion in the plot of the sonata’s first movement. Carl Czerny, who knew Beethoven quite well, says that the recitative ‘must sound like one complaining at a distance’; see Badura-Skoda (1970:43). Marx wrote: ‘Der erste Teil war still geworden, fast bis zum Verstummen; der zweite hebt mit jener ersten Largo-Frage wieder an, die, aus der Tiefe dringender emporgeworfen, in feierlicher Dreizahl ertönt und vergebens der lösende Antwort harrt. Streng vielmehr und ohne Zwischenrede tritt der Hauptsatz wieder an, dringt scharf empor, versinkt wieder zur Tiefe und bringt da . . . die Largo-Frage zurück. Ist sie noch nicht verstanden? es fehlte ja kaum etwas, als das Wort! — Und auch das soll nicht fehlen: aus dem Largo tritt ein Rezitativ hervor, — die Rede, die Musik wird; die Musik, die Rede werden möchte, um endlich verstanden und erhört zu werden’ (Marx 1902 [1859]: 144).

74

Douglass Seaton

receive an answer in the form in which we expect it. While we thought that we had understood how to follow the action of the movement, we must now understand that the meaning of this work goes beyond resolving a problem that the music poses. Instead, we learn that at times we encounter a mind that does not intend to give answers to the questions with which it confronts us. Entering at the moment when tension might be resolved and uncertainties clarified, this intruding ‘voice’ has the effect of commentary on the plot itself, constituting a level of discourse that stands outside the movement’s main action. It serves the function of a narrative voice, interrupting the action in this case explicitly in order to refuse the listener’s desire for clarification. Recognizing this work thus as narrative, though always without any dependence on literature, becomes the only way to expose the real nature of this music. 4. Demonstration (II)—Identifying Narrative Voice by Extramusical Means Musicological narratology may also seek insight beyond the work itself. Equally as intriguing as the “Tempest” Sonata’s first movement is another narrative about the sonata, the report of Beethoven’s sometime secretary Anton Schindler in his highly unreliable memoir of the composer (Biographie von Ludwig van Beethoven, appropriately published in English not as Biography of Ludwig van Beethoven but under the title Beethoven as I Knew Him), which forms the source of the reference to Shakespeare’s The Tempest in relation to op. 31 no. 2. Reporting to the composer the impression made by both this D-minor Sonata and the F-minor Sonata, op. 57, in a gathering of Carl Czerny’s, Schindler asked Beethoven to give him the key to the music. Beethoven’s reply was ‘Just read Shakespeare’s Tempest’ (1966:406).13 Although impossible to verify as factual, Schindler’s story engages the Dminor Sonata narratologically on two levels. First, it contributes to the mythology of Beethoven’s persona, which has undergirded since Beethoven’s 13

The original text reads as follows: ‘Eines Tages, als ich dem Meister den tiefen Eindruck geschildert, den die Sonaten in D moll und F moll (Op. 31 und 57) in der Versammlung bei C. Czerny hervorgebracht und er in guter Stimmung war, bat ich ihn, mir den Schlüssel zu diesen Sonaten zu geben. Er erwiderte: ›Lesen Sie nur Shakespeare’s Sturm.‹ Dort also soll er zu finden seyn; aber an welcher Stelle? Frager, lese, rathe und errathe!’ (Schindler 1970 [1840]:pt. 2, 221). The English edition translates ‘in der Versammlung bei C. Czerny’ as ‘Carl Czerny’s playing’, although the German only suggests a gathering at Czerny’s home or organized by him. Czerny need not have been the performer of the two sonatas on this occasion.

Narrative in Music

75

own time an assumption that the narrative voice in Beethoven’s music is that of the composer himself. It further suggests (possibly in Beethoven’s own mind, if Schindler’s report has any basis in an actual event) a connection between the ‘composer’s voice’14 and the character of Prospero in Shakespeare’s play, which turns out after all to offer a hermeneutic clue to the recitative’s intrusion into the music of the sonata. Assuming that Beethoven actually did say, ‘Just read Shakespeare’s Tempest’, Schindler obviously did not understand what the composer was saying, but he urged his readers to probe the subject: ‘Questioner, it is for you to read, to ponder, and to guess’ (1966:406). As a matter of history, Beethoven’s remark really has led commentators to speculate about how the music of op. 31 no. 2 reflects Shakespeare’s The Tempest. Arnold Schering (1934:80–86), to cite just one notorious instance, tried to link all three movements with the characters and action of the play, and Josef Pembaur (1915:3–4) contrived similar detailed programs for both op. 31 no. 2 and op. 57.15 Donald Francis Tovey suggested, although tongue in cheek, that ‘it will do you no harm to think of Miranda at bars 31–38 of the slow movement’ (1948:128).16 Certainly one could argue that the minor key and animated motion make the first movement sound stormy.17 There are several problems with this approach to relating Beethoven’s statement to the music. For one thing, there is no clear sense in which the structure of the music suggests the action of the play. One can only be left with a sense that the general character of tempestuousness or of feminine grace might emerge from various passages in the sonata. This would, of course, be equally true of any number of Beethoven sonatas. In fact, as Schindler’s story makes clear, the composer’s allusion to The 14

15

16

17

The allusion here recognizes Edward T. Cone’s landmark study The Composer’s Voice (1974), to which musicological narratology is greatly indebted for its groundbreaking consideration of persona in music. Michael Heinemann (1991) lists several of the other early writers who also discussed the question of the relationship of sonata to play. Tovey’s point is not to encourage this, of course. Immediately preceding this he writes ‘[w]ith all the tragic power of its first movement the D minor Sonata is, like Prospero, almost as far beyond tragedy as it is beyond mere foul weather’; and he continues, ‘but people who want to identify Ariel and Caliban and the castaways, good and villainous, may as well confine their attention to the exploits of the Scarlet Pimpernel when the Eroica or the C minor Symphony is being played.’ Tovey’s own treatment of the movement (1948:128–131) is purely descriptive of the musical events. See Owen Jander (1996). Actually, Jander rejects any connection between op. 31 no. 2 and the Shakespeare play, instead attempting to place the music into the Enlightenment tradition of musical depictions of stormy weather and the human response to it. The self-evident problem with this argument is that without the reference to the play there would be no reason to make a particular association between stormy weather and this music as opposed to any other Beethoven sonata movement.

76

Douglass Seaton

Tempest must apply as much to the F-minor Sonata op. 57 as to op. 31 no. 2, and one might argue that op. 57 is even more tempestuous. The name became attached to op. 31 no. 2 only because op. 57 already had its own name, “Appassionata”. We might, however, take Beethoven’s comment in quite a different way—a way that Schindler seems not to have understood. Beethoven’s statement, even if it was a direct response to Schindler’s question, need not have answered that question in its own terms. Let us reconsider the exchange in this way. Schindler asks Beethoven for the key to the sonatas, and Beethoven retorts that he should go read The Tempest. What might Beethoven have meant, if not simply that the sonata told the story of or was inspired by Shakespeare’s play? Central to The Tempest is Prospero’s relationship to the other characters. The former Duke of Milan, he lives on a desert shore in isolation from the world. He came there as a consequence of his earlier withdrawal from everyday affairs into his own world of magical studies, having first entrusted his dukedom to his brother, who then usurped Prospero’s position, treacherously conspired with the rival King of Naples, and cast Prospero asea in a leaky boat with only his books and his infant daughter. In act 1, scene 2, Prospero explains his history: … being so reputed In dignity, and for the liberal arts Without a parallel; those being all my study, The government I cast upon my brother And to my state grew stranger, being transported And rapt in secret studies … I thus neglecting worldly ends, all dedicated To closeness, and the bettering of my mind With that which, but by being so retired, O’er-prized all popular rate, in my false brother Awaked an evil nature, and my trust, Like a good parent, did beget of him A falsehood in its contrary as great As my trust was, which had indeed no limit, …

What can this have to do with Schindler’s question? We might suppose that Beethoven’s response (not answer) to the naive question intended to draw a parallel between the composer and Prospero.18 Not unlike Prospero, 18

Theodore Albrecht (1992) actually does suggest that Beethoven may have had in mind a parallel between himself and Prospero, but he argues that what the composer referred to was the fraternal strife between Prospero and Antonio, which in many respects paralleled difficulties regarding business matters between Beethoven and his brother Karl. He does not suggest any connection to the music or its expressive content, nor give much rationale for the composer’s remark as a response to an inquiry about particular sonatas.

Narrative in Music

77

Beethoven had withdrawn into his own art, leaving the common world behind. Also like Prospero, the composer had little patience for the petty questions of small minds. To tell Schindler to read The Tempest was not an attempt to explain the music but rather a way to indicate the futility of such a conversation. In other words, Beethoven’s reference to The Tempest does not relate directly to the action in the music of the two sonatas about which Schindler asked—still less to the D-minor Sonata in particular—but to the place of the music in the experience of the composer on the one hand and his interlocutor on the other.19 It is not, in fact, unreasonable to think of the character of the magician in The Tempest as a model for the position of Beethoven in Schindler’s anecdote. From this point of view, Beethoven’s response to Schindler’s request does not appear to be a direct answer to the question as it was asked. Instead of giving Schindler an explanation of the sonata’s content, Beethoven’s answer simply indicates that Beethoven, like Prospero, sees himself as different from others, a figure isolated by his own unwillingness to trouble himself with the petty interests and naive questions of duller intellects. 5. Synthesis Understood thus, Beethoven’s response to Schindler shares a viewpoint with the first movement of op. 31 no. 2. In both cases the issue turns out to lie beyond the immediate question that Beethoven’s listener poses. In both cases the response, refusing to be drawn into the realm of thinking in which the question is asked, speaks from an outside position that at worst we do not comprehend at all and at best takes us by surprise. The first movement of the “Tempest” Sonata brings to the fore the experience of levels of discourse. The movement’s plot, peculiar as it is, takes place on one level, the level of the Classic sonata form. The interrupting recitative comes as a voice from another world, a voice that speaks in a different idiom, and a voice that draws attention to and at the same

19

Among his list of ten possible relationships between The Tempest and op. 31, no. 2, Heinemann (1991) includes a somewhat similar point to the one that I am making here: ‘Das beliebte Motiv romantischer Autoren der Abgeschiedenheit des Künstlers von seiner Welt, konkretisiert in der Metapher der Insel, auf der der Komponist als Einsiedler in die Welt der Musik versunken erscheint, findet sich auch in Shakespeares Sturm. Dabei korrespondiert die Gestalt des Prospero, der die Faden der Handlung zu jedem Zeitpunkt in der Hand hält, mühelos mit dem Komponisten Beethoven’ (Heinemann 1991:302). Heinemann does not note any correspondence between this observation and the music of the sonata.

78

Douglass Seaton

time denies us a solution to the questions that trouble conventional expectation. Whose voice? We must, of course, steer away from the idea that it is Beethoven’s voice in an externally biographical sense. It is a fictional voice created for this movement and this meaning. To identify it, however, we can turn, perhaps ironically, to Schindler’s dubious memoir. Given its refusal to belong to the world around it, its focusing of attention at the point where the commonplace question fails, and its insistence on its superior position, we must regard it as a voice of isolated genius. To name it we can hardly do better than to call it a ‘Prospero voice’ or, more generally, an ‘exiled genius voice’. 6. Conclusion As this brief discussion has demonstrated, it is quite legitimate to establish definitions and models to support the possibility of musical narrative parallel to but independent of literature. A music-theoretical analysis of the “Tempest” Sonata’s first movement illustrates how narrativity can operate within a musical work. The examination of a story about the composer (in this case a presumably fictitious story) demonstrates how mythology comes into play as a determinant of authorial persona for a musical narrative. In this case, admittedly a somewhat remarkable one, the return from that story to the music, finally, hermeneutically connects the extramusically constructed narrative voice to the ‘voiced’ passage within the music. The “Tempest” Sonata, both in its tones themselves and in its history (or fictional afterlife, if we prefer), thus offers a lesson in musical narrativity. To the extent that this demonstration can serve as a model, of course, it can do so only in broad terms. Each musical work will have its own plot and narrative persona. The process of narrative analysis will differ from work to work, and the hermeneutic insights offered by music-historical or other extramusical evidence will vary widely.20 Nevertheless, this sonata demonstrates that narrativity as a principle does transcend the boundaries of artistic medium. Further, narratological method, still not well established in musicology, has much to offer to the future study of music, especially 20

Lawrence Kramer (1991) rejects structuralist approaches to narrativity in music and instead offers some examples of the hermeneutic work performed by narratological study of music. While I clearly find more to justify the discovery of narrative in analysis than Kramer does, we would certainly agree that narratological approaches are more likely to help us deal with music that lacks structural coherence than to demonstrate coherence, as well as that narrative thinking offers social and cultural or epistemic and self-reflective insights (or insights that combine any or all of these).

Narrative in Music

79

that of the nineteenth century. Musicologists thus have much to learn from our colleagues who work in narratology within other specific fields and from those who contemplate it more generally. References Abbate 1989 Abbate, Carolyn: “What the Sorcerer Said.” In: 19th-Century Music 12.3, 1989:221–230. Abbate 1991 Abbate, Carolyn: Unsung Voices: Opera and Musical Narrative in the Nineteenth Century. Princeton, NJ 1991. Albrecht 1992 Albrecht, Theodore: “Beethoven and Shakespeare’s Tempest: New Light on an Old Allusion.” In: Beethoven Forum 1, 1992:81–92. Cone 1974 Cone, Edward T.: The Composer’s Voice. Berkeley / Los Angeles 1974. Cooper 1990 Cooper, Barry: Beethoven and the Creative Process. Oxford 1990. Badura-Skoda 1970 Badura-Skoda, Paul (ed.): Carl Czerny: On the Proper Performance of All Beethoven’s Works for the Piano, with excerpts from Czerny’s Memoirs and Anecdotes and Notes about Beethoven. Wien 1970. Dahlhaus 1980 Dahlhaus, Carl: “Zur Formidee von Beethovens d-moll-Sonate opus 31, 2.” In: Die Musikforschung 33.3, 1980:310–312. Dahlhaus 1987 Dahlhaus, Carl: Ludwig van Beethoven und seine Zeit. Laaber 1987. Dahlhaus 1989 Dahlhaus, Carl: Nineteenth-Century Music. Translated by J. Bradford Robinson. Berkeley / Los Angeles 1989. Dahlhaus 1991 Dahlhaus, Carl: Ludwig van Beethoven: Approaches to His Music. Translated by Mary Whittall. Oxford 1991. Hatten 1991 Hatten, Robert: “On Narrativity in Music: Expressive Genres and Levels of Discourse in Beethoven.” In: Indiana Theory Review 12, 1991:75–98. Heinemann 1991 Heinemann, Michael: “Lesen Sie nur Shakespeares Sturm!” In: Musica 45.5, 1991:301–302. Jander 1996 Jander, Owen: “Genius in the Arena of Charlatanry: The First Movement of Beethoven’s ‘Tempest’ Sonata in Cultural Context.” In: Alm, Irene/McLamore, Alyson / Reardon, Colleen (eds.): Musica franca: Essays in Honor of Frank A. D’Accone. Stuyvesant, NY 1996:585–630. Jones 1999 Jones, Timothy: Beethoven: The “Moonlight” and Other Sonatas, op. 27 and op. 31. Cambridge 1999.

80

Douglass Seaton

Kinderman 1992 Kinderman, William: “Integration and Narrative Design in Beethoven’s Piano Sonata in A-flat Major, Op. 110.” In: Beethoven Forum 1, 1992: 111–145. Kivy 2002 Kivy, Peter: Introduction to a Philosophy of Music. Oxford 2002. Kramer 1991 Kramer, Lawrence: “Musical Narratology: A Theoretical Outline.” In: Indiana Theory Review 12, 1991:141–162. Kramer 1998 Kramer, Lawrence: “Beethoven’s Tempest Sonata: Musical Meaning and Enlightenment Anthropology.” In: Beethoven Forum 6, 1998:31–65. Leichtentritt 1948 Leichtentritt, Hugo: Musikalische Formenlehre. 3rd edition, Leipzig 1948. Marx 1902 Marx, Adolf Bernhard: Ludwig van Beethoven: Leben und Schaffen. Leipzig 1902. Maus 1991 Maus, Fred Everett: “Music as Narrative.” In: Indiana Theory Review 12, 1991:1–34. Maus 1997 Maus, Fred Everett: “Narrative, Drama, and Emotion in Instrumental Music.” In: Journal of Aesthetics and Art History 55.3, 1997:293–303. Maus 2001 Maus, Fred Everett: “Narratology, narrativity.” In: Sadie, Stanley (ed.): The New Grove Dictionary of Music and Musicians. 2nd edition, vol. 17, London, 2001:641–643. McClary 1997 McClary, Susan: “The Impromptu That Trod on a Loaf: or How Music Tells Stories.” In: Narrative 5.1, 1997:20–35. Misch 1950 Misch, Ludwig: “Das ‘Problem’ der d-moll Sonate von Beethoven.” In: L.M.: Beethoven-Studien. Berlin 1950:42–55. Nattiez 1990 Nattiez, Jean-Jacques: “Can One Speak of Narrativity in Music?” In: Journal of the Royal Musical Association 115, 1990:240–257. Neubauer 1997 Neubauer, John: “Tales of Hoffmann and Others: On Narrativizations of Instrumental Music.” In: Lagerroth, Ulla-Britta / Lund, Hans / Hedling, Erik (eds.): Interart Poetics: Essays on the Interrelations of the Arts and Media. Amsterdam 1997:117–36. Pembaur 1915 Pembaur, Josef: Ludwig van Beethovens Sonaten op. 31 Nr. 2 u. op. 57. München 1915. Rosen 1971 Rosen, Charles: The Classic Style: Haydn, Mozart, Beethoven. New York 1971. Schering 1934 Schering, Arnold: Beethoven in neuer Deutung. Leipzig 1934.

Narrative in Music

81

Schindler 1966 Schindler, Anton Felix: Beethoven as I Knew Him. Translated by Constance S. Jolly and edited by Donald W. MacArdle. Chapel Hill 1966. Schindler 1970 Schindler, Anton Felix: Biographie von Ludwig van Beethoven. 4th edition, Münster 1871; repr. Hildesheim 1970. Schmalfeldt 1995 Schmalfeldt, Janet: “Form as the Process of Becoming: The BeethovenHegelian Tradition and the ‘Tempest’ Sonata.” In: Beethoven Forum 4, 1995:37–71. Tovey 1984 Tovey, Donald Francis: A Companion to Beethoven’s Pianoforte Sonatas. London 1984. Wolf 2002 Wolf, Werner: “Das Problem der Narrativität in Literatur, bildender Kunst und Musik: Ein Beitrag zu einer intermedialen Erzähltheorie.” In: Nünning, Vera / Nünning, Ansgar (eds.): Erzähltheorie transgenerisch, intermedial, interdisziplinär. Trier 2002:23–104.

Metalepsis as a Transgeneric and Transmedial Phenomenon A Case Study of the Possibilities of ‘Exporting’ Narratological Concepts1 Werner Wolf (Graz) 1. Introduction: The Currently Increasing Export of Narratological Concepts into other Disciplines Narratology is a thriving discipline which has contributed many concepts and technical terms to literary studies, and, furthermore, is a discipline that deals with one of the most fundamental and general activities of ‘man [as] the story-telling animal’ (Swift 1992:62). In view of this it is hardly surprising that narratological terms and concepts which originally referred to narrator-transmitted verbal storytelling (embodied for many scholars in print-mediated fiction)2 have frequently—and in the past few decades increasingly—been ‘exported’ to other theories and fields of knowledge. Thus, ‘metafiction’ and ‘mise en abyme’ have by now become widely-used notions that are also encountered, for instance, in art history (Stoichita 1997:158), whilst ‘story’ has been discussed in drama theory (Pfister 1977: chap. 6.1.1.2), ‘point of view’ is employed in film theory (Hickethier 2001: 130), and the terms ‘narrating voice’ and ‘plot’ have also been used in musicological discussions of instrumental music (Abbate 1991:19; McClary 1997:29). Some of these ‘exports’ are obvious and unproblematic; others, however, seem less so. Thus, it is, for instance, problematic, when the ‘story-discourse’ opposition is promoted to a general dichotomy in the interpretation of lyric poetry, as I once heard in a lecture by a colleague of mine, or when the notion of ‘narrative situations’ is not only applied to films (see Hickethier 2001:130) but also to pictures (see Mosthaf 2000: chap. 2.6.1).

1

2

My thanks for careful proof-reading are once again due to Ingrid Pfandl-Buchegger and Sarah Mercer. Recently, the often exclusive focus on literary fiction has, however, been criticized; cf. Fludernik 1996, and Harald Weilnböck’s contribution to the present volume.

84

Werner Wolf

The current export of narratological notions is a process which has certainly not come to an end yet. Considering the debatable exports, but also generally speaking, it is therefore high time to ask under what circumstances such notional transfer makes sense in the first place. I propose to do this in the following, using the example of ‘metalepsis’, i.e. paradoxical transgressions between (onto-)logical levels suggested by works of various media. I have chosen this phenomenon because research has so far largely ignored its transgeneric and transmedial3 nature and, as a consequence, has neglected the relevance of the corresponding term beyond the realm of narratology. In fact, since its first introduction into contemporary narratology by Gérard Genette (1972:243–251), this term and parallel designations that refer to the same phenomenon have continued to be employed almost exclusively4 with reference to literature, and what is more, to fiction as a specific literary genre. This can be seen in the treatment of the subject by, for example, David Lodge, who speaks of ‘short circuit’ (1977:239–245), David Herman (1997) and others,5 and it was only in the recent ‘Colloque International: la métalepse aujourd’hui’ held in Paris in November 2002 that a remarkable number of papers dealt with non-fictional topics. 2. Conditions for a Fruitful Export of Narratological Concepts Before proposing some conditions that render the application of narratological terms and concepts beyond verbal storytelling fruitful and meaningful, I would like to outline the framework or the preconditions within which I will put forward my proposals. 3

4

5

‘Transmediality’, as I define it, is a subform of ‘intermediality’ (‘intermediality’ designates all kinds of relationships between different media). The distinct quality of ‘transmedial phenomena’—or a transmedial perspective—is the fact that similar phenomena occur in more than one medium and that a possible origin in one medium and an ‘intermedial transposition’ from this origin into another medium is either uninteresting or unknown (see Wolf 2002a:18f). Thus, the imitation of an ‘epic’ narrator in ‘epic’ drama is best regarded not as transmediality but as a case of intermedial ‘transposition’ (since the origin in narrative fiction is clear and illuminating for the export domain); in contrast to this, the origin of ‘metalepsis’, as I will argue, may be verbal fiction, but as it occurs in so many media independently of fiction, the classification as a ‘transmedial’ phenomenon seems to be more appropriate. Among the exceptions must be mentioned as one of the earliest and most recent studies that have attempted to widen the scope of the application of ‘metalepsis’: Douglas Hofstadter’s Gödel, Escher, Bach (1980), Marie-Laure Ryan’s “Logique culturelle de la métalepse, ou: la métalepse dans tous ses états” (forthcoming) and Genette’s brand-new book Métalepse (2004, which appeared after the completion of the manuscript of the present essay). Cf., e.g., Nash 1987:94, McHale 1987 (‘Strange loops, or metalepsis’ [119–121], ‘Shortcircuit’ [213–215]), Wagner 2002, and Malina 2002, who claims that ‘metalepsis inevitably operates within the terms of narrative’ (133).

Exporting Narratological Concepts

85

In accordance with the focus of the present volume I shall only discuss concepts that historically or essentially derive from narratology. This means that my remarks do not extend to notions such as ‘polyphony’ or ‘framing’ that stem from other disciplines (in these two cases from musicology and art history or cognitive psychology), nor can I include transmedial phenomena whose conceptualization has no clear narratological origin, such as ‘character’, ‘event’, ‘closure’, ‘suspense’, etc. Another precondition is my assumption of an at least pragmatically justified differentiation of genres, media and concomitant theoretical discourses. This assumption deserves to be mentioned in the face of deconstructive tendencies among some scholars who are unhappy with ‘essentialist’ distinctions between the media (cf. Goodman 1968; Prieto 2002:60f). As a consequence, such deconstruction would also blur the boundaries between narratology and other disciplines and render the question of the ‘export’ of concepts across disciplinary boundaries futile. Finally, a third precondition concerns the mode in which I propose to conceptualize at least some, if not all of the notional ‘export candidates’: if one wants to consider the exportability of narratological notions beyond the realm of narratology, it is helpful not to start from an overly rigid conceptualization. In particular, one should avoid a conceptualization that would lead to a simple, binary inclusion-exclusion mechanism. Instead, it is advisable to conceive of the phenomenon under consideration along the lines of prototype theory. This implies, first, that one should describe typical features or constituents of the phenomenon as discussed in narratology, preferably using examples which scholars in the field have repeatedly regarded as prototypical. This prototype then would provide a basis that permits a comparison with phenomena outside verbal narratives, which, as a result, could be pronounced to be more or less similar to the prototype. In contradistinction to an approach that would yield binary results (‘yes—no’), this prototypical approach allows us to describe certain phenomena as ‘more or less’ like the prototype, according to the recognizability of more or fewer of its typical features in the object under discussion. In many cases it will depend on the respective frames and intentions of the theorist to what extent he or she accepts deviations from the prototype, while still being willing to use the original concept and term. Lest this opening-up of terminology be viewed as a first step in a direction where finally ‘anything goes’, let me hasten to add the following: While I do think that using the same concepts and terms in an inter- or transdisciplinary context inevitably requires some readiness to allow for a certain breadth and in many cases metaphoricity in their use, this certainly does not mean that I hereby advocate a terminological permissiveness or carelessness. In particular, I do not advocate a dismissal of ‘appropriateness’

86

Werner Wolf

or ‘adequacy’ in favour of ‘surprise’ and ‘cognitive dissonance’ as criteria for the use of terminology, as has been suggested in a similar discussion in a musico-literary context (Prieto 2002:51). So, within this framework, what are the conditions and potentials that enable us to meaningfully employ narratological terminology and concepts for phenomena that can be found in fields other than verbal storytelling?6 I would like to propose the following four criteria: 1. First of all, there is the question of the export-facilitating potential of the phenomena under consideration. Depending on the intensity of their links with the genre of narrator-transmitted stories and the verbal medium, phenomena can be located on a continuum between two poles, a continuum which represents the range of their export potential: On the one hand there are phenomena that are more genre- or mediaspecific, such as the narrator who tells a story, phenomena that are intrinsically related to the defining core of verbal storytelling. Consequently, the corresponding concepts will possess, in their substance, an export potential that extends, but is also restricted, to genres and media that involve narrating voices in the constitution and transmission of their storyworlds.7 On the other hand there are phenomena that are rather more transgeneric and ‘transmedial’. Their conceptualization may have originated in narratology, too, but it is characteristic of them to have only a loose or accidental connection with verbal narratives. The corresponding terms are, therefore, in principle also exportable to areas beyond verbal storytelling and even to non-narrative fields. ‘Metalepsis’ is an example of the latter, trans-narrative type with an extended export potential. 2. For both types of phenomena a clear narratological conceptualization and description of typical features is a precondition of a meaningful export in which the exported concept remains recognizable. As is well-known, the lack of such clarity is often a deplorable reality in our discipline, extending even to its most basic notion, ‘narrative’.8 Yet, clarity does not mean over-restrictiveness: in accordance with what has been said above, as long as terminological precision is not impaired, the conceptualization should be sufficiently flexible so that the notion under consideration retains the maximum of the export potential it may have.

6

7

8

In a semiotic framework ‘term’, ‘concept’ and ‘phenomenon’ correspond to ‘signifier’, ‘signified’ and ‘referent’ respectively. Thus, a concept such as ‘narratorial agency’ that focusses on the narrating voice can be exported from verbal storytelling to ‘epic’ drama, to the ballad as a narrative sub-genre of poetry or—with some restrictions—to the ‘voice over’ in film, but not to painting. For an illuminating attempt at clarifying narrativity as a transmedial phenomenon see Ryan’s contribution to this volume; cf. also Wolf 2002b and 2003.


87

3. Another condition refers to the formal appropriateness of the narratological concept for the target phenomenon in the import domain: Such an appropriateness can be assumed if there is a high degree of similarity between a typical phenomenon in the export domain for whose identification the concept was originally used, i.e. verbal narratives, and the phenomenon in the import field for which an application of the concept and corresponding term is being considered. This similarity can be analysed by comparing the distinctive features of both phenomena. With a view to the afore-mentioned prototypical approach, it should be noted that ‘similarity’ is here a matter of degree in the equivalence between the characteristic traits of the two phenomena. The result of such a comparison will quite frequently not be a verdict of ‘appropriate’ or ‘inappropriate’ but of ‘more or less appropriate’, and opinions may differ as to where to draw the line for a conceptual export. At any rate, some ‘cognitive dissonance’ and metaphoricity should be permitted, and in some cases of apparent surface dissimilarity the possibility of a deep-structural convergence should also be considered. The degree of similarity will not only depend on individual phenomena as such but also on the medial or generic contexts in which they appear. Clearly, narrative genres and media outside verbal stories will present more favourable conditions for the import of narratological concepts than non-narrative genres and media. (It should go without saying that here, as elsewhere in my remarks, ‘narrative’ is not restricted to verbal stories nor to narrative fiction, but extends to other genres and / or media as well, such as drama, the comic strip, film, or certain [series of] pictures.) 4. A final export condition is the heuristic value of the exported notion for the use in the import field. While the manifold possibilities of what may constitute a ‘heuristic value’ elude holistic description, because they depend too much on individual cases and interests, it can generally be said that, given the afore-mentioned conditions, it greatly contributes to the justification of a notional export if no adequate term has as yet been coined in the theory of the import domain. Until recently it appears that this was for instance the case for what nowadays is being described as ‘mise en abyme’ in art history. The use of this originally narratological term can fruitfully highlight practices of embedding pictures in pictures, which Victor Stoichita (see 1997) has described as a frequent feature of what he calls ‘meta-paintings’. A similar fruitfulness can be claimed for the export of ‘metalepsis’, as I hope to be able to show. Even where, in some cases, an ‘indigenous’ terminus technicus is already in use for a given phenomenon in an import field, it can still be illuminating to borrow from narratology in order to highlight, e.g., a common range of functions of the designated phenomenon or a stylistic or other similarity

88

Werner Wolf

among media in a given epoch. This is, for example, the case for some nineteenth-century instrumental music, for whose technical analysis there exists a panoply of musicological terms. However, in some cases (including the second movement of Beethoven’s fourth Piano Concerto; see Wolf 2002b:86–93) the application of narratological terms, and even if it is only the term ‘narrative’, can shed light on cultural contexts and musical effects that would elude a strictly musicological analysis. As the enumeration of ‘export conditions’ for the use of narratological notions in fields other than verbal stories may seem both too theoretical and too vague for the reader, I propose to test them in at least one casestudy, the case of ‘metalepsis’. 3. The Case of ‘Metalepsis’ 3.1 Typical Features of Narrative ‘Metalepsis’ The notion ‘metalepsis’ has occasionally occured in rhetoric since Antiquity—albeit with rather imprecise meanings.9 In contrast to this, and as already mentioned, the first extensive discussion of ‘metalepsis’ in the sense that is relevant here can be found in Genette’s research, more precisely in his influential treatise Figures III (1972:243–251). Thus, metalepsis as explained in the following is clearly an originally narratological concept. Unfortunately, Genette himself does not really provide an explicit formal definition of ‘metalepsis’ (nor does he do so in his 2004 monograph on metalpsis). Yet, his description of the phenomenon, which was based on some examples from the field of ‘high-brow’ narrative fiction,10 as well as subsequent research (in particular Herman 1997, Wagner 2002 and Ryan forthcoming11) enable us to present a clear notion of narrative metalepsis and its typical features as a starting point for the consideration of its export. Thus, a central export-facilitating condition (no. 2 in the above catalogue) is fulfilled. By way of illustration of these typical features let me briefly analyse one of the especially pertinent examples given by Genette (cf. 1972:244) as a 9

10

11

Hollander/Fletcher, in their entry on ‘metalepsis’ mention ‘Quintilian (8.6.37) and later rhetoricians’, in whose writings ‘metalepsis’ appears as ‘an obscure and minor trope, variously defined, ›a change from one trope to another‹, often moving through an associative chain’ (1993:759). He mentions, among other examples, the metaleptic classic, Julio Cortázar’s short story “Continuidad de los parques” (1956, “Continuity of Parks”), and some passages from Sterne’s Tristram Shandy (for one of these see below). Cf. also Lodge 1977:239–245, Genette 1983:58 f. (where he rephrases his 1972 description of metalepsis without substantial alterations), and McHale 1987:119–121, 213–215.


89

prototypical instance: It is an episode in Laurence Sterne’s novel Tristram Shandy (1768), where the narrator Tristram, owing to his constant digressions, despairs of ‘get[ting] my father and my uncle Toby off the stairs [of Shandy Hall], and [...] put[ting] them to bed’ (IV / 13, p. 285). Tristram’s solution to the problem is to call for a ‘chairman’ who should ‘step into [a] bookseller’s shop, and call [...] a day-tall critic’ (ibid.) to help him with his story. This proves to be successful indeed, since at the end of the chapter Tristram is able to address the critic as follows: ‘– So then, friend! You have got my father and my uncle Toby off the stairs, and seen them to bed? [...] Here’s a crown for your trouble’ (287). In deducing characteristic traits of metalepsis from this example and previous research, it should be noted that I am here only dealing with the ‘genuine’ type of metalepsis, which Ryan has called ‘ontological metalepsis’ as opposed to ‘rhetorical metalepsis’ (see Ryan forthcoming and note 19, below). Further, in accordance with the afore-mentioned condition of formulating, where possible, narratological concepts in a flexible, exportfacilitating way, I shall take care to avoid unnecessary restrictions that would not be justified by the phenomenon under consideration. On the basis of these conditions four typical features of metalepsis can be distinguished: 1. the occurrence of metalepsis within artefacts, or—with a view to drama—performances that represent possible worlds in Ryan’s sense (1991);12 as the example of Tristram’s autobiography shows, such possible worlds can be of a narrative nature; yet, as the further traits of metalepsis will reveal, metalepsis is not essentially linked with narrativity, and so a restriction to storyworlds would seem unnecessary; 2. the existence of, or reference to, recognizable, logically distinct levels13 or possible (sub)worlds within these artefacts or performances; these levels or (sub)worlds usually differ ontologically from each other and can be distinguished—as frequently happens in works involving a mise en abyme14—by 12

13

14

This trait excludes some of Ryan’s quasi-metaleptic examples (see 2003), taken from logic and mathematics, from the core of metalepsis, as they do not occur in representations of possible worlds. It is advisable to include ‘level’ as an alternative to ‘possible (sub-)world’ in the traits of metalepsis in order to account for ‘impossible’ triangles, which represent (and therefore are representational!) three dimensions (spatial levels) in the manner of a möbius-‘strip’ so that one can visually follow a paradoxical move from one spatial level to another without having crossed an ‘edge’. There are in fact points of contact between mise en abyme and ‘metalepsis’: The first applies to the majority of metalepses (i.e. those that involve a hierarchical structure of levels, as opposed to parallel possible worlds, see below, note 16): such metalepses always also imply an at least formal mise en abyme, e.g. the recursive appearance of a novel within a novel, a play within a play, a picture within a picture, or a film within a film (such embeddings will also frequently appear in the examples discussed below).

90

Werner Wolf

the opposition ‘reality vs. fiction’15 or ‘actual present vs. evoked past’16 (in our example, the difference is between the diegetic world of the past, in which the story of Tristram’s father and of his uncle Toby’s unfolds, and the extradiegetic ‘present’ world of Tristram as narrator—a world in which we are also asked to imagine the existence of the chairman and the critic); again, these levels or (sub)worlds can include narrative levels, but are not limited to them;17 3. an actual transgression between, or a confusion and contamination of, the (sub)worlds involved (in our case the main transgression resides in the effect which the extradiegetic critic has on the diegetic story);18 4. the paradoxical nature of this transgression or confusion with reference to a ‘natural’ or conventional belief in the inviolability, in ‘normal’ life and ‘normal’ fiction, of the (sub)worlds or levels involved.19 This viola-

15

16

17

18

19

The second point of contact is a special kind of mise en abyme, which Dällenbach has called ‘réflexion paradoxale’ (‘un fragment censé inclure l’oeuvre qui l’inclut’; 1977:38, 51), and to which Ryan has also drawn attention (see Ryan forthcoming). It should, however, be noted, that mise en abyme and metalepsis are nevertheless distinct: for the two cases of contact just mentioned are incidental and not necessary traits of the phenomena involved (there are also non-hierarchical metalepses between parallel worlds, and there are non-paradoxical mises en abyme). In addition, representationality is a distinct trait of metalepsis, but not of mise en abyme, since mise en abyme can also occur in recursively embedded purely geometrical forms that do not represent anything. ‘Fiction’ as used here (in its non-generic sense), covers two meanings: fiction as ‘artefact’ (which I call ‘fictio’) and ‘fiction’ as ‘invention without direct reference to reality’ (which I call ‘fictum’; see Wolf 1993:38f). In most cases the distinctness of the levels or possible (sub)worlds involve a ‘vertical’ hierarchical relation, which may be visualized by ‘stacks’ (cf. Ryan forthcoming); yet, Wagner (2002:247) has pertinently pointed out that in some cases such possible (sub)worlds may also be parallel universes that cannot normally have any contact with each other (e.g. two fictional hypodiegetic novels by different diegetic fictional authors) and that therefore metalepsis can also operate in a lateral or horizontal direction. I here differ from Wagner 2002, who—from a point of view restricted to fiction—speaks of ‘niveaux narratifs’ as the common denominator of metalepsis (243). Nash (1987:54) unduly restricts ‘metalepsis’ to a ›leap‹ between histoire and discours’, which is an even narrower conception than Genette’s. Genette, again from a purely fiction-centred perspective, speaks of ‘[une] intrusion du narrateur ou du narrataire extradiégétique dans l’univers diégétique (ou de personnages diégétiques dans un univers métadiégétique, etc.) ou inversement [...]’ (1972:244). Curiously, one of his examples (cf. ibid.), the reference to chapter I/4 of Tristram Shandy, where the extradiegetic narrator asks the—equally extradiegetic (!)—fictional reader to ‘Shut the door’ (38), does not fulfil the criterion of transgression. Genette calls this ‘un effet de bizarrerie’ (1972:244) and links it to ‘[un] mépris de la vraisemblance’ (245). Wagner is right to specify that the awareness of the metaleptic paradox depends on ‘une norme particulière qui, pour ancienne qu’elle soit puisqu’elle remonte à l’Antiquité, ne saurait prétendre ni à l’éternité ni à l’universalité [...] la mimésis [...]’ (237). At any rate it is the paradoxical, implausible nature of an ‘impossible’ transgression which forms what Ryan (forthcoming) calls ‘ontological’ metalepsis—as opposed to a ‘rhetorical’ variant, which can be rendered plausible (e.g. if a narrator, in the act of storytelling, is shown to be emotionally affected by happenings of his or her story).


91

tion of normal expections, as a rule, is the result of an intentional act and usually serves as an indicator of the fictionality of the represented world in its entirety.20 Thus, according to the convention of regarding fictional narratives as if they were ruled by the same laws that apply to factual storytelling and actuality, it would be impossible that a narrator and a helper of his can interfere with happenings which occurred in a storyworld of the past, as is in fact suggested in the episode from Sterne’s Tristram Shandy. The paradox involved in (ontological) metalepsis often has a startling or also comic effect, which is the main reason why it occurs so frequently in ‘experimental’ texts (such as Sterne’s famous novel) that lay bare their fictionality, but also in comedies, ever since Greek Antiquity (as ‘parabasis’ and ‘out-of-character’ addresses to the audience). Taking these four traits together, ‘metalepsis’ may be defined as a usually intentional paradoxical transgression of, or confusion between, (onto)logically distinct (sub)worlds and / or levels that exist, or are referred to, within representations of possible worlds. It will be noted that, in accordance with my previously mentioned classification of ‘metalepsis’ as a notion that does not involve the essence of narrativity, I have not restricted ‘metalepsis’ in this definition to narratives, let alone print-mediated fiction, but permit an inclusion of a wider class of artefacts and performances, namely representations and performances of possible worlds. Although for historical reasons part of a novel was used as an illustration, there is in fact no need for a restriction to narratives, since none of the typical traits of metalepsis requires the existence of a context with a temporal, causal and teleological dimension that distinguishes narratives from other forms of representation, such as descriptions. 20

There are, however, interesting borderline cases. One of them was highlighted during the discussion following the presentation of Ute E. Eisen’s paper at the conference whose proceedings are published in the present volume. If one, in the spirit of her paper and contribution to this volume, applies the narratological notion of metalepsis to the Bible, the so-called ‘we passages’ in Acts present a problem. If Acts is regarded as a sequel to the gospel according to St Luke, a curious inconsistency in the presentation of the narrator occurs: according to the preamble of the gospel, it is told by a heterodiegetic narrator; in contrast to this, the we-passages in Acts (16, 10–17; 20, 515; 21, 1–18; 27; 28, 1–15 [I am grateful to Ute E. Eisen for precise information]) betray a homodiegetic narrator. At first glance, this ‘inconsistency’ could be regarded as an (unwitting) metalepsis that points to the fictionality of the represented storyworlds (provided one accepts the unity of the two texts). When we take a closer look at the text, this possibility can, however, be rejected: after all, the participation of the narrator in his story only refers to the travels of St Paul and not to Christ’s life (as narrated in the gospel), and there is no reason why the author of Acts should not be heterodiegetic with reference to Christ’s biography (as recounted in the gospel), yet homodiegetic with reference to a part of the story of the Apostles (as recounted in Acts), which took place at a later period.

92

Werner Wolf

Admittedly, the four traits of metalepsis which have emerged from a discussion of the export-conditions nos. 1 and 2, namely from the nature of the phenomenon and the conceptualization of the narratological concept, allow a number of variants (nothing has, for instance, been specified concerning the direction of the transgression—whether it procedes from ‘above’ or from ‘below’, or in a lateral direction—, how frequently it can occur, etc.). However, the purpose of this paper, which aims to show the exportability of ‘metalepsis’ from its home-domain, the analysis of ‘highbrow’ written fiction, to other genres and media, precludes the detailed description of a typology of metalepsis.21 In the following, I propose to illustrate this exportability by first focussing on export condition no. 3, the formal appropriateness of using the term ‘metalepsis’ for various phenomena in domains further and further removed from the home domain. After that I will synoptically deal with export condition no. 4, the heuristic value that justifies a terminological and conceptual export in the first place. As will be seen, all of the following examples are representations of possible worlds, so I need not point out the existence of this feature of metalepsis separately in each case. 3.2 The Transgeneric and Transmedial Applicability of ‘Metalepsis’ to Fields Outside ‘High-brow’ Narrative Fiction The first ‘export field’ is still quite close to the area for which ‘metalepsis’ was originally conceptualized, namely ‘high-brow fiction’. With a view to the subsequent heuristic and functional evaluation of metalepsis this field shall only be mentioned as an example of more ‘popular fiction’ (to which one could add science fiction22), namely children’s fiction. A tendential metalepsis already occurs in a classic of British children’s fiction, Lewis Carroll’s sequel to his famous Alice in Wonderland (1865), Through the Looking Glass (1872), a remarkable metafictional, metalinguistic and epistemological frame tale that questions all sorts of received beliefs (cf. Wolf 1987). In the chapter ‘Tweedledum and Tweedledee’ Alice is confronted in her dream with Tweedledee’s claim that she is ‘a sort of thing in [the Red King's] dream’ (Carroll 1970:238) and that, according to Tweedledum, she is therefore ‘not real’ (239). The question ‘who it was that dreamed it all’ (343), which is actually a question of authorship, still haunts Alice after she has awoken from her dream. It remains pointedly unanswered up to 21

22

For various proposals for a typology of metaleptic forms see Wolf 1993: chap. 3.5.4, Wagner 2002, Malina 2002 (although her typology is unsystematic), Ryan forthcoming, and Fludernik 2003. A particularly intriguing example is Howard Schoenfeld’s short story “Build Up Logically” (1950), which I have discussed elsewhere, erroneously changing the author’s name into ‘Schoenbaum’ (cf. Wolf 1993:368–370).


93

the very last question mark of the main text, which ends with the narrator addressing the reader: ‘Which do you think it was?’ (344) The whole situation bears typical features of metalepsis, namely the paradoxical confusion of distinct worlds: As opposed to the waking Alice (A1) of the diegetic framing (level 1), Alice, in her dream about the looking glass world (level 2), is a hypodiegetic character (A2), who is confronted with the idea that she is at the same time also a hypo-hypodiegetic character (A3), namely the object of another hypodiegetic character's dream world (level 3). The paradox does not reside in Alice allegedly being a character in the dream of another looking-glass world character (which is after all possible), but rather in the fact that she is threatened by irreality, when Tweedledee says that she would cease to exist if the Red King awoke from his dream. In other words, Tweedledee paradoxically claims that A2 depends on A3 and thus on a being that is located on an ontologically lower level governed by the Red King. However, this is only a tendential metalepsis, since the confusion of subworlds (levels 2 and 3) is here merely an idea, not a ‘reality’ enacted in the fictional world. Moreover, this confusion is naturalized by the fact that even as a mere idea it occurs only in a dream (on level 2), where notoriously all sorts of illogicalities may happen. A much more extended use of metalepsis in children’s fiction can be observed in another frame tale, Michael Ende’s metafictional novel Die unendliche Geschichte (1979). In this intricate self-reflexive allegory—and praise—of the effects of storyreading, the hero, ten-year old Bastian Bux, is shown in the frame to be reading a fairy tale book, whose title Die unendliche Geschichte is an obvious mise en abyme of the title of Ende’s own novel. The distinctness of both (sub)worlds is underlined by the use of different colours for the passages referring to either subworld. Yet, from p. 81 onwards, curious transgressions between the frame and the framed story occur: at first it is ‘only’ a cry uttered by the empathetic reader Bastian which reverberates in the fairy-tale world of ‘Phantasien’, but shortly after this Bastian is even able to access this world in person. Although this ‘impossible’ fact of a diegetic character entering a hypodiegetic story can be explained as an allegory of aesthetic illusion, the text, unlike Through the Looking Glass, virtually suggests the ‘reality’ of such a transgression and thus unquestionably fulfills all the typical features of a metalepsis. Another kind of metaleptic leap between a possible world written for children and, in this case, the world of the extratextual reader occurs as an integral part of the reading process in Edward Packard’s story of adventure Sugarcane Island (1976). The text is written in the present tense and is hence a ‘simultaneous narration’. Its protagonist is moreover consistently addressed in the form of second person singular pronouns. The effect of this device is not only to offer a strong incentive to the reader

94

Werner Wolf

to identify with this protagonist but also a paradoxical double reference of the pronoun ‘you’: it here not only refers to an extradiegetic addressee (as would be the case in traditional narratives), but also to an agent in the intradiegetic fictional world (analogous to the I-as-experiencer in a first-person narration). As in other narratives written in the second person (including Michel Butor’s classic La Modification [1957]) the use of the pronoun ‘you’ with this double reference could here already be classified as a tendential metalepsis: for it positions an addressee who exists outside a fictional world simultaneously (as the employment of the present tense suggests) inside it. Yet, Packard’s text goes further and by this constitutes a prefiguration of, or a literary parallel to, interactive computer games and certain forms of hyperfiction23: it offers the reader a series of possible choices, e.g., whether he or she, after having been stranded on an unknown island, decides to explore it by walking along the beach or by climbing up a rocky hill. Depending on the choice made, the text invites the reader to continue on different pages, and this eventually leads to a plethora of separate endings, some of which even include the reader’s death. The suggestion that the recipient can thus interfere with the action is diametrically opposed to traditional conceptions of fixed storyworlds and constitutes a ‘para-dox’ at least in the etymological sense of the term as ‘something that runs counter to established notions’. For normally a reader may become immersed in a represented possible world (this state of aesthetic illusion is apparently intended in this text, too) but only in a basically passive way, i.e. without being able to modify it. In the last three examples metalepsis was only ‘exported’ from the theory of high-brow fiction to the description of children’s fiction. A more marked ‘export’ occurs when metalepsis is said to occur in drama, that is, in another literary genre or even medium. It may suffice to just recall metadramatic works such as Luigi Pirandello’s Sei personaggi in cerca d’autore (1921) or Tom Stoppard’s The Real Inspector Hound (1968) to show that ‘metalepsis’ is indeed applicable here, too. Owing to a playwithin-a play structure,24 both fictional representations possess ontologically different worlds which in both cases are paradoxically confused, a confusion of levels for whose description one may export yet a further narratological concept (this time into drama theory), namely the opposition of ‘diegetic’ vs. ‘hypodiegetic’ levels. 23

24

The form relevant in this case is what Helbig, in the framework of a general typology of hyperfiction, calls a ‘reader-guided’ or ‘collaborative hyperfiction’ (2003:304 and 310). Formal mise en abyme structures of the type ‘tale within a tale’ or ‘play within a play’ are indeed frequent in metaleptic possible worlds, but they are not a precondition, as the example from Tristram Shandy shows.


95

In Pirandello’s play this transgression happens, for instance, when characters devised by an anonymous fictional author, hence hypodiegetic characters, appear on the diegetic stage of a fictitious theatre and ask for a ‘realization’ so that they can re-enact their sad story. This ‘impossible’ contamination of two separate worlds is clearly a metalepsis. In addition, narratologically speaking, these characters appear as mere story elements and request that they be given a discourse. The fact that story elements never actually can appear without a discourse adds to the paradoxical, metaleptic nature of a situation whose description, thus, shows a further possibility of exporting narratological terms. As for Stoppard’s The Real Inspector Hound, the framing diegetic situation is here equally a theatre. In this fictional theatre a whodunnit is performed, witnessed by an audience which includes two theatre critics. In the course of the embedded performance these critics become paradoxically involved in the hypdodiegetic play within a play, an involvement which even leads to the death of one of them. Thus, as in the case of Pirandello’s Sei personaggi, the typical traits of a metalepsis can here also be recognized: a fictional representation consisting of several distinct worlds and levels, among which unorthodox transgressions occur. While drama could arguably still be classified as a literary genre, we definitely deal with another medium when referring to a comic strip. Surpisingly, metalepsis can occur here, too. This is, for instance, the case in a Walt Disney-licenced story entitled, in its German original, “Der Vielfraß aus dem All” (1996). In this strange, self-reflexive comic, a small, voracious and virtually omnivorous alien suddenly emerges from a meteorite, whose impact Donald Duck witnesses during a picknick outing with Daisy. The little alien proves to be such a nuisance that he is captured by the local authorities, but he thwarts all endeavours to stop his destructive appetite, including an attempt at exiling him lastingly to another celestial body. Finally, Donald has what he thinks is a clever idea, and it is at this moment that the first metalepsis occurs: he shows the alien a sandwich, lifts the bottom left corner of the following drawing and throws the sandwich out of the picture—no doubt an extremely surprising and paradoxical transgression between the diegetic, representational level of the story and its material framing (illustration 1).25 25

The paradox is even overtly thematized on the diegetic level, when Donald admits: ‘[...] ich hab’ ihn aus dem Comic rausgeworfen! Ich weiß, daß das nicht ganz richtig war, aber das war die einzig mögliche Lösung’ and Daniel Düsentrieb answers: ‘Das ist unmöglich!’ (Disney 1996:169). Interestingly, metalepsis has recently been used in a similar way in David Wiesner’s illustrated postmodernist version of The Three Pigs (2001), in which, as a prelude to further complications, the wolf inadvertedly blows one piglet after the other ‘out of the story’, i.e. out of the frames of the illustrations (I am grateful to Malte Stein/Hamburg for having drawn my attention to this text).

96

Werner Wolf

Illustration 1: “Der Vielfraß aus dem All” (Disney 1996:169)

However, this solution misfires, too, as the alien, from his seemingly extradiegetic position, starts to nibble away the comic strip pictures on the following page, leaving less and less diegetic space for Donald’s world. The final twist of this funny, metafictional story is that at this point Donald awakes from what eventually turns out to have been no more than a dream dreamt during his picknick outing. Owing to the fact that initially the ‘discourse’ did not indicate the down-stepping onto a hypodiegetic level, this surprise can be described once again by means of a conceptual export from narratology, namely by the term ‘missing opening frame’ (Wolf 1992:33).26 Even if we, as in Carroll’s Through the Looking Glass, 26

I coined this term as a complement to Brian McHale’s narratological notion of the ‘missing end-frame’ (1987:117).


97

read the dream as a possibility of ‘naturalizing’ and thus defusing the paradoxical character of the story, the comic strip still shows all the typical features of a metalepsis and thus justifies the application of this narratological term: the story includes the representation of several hierarchically structured levels, and involves repeated ‘illogical’ transgressions between ontologically distinct levels. Interestingly, the illogical nature of these transgressions cannot really be naturalized by the dream structure, for it is hardly probable that anyone will dream of him- or herself as part of a comic strip adventure. Another medium in which metalepsis has occurred, in particular in the past few decades, is film. Examples include The Matrix (1999, directors Andy and Larry Wachowski), Pleasantville (1998, director Garry Ross) and numerous films by Woody Allen. It may here suffice to give just one illustration, namely Woody Allen’s The Purple Rose of Cairo (1984). As with the plays mentioned earlier, this meta-film uses a mise-en-abyme structure, the device of a film within a film, as the precondition of numerous metaleptic surprises. The first of them is doubtless most surprising: The story is centred on a young married woman, Cecilia, who leads a difficult life during the days of the Depression and seeks relief from her drab job and frustrating marriage in the local cinema. Here a movie is shown, whose romantic qualities appeal to Cecilia to an extraordinary extent, ‘The Purple Rose of Cairo’. In particular she admires the actor playing ‘Tom Baxter’. When Cecilia watches this film for the fifth time, her obvious interest elicits the following startling reaction from ‘Tom’: after having ‘look[ed] out at the offscreen audience’ (Allen 1990:350), he realizes that Cecilia is again among the spectators and exclaims, ‘shaking his head, looking out at the offscreen Cecilia: ›My God, you must really love this picture.‹’ (ibid.) This ‘impossible’ apostrophe, by a hypodiegetic character on a canvas, of someone located on a superior, diegetic level, is already technically a metalepsis. Yet, this strange situation becomes even more strange, when, shortly afterwards, Tom decides: ‘I gotta speak to you’ (351) and begins to leave the black-and-white screen. The audience, reacting, begins to gasp. The film cuts to a shocked Cecilia, immobilized in her seat. As the other patrons cry out in the background, the film quickly cuts back to Tom: He actually walks off the black-and-white movie screen turning into living color as he enters the theater (351).

This is the beginning of a series of complications, in the course of which Cecilia becomes entangled in a romance with Tom. He, however, later abandons her, and so she is eventually left with the same poor substitute for real happiness as at the beginning, namely watching the glamorous

98

Werner Wolf

fictions of Hollywood movies. The levels of representation which the metaleptic romance has thus paradoxically confused seem to be re-established in their original, inviolable nature—order reigns again, although it is a sad kind of order. 3.3 The Transmedial Applicability of ‘Metalepsis’ to Fields Outside Narrative Genres and Media In the previous examples taken from children’s fiction, drama, comic strip, and film we still remained within the realm of narrative genres or media. In accordance with the claim that ‘metalepsis’ is one of the narratological concepts that can be exported even beyond the confines of narrativity, at least one example will be given from a medium that is not usually counted among narrative media, namely from a pictorial medium (illustration 2).

Illustration 2: M.C. Escher, Picture Gallery (Escher 1956)


99

Illustration 3: René Magritte, Le Maître du plaisir (Magritte 1928)

M.C. Escher’s lithography Picture Gallery (1956) refers, as the title indicates, to the location of a young man depicted in the left bottom corner of a picture gallery with curiously distorted windows, walls and paintings and gazes at the representation of a city situated at the banks of a river or lake. If we let our gaze travel in a circle towards top left and right, then down to bottom right and back again to the young man, we notice that, paradoxically, the man seems to be contained in the very picture he is gazing at. The picture is clearly not narrative, since the represented ‘scene’ is entirely static (the temporal dynamics which the above description may have implied is restricted to a mere act of reception). And yet the picture features all the typical traits of a metalepsis, as previously outlined: it is clearly the representation of a possible world; it consists of at least two logical levels (a level of represented ‘reality’: the picture gallery, and the level of a represented ‘fiction’ or artefact: the pictures in the gallery); there is a confusion between these levels; and this confusion is an ‘impossible’ paradoxical one. While it would not be difficult to find other examples of metalepsis in the realm of ‘experimental’ pictures both by Escher and others, e.g. Magritte (cf. his Le Maître du plaisir [1928], illustration 3), it would be a challenge to try and extend the enquiry to yet other non-narrative fields.

100

Werner Wolf

Even if there is insufficient space to do so here, the features that have been outlined as typical of metalepsis at least allow some prognosis: Given that, for instance, sculpture as well as lyric poetry are representational media, they would be obvious areas suitable for such an investigation.27 As opposed to them, abstract painting and (instrumental) music would be less obvious areas and would predictably not yield convincing results in a search for metaleptic structures, as in both cases the fundamental condition of representationality is not fulfilled.28 In a similar way, it would be surprising to find genuine (in Ryan’s terminology ‘ontological’) metalepses, for instance, in authentic biography or historiography: while both are 27

At least as far as poetry is concerned, some examples come to mind. Among these the ending of T. S. Eliot’s “The Love Song of J. Alfred Prufrock” (1915/17) should be mentioned as containing a potential metalepsis. Towards the end of the poem the speaker, reflecting on past and future walks on a beach, also imagines a subaquatic world in which mermaids sing to each other, but not to him. Yet, in the last stanza the speaker has unaccountably ‘descended’ into the world of his own imagination and now muses about an unspecified ‘we’: ‘We have lingered in the chambers of the sea/[...]/Till human voices wake us, and we drown’ (Eliot 14). While the paradoxical quality of this change of perspective—and of ontological quality (the human ‘I’ that was opposed to the nonhuman mermaids suddenly becomes part of a group that now appears to include the mermaids and is opposed to humans)—may still be naturalized as the workings of the speaker’s imagination, no such naturalization is possible in one of Christian Morgenstern’s (1871–1914) nonsense-poems: in this poem ‘an aesthetic weasel’ is seen sitting on a pebble near a babbling brook ‘for the sake of the rhyme’—a clear confusion between the levels of enunciation and enounced: Das ästhetische Wiesel Ein Wiesel saß auf einem Kiesel inmitten Bachgeriesel. Wißt ihr, weshalb? Das Mondkalb verriet es mir im Stillen: Das raffinierte Tier tats um des Reimes willen. (Conrady 1978:608)

28

In music a—highly debatable—borderline case could be the recitativo in the last movement of Beethoven’s “Ninth Symphony”, in which a voice admonishes the instruments and the choir: ‘O Freunde, nicht diese Töne, sondern laßt uns angenehmere anstimmen, und freudenvollere!’, an intrusion which has an audible effect on the ensuing musical development. However, this can only be construed as a metalepsis, if one regards this voice as something like an authorial voice coming from a realm beyond the field of instrumental music (which is questionable, since the recitativo melody is prefigured in a preceding instrumental passage). Even so, the metalepsis would be linked to the appearance of a verbal text and hence would not occur within ‘pure’ instrumental music alone.


101

representations, and even typically narrative ones, they nevertheless are bound to the logic of the real world, in which paradoxical metalepses, according to common ‘doxa’, do not occur. However, this would not be a true obstacle to quasi metaleptic ways of representation, in particular not to what Ryan has described as ‘rhetorical metalepsis’.

4. Why Exporting Concepts such as ‘Metalepsis’ to Fields Outside Narratology Matters The preceding survey shows a clear result: from the point of view of formal appropriateness it is indeed possible to export notions such as ‘metalepsis’ from their home domain to other fields and use them for the description of transmedial and transgeneric phenomena that do not only apply to verbal stories. Yet, why would one do so in the first place? Why is it not sufficient to call Escher’s drawing just an ‘illusion’, as has been done (Ernst 1986: 26 and passim), or to refer to what happens in The Purple Rose of Cairo as an ‘artfully constructed picture puzzle game’ (‘kunstvoll konstruierte[s] Vexierspiel’ [Bornemann 1995:474])? This question deals with the heuristic value (and hence the above export condition no. 4) which terminological exports have or can be expected to yield. One heuristic advantage of the export of ‘metalepsis’ is that there are no received ‘indigenous’ terms in use for fields beyond verbal narratives that would describe the phenomena in question with the same precision (‘Illusion’ and ‘Vexierspiel’ are cleary less precise). Another advantage is implied in the fact that the transmedial and transgeneric employment of a term such as ‘metalepsis’ highlights similarities and contact zones between different media and genres. This seems to be a desirable effect indeed, as it can establish a basis for interdisciplinary comparisons and thus enhances our sensitivity for structural and formal variations both within and across genres and media. Using the same notion can also increase the awareness that different cultural practices can contribute to similar purposes and share a common functional potential. In addition, applying a common terminology to at least some of the often deplorably separate disciplines of the humanities facilitates an interdisciplinary discussion and analysis of the arts and media. A perhaps even more important heuristic gain which can be reaped from the use of a common terminology refers to the functional dimension of formally similar phenomena. In fact, it is thanks to the use of a term such as ‘metalepsis’ that one becomes aware of a number of common functions which this device can fulfil in different genres and media. From

102

Werner Wolf

a comparative point of view, Ryan (forthcoming) has already enumerated most of them, so that I can restrict myself to recalling them and occasionally making some additions. Ryan mentions, for example, a ludic function (which is, for instance, discernible in Packard’s Sugarcane Island as well as in the Disney comic strip). This may be reminiscent of what arguably can be regarded as one of the ‘natural’ origins of metalepsis, namely ‘frame-breaks’ in children’s games of impersonation (e.g. by means of ‘out-of-character-speaking’). With a view to the Disney comic strip but also to Stoppard’s The Real Inspector Hound, I would add to the ludic function a comic valorization (an addition called for by the remarkable frequency of metalepses in comedy and comic fiction29). It is also worth mentioning that in many cases a certain toying with the surprise element implied in metalepsis also plays a role. One may, thus, perhaps also speak of a sensational function of metalepsis, which can in fact be observed in almost all the examples given. In today’s competitive market the insertion of such surprises into comics, films etc. may also have the (additional) effect of enhancing the attractiveness of the ‘product’. If we follow this line of argument, Ryan is right to point to a further function of metalepsis: namely to covertly celebrate the creator and manager of metaleptic surprises: the author (this, remarkably, even holds true in the postmodernist context in which Woody Allen’s films are set and which is theoretically hostile towards the very idea of authorship30). Apart from the author, metalepsis can also have the effect of celebrating, or at least highlighting, the faculty which renders such authorial ‘games’ possible in the the first place, namely the imagination (or, in more modern wording, the faculty of the human mind to engage in the field of ‘the imaginary’ regardless of rational ‘impossibilities’). Another function mentioned by Ryan is equally transgenerically and intermedially valid, namely the fact that metalepsis, as a rule, is a marker of fictionality in the sense of ‘invention’. This is also the reason why genuine (i.e. ‘ontological’) metalepsis can hardly be expected in fact-oriented genres and media such as biography, historiography or documentary film. In addition to this marking of fictionality, which can in fact be observed in all of our examples, one may also mention an almost equally general 29

30

In fact, before modernism (where metalepsis occurs in serious contexts as in Pirandello’s Sei personaggi in cerca d’autore or Miguel de Unamuno’s Niebla [1914]) metalepsis almost invariably appears in a comic frame (which continues to apply, though, as e.g. Flann O’Brien’s hilarious novel At Swim-two-Birds [1939] shows). This function becomes particularly clear in cases such as John Fowles’s novel The French Lieutenant’s Woman (1969), where a figure that unmistakably refers to Fowles himself repeatedly enters the nineteenth-century world of the novel.


103

function of metalepsis, namely the strong anti-illusionist effect this device usually has. This effect is a consequence of the fact that metalepsis prevents the recipients from becoming ‘recentred’ or ‘immersed’ (cf. Ryan 1991 and 2001) in a coherent possible world and that it thus reminds them, through the frequently comic ‘impossibility’ of the metaleptic transgression, of the fictionality of the represented world.31 As such, metalepsis is also an implicitly meta-fictional or meta-aesthetic device, and this is why it is so frequently found in meta-works: in Carroll’s metafiction, in Pirandello’s and Stoppard’s metadramas, in Disney’s metacomic, in Woody Allen’s metafilms and in Escher’s and Magritte’s metapictures. Generally, as metalepsis typically operates across the boundary between ‘reality’ and ‘fiction’, it can serve as a trigger to reflections on this boundary or—as in the postmodernist use of the device—it can also undermine it under the auspices of a deconstructionist belief in ‘pan-fictionality’. All this could be regarded—if the employment of metalepsis stems from a serious and not merely ludic or sensational motivation—as a contribution to yet another function mentioned by Ryan, namely exploring (or acknowledging) the limits of human knowledge and concepts, including the concept of the subject, which, according to Malina 2002 (and in contrast to what has been said about metaleptic celebrations of the author) can advance to a favourite object of metaleptic deconstruction.32 Such epistemological reflections are, however, only a special case of another general effect of metalepsis, namely that it activates the recipient. This activation need not be of a genuinely interactive nature as in Sugarcane Island, yet some activity will always be elicited in the recipients by the paradoxical nature of metalepsis: for the ‘madness’ of this device will inevitably make them look for a ‘method in’t’ and will thereby refer them to some of the abovementioned functions or to still other functions, for the list given here does not purport to be complete.33 31

32

33

For the link between the undermining of aesthetic illusion and comedy (or a comic frame) see Wolf 1993: chap. 3.7. Interactive texts such as Sugarcane Island and computertransmitted virtual realities, however, seem to be exceptions: even though the systematic transgressions between a possible world and its ‘outside’ (the world of the recipient or player) show a typically metaleptic structure, in these cases the generic conventions of the ‘game’ neutralize the anti-illusionist effect; rather, the interactiveness, which apparently aims at deleting the ontological difference between the player/recipient and his or her diegetic representative (a character), appears, on the contrary, to enhance the impression of immersion. Malina—somewhat onesidedly—emphasises the ‘undermin[ing]’ of ‘illusions of bounded, essential selves and worlds’ (2002:138) and even contends that metalepsis can lead to a feminist blurring of ‘traditional [genderized] boundaries’ (24). In the discussion at the conference in Hamburg Harald Weilnböck pointed out that from a psychotherapeutic point of view the ‘madness’ of metalepsis can appear as a kind of narrative blocking caused by a traumatic experience.

104

Werner Wolf

The use of a common term not only highlights functional and structural similarities; it can also draw attention to historical developments and to the profile of individual epochs or cultural currents. Thus, the extent to which metalepsis has occurred in the art and media since modernism and in particular in the postmodern era is quite remarkable. This is, of course, no coincidence, since metalepsis is so akin to major postmodern concerns: a metamedial ‘self-consciousness’, the undermining of ontological certainties,34 and the use of the media for games and experiments as an alternative to, or substitute for, traditional truth claims, which the postmodernists no longer accept. To conclude: The highlighting of formal, functional and historical similarities together with the furtherance of interdisciplinarity appear to be sufficient reasons and heuristic gains to promote the use of narratological concepts in as many cultural fields as possible. Such ‘exports’ should, however, not be carried out thoughtlessly, and the caveats implied in the previous reflections were meant to serve as as a guide in this respect. Another caveat should be added: interdisciplinarity presupposes multidisciplinary competence, and this requirement imposes obvious limits on each scholar. As narratologists we can, and ought to, make offers to other disciplines by, e.g., formulating our concepts in an export-friendly way, but whether these offers are plausible and helpful is often not for the narratologist to decide. Yet, one should not be too shy or respectful of disciplinary boundaries either. After all, the common endeavour which has always motivated the humanities, namely to throw light on human cultural activities, is justification enough at least to try to cross disciplinary boundaries where appropriate. Exporting, is, of course, only one side of the interdisciplinary ‘trade’. The other side is importing notions and theories from other disciplines and media into narratology and literary studies as a whole. Such importing (e.g. from psychology, sociology and, more recently, the cognitive sciences) has been going on for decades. Yet it seems to me that it has been carried out in a predominantly unreflected way and has often merely followed the fashions of the day. It would therefore be high time to start reflecting in a more systematic way about these import fields and further possible import domains, about the profits such imports have or would have for narratology, and about conditions under which borrowings from other fields are fruitful for the elucidation of narratives. But this inversion of the perspective governing the present volume should rather be the subject of another conference, and another volume of proceedings. 34

Metalepsis, thus, illustrates the ontological ‘dominant’ which Brian McHale has regarded as typical of postmodernism, in contrast to the ‘epistemological dominant’ of modernism (1987:10 and 9).


105

References Abbate 1991 Abbate, Carolyn: Unsung Voices: Opera and Musical Narrative in the Nineteenth Century. Princeton, N.J. 1991. Allen 1990 Allen, Woody: “The Purple Rose of Cairo”. In: Allen, Woody: Three Films. London 1990:317–473. Bornemann 1995 Bornemann, Christiane: “The Purple Rose of Cairo.” In: Töteberg, Michael (ed.): Metzler Film Lexikon. Stuttgart 1995:474–475. Carroll 1970 Carroll, Lewis: The Annotated Alice. Ed. Martin Gardner. Harmondsworth 1970. Conrady 1978 Conrady, Karl Otto (ed.): Das große deutsche Gedichtbuch. Königstein / Ts. 1978. Dällenbach 1977 Dällenbach, Lucien: Le Récit spéculaire: Essai sur la mise en abyme. Collection Poétique. Paris 1977; Engl.: The Mirror in the Text. Transl. J. Whiteley / E. Hughes 1989. Disney 1996 Disney, Walt: “Der Vielfraß aus dem All.” In: Disney, Walt: Der erste Held von Olympia. Lustiges Taschenbuch 222. Leinfelden-Echterdingen 1996: 124–171. Eliot 1972 Eliot, T. S.: The Waste Land and Other Poems. London 1972. Ende 1979 Ende, Michael: Die unendliche Geschichte. München 1979. Escher 1956 Escher, M. C.: Picture Gallery (1956). In: Ernst, Bruno: Im Zauberspiegel des M.C. Escher. Transl. Ilse Wirth. Berlin 1986:32. Fludernik 1996 Fludernik, Monika: Towards a ‘Natural’ Narratology. London 1996. Fludernik 2003 Fludernik, Monika: “Scene Shift, Metalepsis, and the Metaleptic Mode.” In: Style 37, 2003:382–400. Genette 1972 Genette, Gérard: Figures III. Paris 1972. Genette 1983. Genette, Gérard: Nouveau discours du récit. Paris 1983. Genette 2004. Genette, Gérard: Métalepse: De la figure à la fiction. Paris 2004. Goodman 1968 Goodman, Nelson: Languages of Art: An Approach to a Theory of Symbols. Indianapolis 1968. Helbig 2003 Helbig, Jörg: “Wie postmodern ist Hyperfiction? Formen der Rezeptionslenkung in fiktionalen Hypertexten.” In: Alber, Jan / Fludernik, Monika

106

Werner Wolf

(eds.): Moderne / Postmoderne. Literatur – Imagination – Realität. Trier 2003:299–313. Herman 1997 Herman, David: “Toward a Formal Description of Narrative Metalepsis.” In: Journal of Literary Semantics 26, 1997:132–152. Hickethier 2001 Hickethier, Knut: Film- und Fernsehanalyse. Stuttgart 2001. Hofstadter 1980 Hofstadter, Douglas R.: Gödel, Escher, Bach: An Eternal Golden Braid. A Metaphorical Fugue on Minds and Machines in the Spirit of Lewis Carroll. Harmondsworth 1980. Hollander / Fletcher 1993 Hollander, John / Fletcher, Angus: “Metalepsis or Transumption.” In: Preminger, Alex / Brogan, T. V. F. (eds.): The New Princeton Encyclopedia of Poetry and Poetics. Princeton, N.J. 1993:759–760. Lodge 1977 Lodge, David: The Modes of Modern Writing: Metaphor, Metonymy, and the Typology of Modern Literature. London 1977. McClary 1997 McClary, Susan: “The Impromptu That Trod on a Loaf: or How Music Tells Stories.” In: Narrative 5.1, 1997:20–35. McHale 1987 McHale, Brian: Postmodernist Fiction. London 1987. Magritte 1928 Magritte, René: Le Maître du plaisir. In: Ausstellungskatalog der Kunsthalle Hypo-Stiftung. München 1987:Bild 28. Malina 2002 Malina, Debra: Breaking the Frame: Metalepsis and the Construction of the Subject. Columbus, Ohio 2002. Mosthaf 2000 Mosthaf, Franziska: Metaphorische Intermedialität: Formen und Funktionen der Verarbeitung von Malerei im Roman. Theorie und Praxis in der englischsprachigen Erzählkunst des 19. und 20. Jahrhunderts. Trier 2000. Nash 1987 Nash, Cristopher: World-Games: The Tradition of Anti-Realist Revolt. London 1987. Packard 1976 Packard, Edward: Sugarcane Island. Waitsfield / Vt. 1976. Pfister 1977 Pfister, Manfred: Das Drama. München 1977. Prieto 2002 Prieto, Eric: “Metaphor and Methodology in Word and Music Studies.” In: Lodato, Suzanne M. / Aspden, Suzanne / Bernhart, Walter (eds.): Word and Music Studies: Essays in Honor of Steven Paul Scher and on Cultural Identity and the Musical Stage. Amsterdam 2002:49–67. Ryan 1991 Ryan, Marie-Laure: Possible Worlds, Artificial Intelligence and Narrative Theory. Bloomington, Indiana 1991.


107

Ryan 2001 Ryan, Marie-Laure: Narrative as Virtual Reality: Immersion and Interactivity in Literature and Electronic Media. Baltimore 2001. Ryan [forthcoming] Ryan, Marie-Laure: “Logique culturelle de la métalepse, ou: La métalepse dans tous ses états.” In: Pier, John / Schaeffer, Jean-Marie (eds.): Métalepses. Entorses au pacte de représentation. Paris [forthcoming]. Sterne 1967 Sterne, Laurence: The Life and Opinions of Tristram Shandy, Gentleman. Ed. Graham Petrie. Harmondsworth 1967. Stoichita 1997 Stoichita, Victor I.: L’Instauration du tableau: Métapeinture à l’aube des temps modernes. Paris 1993; Engl.: The Self-Aware Image: An Insight into Early Modern Meta-Painting. Transl. Anne-Marie Glasheen. Cambridge 1997. Swift 1992 Swift, Graham: Waterland. London 1992. Wagner 2002 Wagner, Frank: “Glissements et déphasages: Note sur la métalepse narrative.” In : Poétique 130, 2002:235–253. Wiesner 2001 Wiesner, David: The Three Pigs. New York 2001. Wolf 1987 Wolf, Werner: “Lewis Carrolls ‘Alice’-Geschichten als sprach- und erkenntniskritische Metafiktionen: Ein Beitrag zur Geschichte des metafiktionalen Romans im 19. Jahrhundert.” In: Germanisch-Romanische Monatsschrift. n.s. 37, 1987:423–446. Wolf 1992 Wolf, Werner: “The Fiction of David Lodge in the Eighties. Discontent with Life and Literature Expressed from a Moderate, Liberal and Undogmatically Christian Point of View.” In: Anglistik und Englischunterricht 48, 1992: 19–42. Wolf 1993 Wolf, Werner: Ästhetische Illusion und Illusionsdurchbrechung in der Erzählkunst: Theorie und Geschichte mit Schwerpunkt auf englischem illusionsstörenden Erzählen. Tübingen 1993. Wolf 2002a Wolf, Werner: “Intermediality Revisited: Reflections on Word and Music Relations in the Context of a General Typology of Intermediality.” In: Lodato, Suzanne M. / Aspden, Suzanne / Bernhart, Walter (eds.): Word and Music Studies: Essays in Honor of Steven Paul Scher and on Cultural Identity and the Musical Stage. Amsterdam 2002:13–34. Wolf 2002b Wolf, Werner: “Das Problem der Narrativität in Literatur, bildender Kunst und Musik: ein Beitrag zu einer intermedialen Erzähltheorie.” In: Nünning, Ansgar / Nünning, Vera (eds.): Erzähltheorie transgenerisch, intermedial, interdisziplinär. Trier 2002:23–104. Wolf 2003 Wolf, Werner: “Narrative and Narrativity: A Narratological Reconceptualization and Its Applicability to the Visual Arts.” In: Word & Image 19, 2003:180–197.

Measuring Narrativity in Literary Texts Vyacheslav Yevseyev (Astana, Kazakhstan) 1. The Problem How can we measure narrativity? The answer to this question depends primarily on what we mean by ‘narrativity’ in the first place. It is not possible to cover all approaches to quantifying narrativity here, so I begin this article by considering a selection of possible ways of determining levels of narrativity, where ‘narrative’ is understood as meaning a text with a plot. I then proceed to discuss the minimal representation of a narrative and outline a method for calculating narrativity levels in terms of the density of minimal narrative elements in a text. Finally, I present selected results from a corpus analysis of German, Russian, and English texts. We should begin by recalling that a narrative text usually contains both narrative and non-narrative elements (the latter including descriptions, reflections, dialogues, etc.). Todorov remarks that narrative ‘does not exclude description’, although description alone ‘is not enough to constitute a narrative’ (1971:38). Onega and García Landa note that ‘one of the main differential traits of the novel is its ability to mix reflection and narrative’, whereas the drama ‘is much more directly focused on action’ (1996:3). The point behind these helpful observations is perhaps most cogently formulated by Prince: a passage where signs of the narrated (referring to events) are more numerous than signs of the narrating (referring to the presentation of events and its context) should have a higher degree of narrativity than a passage where the reverse is true (Prince 1982:146).

The above quotations imply that comparing the amount of narrative and non-narrative elements in a text could provide a means of measuring its narrativity. For example, we could go through all the clauses in a text in order to count how many of them contain description and how many narration. The inadequacy of this approach can be illustrated with two example sentences.

110

Vyacheslav Yevseyev

(1) A man came in and took a seat. (2) A good-looking young man with long dark hair, fashionably dressed, came in and took a seat in the corner of the festively decorated room, at the table by the window. Sentence (2) expresses exactly the same number of events as (1) but contains a series of covert descriptions in the form of numerous attributive and adverbial elements. On this basis, we can conclude that (1) is more narrative (both sentences represent two events, but the narrativity of (2) is reduced because of the presence of descriptive elements in it). It will be noted that (1) may seem overly terse and hardly the kind of sentence we would expect to find in a good novel. This shows that narrativity, if seen in this way, is to be treated as a purely technical property of literary texts. That is to say, I share Prince’s view that a high degree of narrativity may not be equated with a high level of aesthetic quality: ‘saying that one narrative has more narrativity than another does not necessarily mean that it is better or worse’ (Prince 1982:160). In what follows, my aim is to identify the typical levels of narrativity for a number of literary genres. Sturgess (1992:15) justifiably objects to Prince’s description of narrativity (see above) on the grounds that it does not take the length of narratives into account. In fact, if we develop Sturgess’s observation further, we see that calculating the ratio between the number of narrative elements in a text and the length of that text provides a possible solution to our problem. Since objective criteria (pages, lines, words, or characters) can be used to measure the length of written texts, it is possible to provide concrete, statistically verifiable statements when we gauge the level of narrativity in a text. For example: ‘telling a story, the narrator used on average x narrative elements per minute of telling / page of text’ or ‘listening to / reading a story, the listener / reader perceived on average x narrative elements per minute / page’. Different densities of narrative elements can be assumed to represent different levels of narrativity. The formula for calculating the density of narrative elements is as follows: density of narrative elements in text =

number of narrative elements length of text

Trivial as it may be, this method does not appear to have been applied previously to the analysis of levels of narrativity.

Measuring Narrativity in Literary Texts

111

2. Identifying Narrative Elements I turn now to the question of what we mean by the narrative elements whose density in texts we are going to measure. At first glance, the event stands out as the most appropriate unit to quantify.1 Many authorities, for example Genette (1988), have argued that a single change in a state of affairs should be taken as the minimal narrative. Danto (1965:252) proposes a more complex model. It is composed of three propositions (initial state, transition, and final state) and is labelled the ‘kernel narrative’ by Prince (1982:83). Meister (2003:146) develops the idea a step further by introducing the concept of the event matrix. An alternative approach treats the minimal narrative as consisting of the connection of two events. Labov, for example, defines a minimal narrative as ‘a sequence of two clauses which are temporally ordered’ (1972: 360). This view is also one of those expressed by Prince in his not always consistent pronouncements on the topic: ‘narrative is the representation of at least two real or fictive events or situations in a time sequence’ (1982: 4) and ‘any representation of non-contradictory events such that at least one occurs at a time t and another at a time t1 following time t constitutes a narrative’ (1982:145).2 A third response to the problem of defining the minimal narrative should also be considered. In the context of his work on the concept of shifters and the Russian verb, Roman Jakobson, who was both a linguist and a literary theorist, identifies a specific category that he calls taxis and defines as follows: ‘[t]axis characterizes the narrated event in relation to another narrated event and without reference to the speech event’ (Jakobson 1971:135). According to Jakobson, taxis is a non-shifter (i.e. a non-deictic category) since it lacks reference to the speech act and / or its participants (unlike tense, for example).

1

2

The word ‘event’ as I use it here refers to any change (as opposed to a static situation), irrespective of whether the change stands out aesthetically as an unusual occurrence in the theory of Lotman (1977). Contrast this with Prince’s statement that a story is ‘a sequence containing at least three narrative events’ (1982:150). Given that Prince occasionally fails to distinguish between changes and static situations in his use of the word ‘event’, we might recall Danto’s tripartite theory and conclude that Prince is thinking of a single change sandwiched between the preceding and following states. His examples, however, suggest that this assumption is wrong. In ‘John was very sickly, then he ate an apple, then, as a result, he became very healthy’ (Prince 1982:150), for example, one state (being sickly) and two changes (eating an apple and becoming healthy) are involved. Elsewhere, Prince shares Genette’s view that ‘narrative can be minimally defined as the representation of at least one event, one change in a state of affairs’ (Prince 1996:95).

112

Vyacheslav Yevseyev

Although he uses grammatical terminology to say it, Jakobson is really talking about the same thing as Labov and Prince. As far as I know, Jakobson’s idea has not previously been transferred to narratological research. According to the theory of the St. Petersburg school of Functional Grammar, where the most significant development of Jakobson’s idea over the past two decades is to be found, we should distinguish between simultaneous taxis (parallel states, inclusion of an event in a state, or overlapping events) and sequential taxis (contiguous sequences or discontiguous sequences). One essential requirement of taxis is that two events should be expressed in a coherent syntactic structure based on either co-ordination or subordination. In other words, the direct proximity of clauses in a text is a prerequisite of taxis (Bondarko 1991). For the purposes of this article, taxis of simultaneity will not be considered further, and instances of sequential taxis will be referred to as taxis units. My intention is to draw connections between narratology and taxis research, and to show that the taxis unit is the narrative element that should be counted when measuring narrativity. The presence of events is clearly a sine qua non of narration, but a single isolated event on its own cannot be treated as the minimal unit of narrative representation. This can be illustrated by analogy with an example from sociology. The individual is the fundamental element of human society but is not the minimal form of it. Since the essence of human society consists of links between individuals, at least two associated individuals are necessary to constitute the minimal form of society—a couple, a pair of friends, and so on. Returning to narratology, we can say that, since the plot is a complex of events on the temporal axis of a narrative, the minimal form of plot representation consists of exactly two associated events. Every narrative starts at a certain point in time and ends at another point in time later than the first. The most important characteristic of the taxis unit as a bipartite model of the minimal narrative is its ability to represent the temporal progression of narration. In the ideal case, there are no instances of anachrony and no descriptive elements, and each new clause refers to a point in time located after the reference point of the preceding clause, as described using Reichenbach’s terminology by Ehrich (1992:155; my translation): ‘each new statement si increments the reference time Ri in the direction in which time is advancing: Ri is temporally posterior to Ri-n (Ri > Ri-n).’3 A pair of temporally conjoined events represents a segment of the timeline and thus captures the movement of the reference time point in the course of narration.

3

See Nerbonne (1986) for a representative analysis of reference time.


113

It is thus reasonable to ask whether an isolated event can increment the reference time point. First, it must be ascertained for each individual case whether the initial state and the final state (i.e. the first and third constituents of the event matrix), or at least one of them, are explicitly indicated in the text. If this is not the case, they are merely pieces of implied information and thus cannot be considered part of the referential content of the text. Bearing this in mind, I restrict my analysis of the temporal period covered by an event to the temporal boundaries of the event itself.4 Second, it follows that we cannot treat a punctual event, an event without internal temporal extension (see Comrie 1976:41–43), such as to burst, to shoot, or to cough, as contributing to the temporal progression of the narrative. In general, every event whose internal temporal extension does not exceed 3 seconds (the upper limit of the ‘Psychological Now’—see for example Lehmann 1992:156, 160–162) should be projected onto the timeline as a point rather than a segment (examples are to enter the room, to notice something, and to greet a person). Therefore, in texts containing solely predicates of this type, i.e. in texts whose narrative speed is close to isochrony, a substantial proportion of the events should be depicted as points on the timeline. It follows that separate events by themselves do not represent the temporal progression of the narrative, they can do that only if they appear in pairs or larger groups.

4

Two examples from the extensive literature on aspect and Aktionsart will serve to show that it is erroneous to postulate initial and final states when only a change is mentioned explicitly. Flämig (1965:5–6) classifies the German verb platzen ‘to burst’ as an ‘Egressivum’ because it marks the ‘(sudden) conclusion’ of the action of being stretched; there are no real objections to this analysis. Chung and Timberlake (1985:218), meanwhile, classify the verb to die as an ‘inception of state’ (i.e. the beginning of the state of being dead), a much less convincing analysis. The juxtaposition of these two views shows that we could actually describe dying as the final phase of living and bursting as the initial stage of being torn. It would seem that identifying an implicit initial and/or final state for an event when only a change of state is explicitly depicted is a purely interpretive enterprise. Furthermore, whether or not we can categorize a particular event as the initial or the final phase of an (implied) state depends on whether or not there are explicit features that modify the meaning of a statal verb to give it phasal semantics. In Russian and German, special prefixes attached to the verb stem can fulfil this function. Classifying verbs as ingressive (inchoative) or egressive (conclusive) is thus only reasonable if we deal with derivatives; so German verblühen ‘to wilt’ (a change) is correctly identified as a ‘Konklusivum’ by Flämig (1965:5–6) since it is derived from blühen ‘to bloom’ (a state); an example for an inception of state is erblühen ‘to begin to bloom’ (a change). Returning to the above examples, we note that the German platzen ‘to burst’ is not morphologically derived from dehnen ‘to stretch’, and nor is the English to die directly morphologically derived from to be dead, let alone to live. To avoid arbitrary interpretations, it would seem sensible to confine our attention to the change in a narrative clause (without trying to identify associated initial and final states) and restrict our analysis to its temporal boundaries.

114

Vyacheslav Yevseyev

A further problem with the single-event model of the minimal narrative is posed by instances of coincident or parallel events. If two or more simultaneous events occur, they are inevitably and necessarily placed in a linear order when a text is written. Because of this, the reference time point moves cyclically rather than advancing forward in a linear manner when the resultant segment of the text is perceived. The reference time point returns repeatedly to the initial time point shared by all the simultaneous events. Furthermore, if a text displays a significant number of cases of anachrony, which occur when the position of events in the text is contrary to the principle of iconic order, it is scarcely possible to speak of (more or less obvious) forward movement of the reference time point between the beginning and end of narration. There is one final reason to stipulate that the minimal narrative element is composed of more than a single event: in some cases, it is not possible to determine the eventfulness or episodicity of a predication (see Schmid 2003 and Lehmann 1994 respectively) without reference to context. Andersson (1972:28ff.) notes that Grenzbezogenheit (‘telicity’ or ‘boundedness’) can be manifested in lexemes (he returned), in the connection of verbs to objects (he built a house), and on the level of complex sentences, for example ‘When Dufarge […] heard a queer exhalation of agony from the direction of the target, he was overjoyed’ (J. D. Salinger, The Laughing Man). In languages without a morphological category of aspect (e.g. German), the eventfulness of verbs for which differences in telicity are not specified becomes apparent only on the level of connection of predicates. According to Maslov (1984:28), there is a reciprocal relationship between the narrativity of a text and the event function of Russian perfective verb forms. In a sense, the relation between eventfulness and narrativity forces us into a hermeneutic circle: without determining the eventfulness of a situation we cannot speak of a narrative context, but the eventfulness of a situation, in turn, depends to a great extent on how it relates to the rest of the text in terms of, for example, logical relations and the principle of iconic order. In setting out the advantages of the taxis unit as the minimal narrative, it is in no way my intention to devalue work based on the singleevent model. I have simply attempted to show that the eventfulness of a predication is much more easy to recognize in the minimal context of two clauses than in isolation.


115

3. Taxis-based Narrative Strategy Let us assume that a text contains as many narrative clauses (N) as descriptive ones (D). There are then two basic ways in which the clauses can be distributed in the text. (1)

D–N–D–N–D–N–D–N–D–N–D–N–D–N–D–N–…

In (1), narrative clauses alternate with descriptive ones in the course of the story. The result is that no or hardly any taxis units can be identified. (2)

D–D–N–N–D–D–N–N–D–D–N–N–D–D–N–N–…

In (2), the narrative clauses come in pairs. According to the theory put forward above, texts with pattern (2) should be treated as more narrative than those with pattern (1), for a text with many grouped events provides a clearer representation of the temporal coherence of the plot. Yet, another possibility is conceivable: (3)

D–D–D–D–D–D–D–D–N–N–N–N–N–N–N–N–…

Pattern (3) represents narrative passages, which will be analysed below. Developing the idea that the microstructure of narrative consists of combinations of two events, I propose introducing the term ‘taxis-based narrative strategy’ to refer to the macrostructure of a narrative. This term describes how events are handled by the narrator, who can either – regularly express events in coherent syntactic structures, so that the temporal relations between the elements of the narrative chain are highlighted (many taxis units appear); or – scatter event-expressing clauses across the surface layer of the text, separating them from one another with non-narrative clauses, with the result that the temporal connections in the narrative are only latent ones—by placing events in isolation, the author causes the temporal connections in the narrative to remain hidden (hardly any taxis units can be recognized). It is legitimate to treat the distribution of taxis units in a text as a strategy rather than a statistical pattern because narrators, even if they are unaware of the concepts of taxis or the minimal narrative, nonetheless highlight or conceal the temporal connections in a narrative by intentionally placing narrative clauses in groups or isolation, thereby following or breaking the rules of traditional narration.5 5

Consider realist prose on the one hand and avant-garde texts on the other.

116

Vyacheslav Yevseyev

4. The Corpus and Its Analysis Applying the idea of narrativity levels not to texts but to event pairs (thus favouring again the bipartite model of the minimal narrative), Prince (1982: 145) links the level of narrativity in short textual segments containing such pairs to the plausibility of the connection between the events. Sentences like John was very rich, then he lost a lot of money should have a higher degree of narrativity than ones like The water boiled, then World War II started. In sentences of the latter type, the two events concern completely different areas of extra-linguistic reference. Such sentences were not considered in the corpus analysis carried out for the purposes of the project behind this article (they are, incidentally, not very common in real-life literary texts anyway). Care was also taken to bear in mind that the presence of two events in a coherent syntactic structure does not automatically indicate the presence of a taxis unit. Occasionally, two event clauses can communicate two aspects of what is really a single situation. For example, a taxis unit cannot be found in ‘[h]is companion who drove the car had broken silence for the first time since they left Belgrade and called to him’ (Graham Greene, Stamboul Train). Only syntactic structures expressing non-iterative events were considered; as repetition relates narration to description, iterative events were excluded from the analysis.6 If a syntactic structure contained three events forming a temporal succession, only one taxis unit was recorded, as in ‘[h]e wrote for ten minutes more and then folded the paper and slipped it into the pocket of his mackintosh’ (Graham Greene, Stamboul Train). In sentences with four events, two taxis units were recorded, as in ‘[1] [t]he Chief took his dollar Ingersoll out of his watch pocket, looked at it, [2] then swung around in his seat and started up the motor’ (J. D. Salinger, The Laughing Man). As these examples show, only clauses in the past tense were considered; clauses in the present tense (apart from the historic present) were excluded from consideration. It should be mentioned that certain intricacies of temporal meaning were ignored so that the material could be assigned to one of two dichotomic classes: simultaneous events were treated as irrelevant and sequential events as relevant. Cases in which it was difficult to decide whether the events were simultaneous or sequential (ten per cent in total) were not considered further. An example is: ‘[a] whistle screamed and a long line of lights came into view moving slowly’ (Graham Greene, Stamboul Train).

6

In this respect, I support Lotman’s definition of eventfulness; see also Schmid (2003: 29).


117

Having clarified the process by which taxis units are identified, I now turn to the problem of measuring the length of a text. It does not seem sensible to take pages, lines, or words as units of length. The format of pages and the length of lines can vary considerably, and words as such are not a suitable unit of measurement either, although reliable statistical calculations of the average word length in different languages have been carried out. A comparison of equivalent texts in languages with different morphological systems shows that the lower the average length of the words in a language, the higher the number of words usually needed to encode a given piece of information. It follows that equivalent texts can have comparable lengths, which can then be measured in the form of character counts. From a cross-linguistic point of view, counting the number of words in texts and adjusting the results to account for differences in word length is unnecessary when the problem can be approached more directly by measuring the length of the texts in characters. It might also be objected that equivalent texts in Russian, German, and English can have different lengths. I have carried out rough calculations using thirty German, Russian, and English original texts and their literary translations in the other two languages (e.g. an English original text would be compared with its Russian and German translation and vice versa). The outcome showed that English and Russian texts have approximately the same length, whereas German texts are an average of twenty per cent longer than their Russian or English equivalents.7 The issue needs further investigation; for the purposes of this article, however, no numerical adjustments were made as a result, for generalizations about the nature of narrativity levels are not made on the basis of a fraction of a percent but are derived from far more substantial differences. To make the calculations more manageable, passages of one thousand characters were adopted as a unit of reference. The length of every text was recorded in units of one thousand characters. As an example of this and the subsequent calculation, take “The Boarding House” in James Joyce’s Dubliners. The number of taxis units identified in this text was 22. The length of the text is 15.3 (i.e. 15,300 characters). So, the density of taxis units in the text is 22 / 15.3 = 1.44. The full results of the analysis

7

8

Modern German texts (e.g. those by Erich Maria Remarque and Heinrich Böll) are as long as their translations into English or Russian, whereas nineteenth- and early twentieth-century German novels (e.g. those by Theodor Fontane and Thomas Mann) are significantly (up to forty per cent) longer than their equivalents in other languages. The tables and diagrams in this article are based exclusively on figures given in Yevseyev (2003); the English examples, however, come from an additional study, the results of which will be published in a forthcoming article.

118

Vyacheslav Yevseyev

of seventy-six German and Russian texts from various historical periods (from the Lay of Hildebrand to Grass’s The Tin Drum and from Nestor’s Chronicle to Solshenitsyn’s Cancer Ward), including tables and diagrams, can be found in Yevseyev (2003:189–191, 258–261).8 5. Main Findings (1) Whole texts (not narrative passages) display levels of narrativity ranging from about 0.3 to about 3.0 taxis units per 1,000 characters. Put more simply, on each ordinary book page of a narrative text, the reader normally encounters between 1 and 10 cases of events encoded in such a way that they are connected. (2) The value of 0.3 taxis units per 1,000 characters (approximately one taxis unit per book page) can be provisionally identified as the threshold of narrativity. I see this threshold as providing statistical confirmation of the idea of the ‘minimum acceptable measure of narrativity’ (Kermode 1983:137) and the ‘initial degree of narrativity’ (Sturgess 1992:24). Of course, ‘just a little bit’ of narrativity is not enough to allow us to describe a text as narrative—‘saying that a text has some narrativity does not mean that it is a narrative’ (Prince 1982:161). Strikingly, some symbolist texts (e.g. Andrei Bely’s Kotik Letaev) do not contain any taxis units. Further investigation is necessary to increase our understanding of this area. (3) One of the most telling results of the study was that longer texts are, as a rule, less narrative than shorter ones in terms of the density of taxis units in them. Arranging texts into groups on the basis of their length and calculating the average narrativity value of each group produced the results shown in the following table.

Length of text (in characters)

Average narrativity value

1,000–10,000

2.70

10,000–100,000

1.67

100,000–1,000,000

1.38

1,000,000 and above

0.76

Table 1: Narrativity values in relation to text lengths

It is easy to see that narrativity is inversely proportional to text length. The following graph illustrates the situation in more detail:

119

Measuring Narrativity in Literary Texts Density of taxis units per 1,000 characters

Key thick line: average values of narrativity broken lines: approximate boundaries of the area

4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0

0

500

1,000

1,500

2,000

2,500

3,000

Length of the text in 1,000 characters

Graph 1: Narrativity levels in relation to text lengths

The graph shows that a low level of narrativity can be present in both long and short texts, whereas a high level of narrativity occurs only in short texts. There is no correlation between long texts and a high level of narrativity: short texts

high narrativity

long texts

low narrativity

Figure 1: Schema of narrativity / length correlations

A possible explanation for this lies in the fact that it is almost inconceivable that a text of medium length or above could refer solely to actions without containing any description, reflection, dialogue, and so on.9 It is of paramount importance that the mentioned schema applies to both German and Russian texts, and this fact might serve as an impetus to undertake research into the status of this regularity as a cross-linguistic universal.

9

Only such texts as the Old Russian Nestor’s Chronicle and the Middle Low German tales of Till Eulenspiegel revealed a high degree of narrativity in spite of their considerable length. The absence of a unified plot from these texts and their status as a series of shorter narratives may account for this fact.

120

Vyacheslav Yevseyev

6. The Homogeneity of Narrativity Levels in Texts Even if they do not accept the taxis unit as the minimal unit of narrative representation, most narratologists and non-specialist readers alike would agree that real-life literary texts do not display a uniform level of narrativity throughout their length. Prince (1982:146) distinguishes between narrative and descriptive passages. On the assumption that taxis units are not distributed evenly throughout a text, I tested my corpus for passages which are both as short as possible and contain as many taxis units as possible (five being the minimum). One such narrative passage occurs in James Joyce’s Dubliners: [1] Polly sat for a little time on the side of the bed, crying. [2] Then she dried her eyes and went over to the looking-glass. [3] She dipped the end of the towel in the water-jug and refreshed her eyes with the cool water. [4] She looked at herself in profile and readjusted a hairpin above her ear. [5] Then she went back to the bed again and sat at the foot. [6] She regarded the pillows for a long time and the sight of them awakened in her mind secret amiable memories. [7] She rested the nape on her neck against the cool iron bed-rail and fell into a reverie. [8] There was no longer any perturbation visible on her face (James Joyce, Dubliners).

This passage, in which 5 taxis units can be identified, contains 600 characters. The density of taxis units is therefore 5 / 0.6 = 8.33. This value is nearly 2.5 times greater than the upper limit of the range of typical levels of narrativity in complete texts (approx. 3.0).10 The extract deserves further examination. The first sentence contains a typical description: the actions of sitting and crying are simultaneous (the exact relation between them is a matter of interpretation). Sentences (2), (3), (4), (5) and (7) are typical taxis units. The reason why the events that form pairs do not coincide temporally lies in the status of the corresponding action sequences as scripts or frames (see, for example, Lehmann / Rauchenecker 1995:212 for an analysis of these terms); the events iconically represent the course of the story. It cannot be determined for sure whether the actions in sentence (6) are simultaneous, whether ‘regarding pillows’ temporally includes ‘awakening memories’, or whether the latter is a logical consequence of the former. 10

The highest narrativity value obtained for a complete text in the corpus analysis was 4.56. It was produced by Skazanie o Vavilone, a fifteenth-century Russian work; the exceptionally high narrativity value is made possible by the extreme brevity of the text (5,700 characters, or approximately 2.5 pages) and the abundance of events in it. Both factors lead to frequent occurrence of two (or more) events in a complex sentence and, as a result, to high density of taxis units. For this reason, the value of 4.56 was not considered typical of complete texts.


121

My corpus contained narrative passages with a density of 10 to 15 taxis units per 1,000 characters. If we abandon the condition that narrative passages should contain a minimum of 5 taxis units, and accept narrative passages consisting of only a single taxis unit, interesting results are obtained. For example, He came in and greeted us (25 characters = narrativity value of 40.0), No sooner said than done (24 characters = 41.7), or even Easy come, easy go (18 characters = 55.6). It goes without saying that a long narrative text of high aesthetic quality cannot consist solely of proverbs or simple sentences composed principally of subject and predicate. A real narrative needs a little bit more embellishment in the form of description (descriptive sentences, and attributes and adverbials in narrative sentences), dialogue, and so on. The literary technique of exhausting the narrative potential of a text is one that should be used with caution. 7. Typical Levels of Narrativity for Particular Literary Genres As mentioned above, I do not equate high narrativity in a text (in the sense of a high density of taxis units, the minimal units of plot representation) with a high level of aesthetic quality. My aim was to determine the typical level of narrativity for certain literary genres, not to evaluate the quality of individual texts. The following table shows the narrativity values recorded for the texts in each genre; the list of numbers on the right summarizes the range of narrativity values for each genre (rounded to the nearest integer or half-integer).11

Graph 2: Narrativity values across genres

11

See Yevseyev (2003:189–191) for more details.

122

Vyacheslav Yevseyev

Two groups of texts can be identified. (1) Text types in which the upper limit of the range of typical levels of narrativity is less than 2.0: the courtly romance, the heroic lay, the nineteenth- and twentieth-century novel, and the travel literature. (2) Text types in which the upper limit of the range of typical levels of narrativity exceeds 3.0: the fairytale and the folk-tale, the nineteenthand twentieth-century short story, the novella, the biography and the hagiographic text, and the chronicle. Genres which can produce texts with higher levels of narrativity (2.0 up to 3.0) should be regarded as (potentially) more narrative than those genres within which levels of narrativity up to maximally 2.0 have been observed. 8. Conclusion The method described in this article is no more than a first step in the direction of determining levels of narrativity on the basis of numerical criteria. The density of the occurrence of minimal units of plot relative to the length of a text can be used to measure narrativity, and one of the basic assumptions behind this article was that the minimal unit of plot is not an isolated event surrounded by descriptive elements but a pair of temporally, causally, and logically connected events. Consideration of aspect and Reichenbach’s concept of reference time provided additional insight into the minimal form of narrative representation. Even if it has not led to ground-breaking results, the method described here has at least provided statistical evidence that allows us to confirm and challenge some established principles and rules that characterize the narrative in general and certain narrative genres in particular. One of the most important insights is that longer texts are usually less narrative than shorter ones. In addition, we have obtained more precise information than was previously available about the range of narrativity levels associated with particular literary genres. Extending the corpus to include further genres or texts in additional languages would allow us to build on the results obtained using the method described here.


123

References Andersson 1972 Andersson, Sven-Gunnar: Aktionalität im Deutschen: Eine Untersuchung unter Vergleich mit dem russischen Aspektsystem (I). Die Kategorien Aspekt und Aktionsart im Russischen und im Deutschen. Uppsala 1972. Bondarko 1991 Bondarko, Alexander: Functional Grammar: A Field Approach. Amsterdam / Philadelphia 1991. Chung / Timberlake 1985 Chung, Sandra / Timberlake, Alan: “Tense, Aspect, and Mood.” In: Shopen, T. (ed.): Language Typology and Syntactic Description (III): Grammatical Categories and the Lexicon. Cambridge et al. 1985:202–258. Comrie 1976 Comrie, Bernard: Aspect: An Introduction to the Study of Verbal Aspect and Related Problems. Cambridge et al. 1976. Danto 1965 Danto, Arthur C.: Analytical Philosophy of History. Cambridge et al. 1965. Ehrich 1992 Ehrich, Veronika: Hier und Jetzt: Studien zur lokalen und temporalen Deixis im Deutschen. Tübingen 1992. Flämig 1965 Flämig, Walter: “Zur Funktion des Verbs (III): Aktionsart und Aktionalität.” In: Deutsch als Fremdsprache 2.2, 1965:4–12. Genette 1988 Genette, Gérard: Narrative Discourse Revisited. Ithaca, NY 1988. Greene 1973 Greene, Graham: Stamboul Train. Harmondsworth et al. 1973. Jakobson 1971 Jakobson, Roman: “Shifters, Verbal Categories, and the Russian Verb.” In: Jakobson, R.: Selected Writings (II): Word and Language. The Hague / Paris 1971:130–147. Joyce 1982 Joyce, James: “Dubliners.” In: J.J.: Dubliners. A Portrait of the Artist as a Young Man. Moscow 1982:39–219. Kermode 1983 Kermode, Frank: Essays on Fiction, 1971–82. London 1983. Labov 1972 Labov, William: Language in the Inner City: Studies in the Black English Vernacular. Philadelphia 1972. Lehmann 1992 Lehmann, Volkmar: “Grammatische Zeitkonzepte und ihre Erklärung.” In: Kognitionswissenschaft 2, 1992:156–170. Lehmann 1994 Lehmann, Volkmar: “Episodizität.” In: Mehlig, H. R. (ed.): Slavistische Linguistik 1993. Munich 1994:153–179.

124

Vyacheslav Yevseyev

Lehmann / Rauchenecker 1995 Lehmann, Volkmar / Rauchenecker, Eva: “Temporale Implikationen nichttemporaler Kategorien.” In: Jachnow, H. / Wingender, M. (eds.): Temporalität und Tempus: Studien zu allgemeinen und slavistischen Fragen. Wiesbaden 1995:199–223. Lotman 1977 Lotman, Jurij M.: The Structure of the Artistic Text. Ann Arbor, MI 1977. Maslov 1984 Maslov, Jurij S.: “Tipologija slavjanskix vido-vremennyx sistem i funkcio. nirovanie form preterita v ‘epicˇeskom’ povestvovanii.” In: Bondarko, A. V. (ed.): Teorija grammaticˇeskogo znacˇenija i aspektologicˇeskie issledovanija. Leningrad 1984:22–42. Meister 2003 Meister, Jan Christoph: Computing Action: A Narratological Approach. Berlin / NewYork 2003. Nerbonne 1986 Nerbonne, John: “Reference Time and Time in Narration.” In: Linguistics and Philosophy 9, 1986:83–95. Onega / García Landa 1996 Onega, Susanna / García Landa, José Angel (eds.): Narratology: An Introduction. London et al. 1996. Prince 1982 Prince, Gerald: Narratology: The Form and Functioning of Narrative. The Hague / Berlin / New York 1982. Prince 1996 Prince, Gerald: “Remarks on Narrativity.” In: Wahlin, C. (ed.): Perspectives on Narratology: Papers from the Stockholm Symposium on Narratology. Frankfurt / Main et al. 1996:95–106. Salinger 1982 Salinger, Jerome David: “The Laughing Man”. In: J.D.S.: Nine Stories. Franny and Zooey. Raise High the Roof Beam, Carpenters. Moscow 1982: 66–80. Schmid 2003 Schmid, Wolf: “Narrativity and Eventfulness.” In: Kindt, T. / Müller, H.-H. (eds.): What Is Narratology? Questions and Answers Regarding the Status of a Theory. Berlin / New York 2003:17–33. Sturgess 1992 Sturgess, Philipp J. M.: Narrativity: Theory and Practice. Oxford 1992. Todorov 1971 Todorov, Tzvetan: “The Two Principles of Narrative.” In: Diacritics: A Review of Contemporary Criticism. Fall, 1971:37–44. Yevseyev 2003 Yevseyev, Vyacheslav: Ikonizität und Taxis: Ein Beitrag zur Natürlichkeitstheorie am Beispiel des Deutschen und Russischen. Frankfurt / Main et al. 2003.

Quantitative Methods in Narratology: A Corpus-Based Study of Motion Events in Stories David Herman (Columbus, Ohio) 1. Two Approaches to Narrative Analysis The present chapter seeks to build on the supra-literary impulses of structuralist narratology—impulses that were evident in works like Roland Barthes’s “Introduction to the Structuralist Analysis of Narratives” (Barthes 1977) but that have begun to find fuller expression only in recent research, thanks to an ever more voluble and productive dialogue between practitioners of humanistic, social-scientific, and other kinds of work on narrative.1 Integrating ideas and methods from corpus linguistics, discourse analysis, cognitive science, and narratology, and focusing on how different kinds of stories represent events involving motion within or across spatial zones of the storyworlds that interpreters mentally model as they work to comprehend a narrative (Herman 2002), my chapter aims to foster greater synergy between two broad approaches to narrative analysis.2 One approach is grounded in what might be called the phenomenology of reading; this was the approach I myself used, for example, in a recent account of the processes of ‘spatialization’ and ‘perspective-taking’ that can be characterized as components or dimensions of narrative interpretation (Herman 2002:chaps. 7 and 8; see below). In this first approach, analysts consult their own readerly intuitions, which are made possible, in turn, by the nomenclatures and structural categories that derive from the collective intuitions of a larger community of expert readers (cf. Toolan 2001:22). Drawing on these resources, theorists propose hypotheses concerning how 1

2

To mention only some of the research that has emerged from or been shaped by this dialogue, see Bamberg (1997), Bruner (2003), Bortolussi and Dixon (2003), Fludernik (1996), Gerrig (1993), Herman (2001a, 2001b, 2003, 2004), Herman/Jahn/Ryan (2005), Mateas/Sengers (2003), Ochs/Capps (2001), and Van Peer/Chatman (2001). In parallel with Herman (2002), I define storyworld as a mental model of who did what to and with whom, when, where, why, and in what fashion in the world to which story-recipients relocate—or make a deictic shift (Zubin/Hewitt 1995)—as they work to comprehend a narrative.

126

David Herman

specific kinds of textual cues prompt interpreters to draw particular kinds of inferences about the nature, inhabitants, and spatiotemporal profile of storyworlds. In this approach goodness-of-fit between hypotheses and data stems from the overall ecological validity of the hypotheses in question—i.e. the extent to which they can be mapped onto the inferential activities used by readers engaged in the process of interpreting real-world narratives. Here as elsewhere, however, there is a basic tension between the ecological validity and the generalizability of research hypotheses: the logical end-point of an ecologically valid theory of reading (narratives) would be a theory that captures or emulates the experience of reading a particular text, in all its individual richness and complexity. Conversely, the second broad approach to narrative analysis trades ecological validity for increased generalizability; its logical end-point is a theory of some core set of cue-inference correlations underlying all narrative processing, no matter what kind of narrative is under examination (e.g. whether it is experimentally constructed or naturally occurring). This second approach derives from work in disciplines (including cognitive psychology, artificial intelligence, and corpus and computational linguistics) that have developed methods for gathering, sorting, counting, and analyzing empirical data. As Mick Short (2001) notes, the data in question may derive either from (1) informant-based research involving questionnaires or the measuring of readers’ responses, or from (2) the coding and analysis of identifiable textual structures. (As I discuss in more detail in section 3, the present chapter uses the second kind of data.) In general, the work in this second approach focuses on attested, measurable correlations between what Peter Dixon and Marisa Bortolussi (2003) divide into text features and text effects—i.e. semiotic cues and the inferential activities triggered, more or less consistently and reliably, by their occurrence in a given text. In this approach goodness-of-fit between hypotheses and data derives from several factors, including the reproducibility of the methods used to code and quantify the data at issue (will subsequent analysts be able to replicate my results?); the robustness of the distributional patterns that emerge when the data are coded and counted in that way (does text feature X always correlate with text effect Y, or only in some environments, and then only probabilistically versus categorically?); and the extent to which a given hypothesis adequately explains the distributional patterns revealed during the analysis (does hypothesis H account for why a particular correlation between cues and inferences tends to occur in environments of type E1 but not those of type E2?).3 3

Here I echo Chomsky’s (1965) three standards for goodness-of-fit between linguistic models and linguistic data, namely, observational, descriptive, and explanatory adequacy.

Quantitative Methods in Narratology

127

What I seek to demonstrate in this chapter, overall, is the need to dovetail these two broad approaches to narratological inquiry—approaches that are in fact interdependent to a degree that has not yet been sufficiently recognized. In social-scientific parlance, the two approaches can be classified as qualitative and quantitative, respectively (cf. Johnstone 2000). Qualitative (= ‘phenomenological’ or ‘intuition-driven’) methods address questions about how and why data have the particular character that they do. By contrast, quantitative (= ‘empirical’ or ‘data-driven’) methods address questions about how much (the degree to which) and how often (the frequency with which) those data display a given property or set of properties. As outlined more fully in the next section, I use here a combination of these two methods to revisit one of the central research hypotheses proposed in my previous, strictly qualitative work on the role of spatial reference in narrative domains (Herman 2001a; 2002:chap. 7). That work suggests that modes of spatial reference, like other design principles regulating storyworld (re)construction, take their character from the way they are anchored in preference-based systems for ‘spatializing’ narrated worlds, systems that in turn help constitute particular narrative genres as different kinds of stories. The present study tests the explanatory limits of this hypothesis by examining verbs of motion across eight narrative genres or text-types.4 Intended to be illustrative rather than exhaustive, the chapter is a pilot-study designed to promote further research in this area; it is, in effect, a heuristic device or ‘proof-of-concept’ tool. My specific aim is to throw light on the nature of the link between spatialization and genre by mapping out patterns in the use of motion verbs in a narrative corpus consisting of approximately 212,000 words. My broader aim is to suggest why this sort of endeavor can help disentangle the project of narrative analysis from the practice of interpreting particular (literary) narratives. 2. Spatialization, Motion Verbs, and Motion Events: Toward an Integrative Approach To be sure, it is a basic function of narrative to situate things (more or less definitely) in time. As suggested both by the Russian Formalist theorists (Shklovskii 1990) and by structuralist narratologists, stories are marked by a complex, double temporality, being sequentially organized accounts of sequences of events (Chatman 1990). Yet narratives can also be thought of as semiotic systems prompting their readers to spatialize storyworlds into evolving configurations of participants, objects, and places, with spatial reference playing a crucial, not an optional or derivative, role in stories 4

In this chapter, I treat the terms genre and text-type as interchangeable.

128

David Herman

(Herman 2002:chap. 7; cf. Herman 2001a). From this perspective, narratives should be viewed not just as temporally structured representations of event-sequences, but also as systems of verbal or visual prompts anchored in mental models having a particular spatial configuration. In fact, narrative affords a range of structures for encoding emergent spatial relationships among agents, places, and objects—for representing potential paths of motion, or possible routes through space. In this connection, recent work in narrative theory incorporating discourse-analytic and cognitive-scientific ideas has shown that narrative comprehension entails grasping the where of a story and not just the who, what, and when (cf. Zubin / Hewitt 1995); at issue is a process whereby interpreters use textual cues to build spatial information into the mental models that I have termed storyworlds. Emmott (1997), for example, has developed the notion of ‘contextual frames’ to discuss how readers of written narratives supplement text-based or propositional information with situation-based information (cf. Speelman / Kirsner 1990). When people read they do not automatically and iteratively assign referents to third-person pronouns, for instance, by attaching them to entities previously mentioned in the discourse. Rather, reference assignment is made possible when narrative texts cue readers to activate contextual frames, i.e. knowledge representations that store specific configurations of characters located at specific spacetime coordinates in the storyworld. Referring expressions thus evoke not just fictional individuals but whole contextual frames; pronouns like he and she and definite descriptions like the city and the person I met today are anchored not so much in particular entities as in the spatiotemporal contexts of those entities. Hence to know who or what is being referred to at a given point in a narrative text is to have the ability to build (as well as update) a mental model of where, within the storyworld, the person, object, situation, or event referred to is located in time and space. More generally, Emmott’s model, like Fludernik’s (1996) emphasis on viewing and experiencing as basic cognitive parameters for telling and interpreting stories, suggests that framing representations of the where in a story involves much more than simply filtering out descriptive detail to form interpretations of ‘core’ narrative elements (e.g. who did what to whom and why). I build on this research to explore a number of design principles in terms of which producers and interpreters of narrative spatialize storyworlds, i.e. construct mental representations of narrated environments as spatially configured, albeit in a dynamic, temporally emergent way (Herman 2002). The principles in question include: – deictic shifts (Zubin / Hewitt 1995), whereby a storyteller prompts his or her interlocutors to relocate from the here and now of the current


– – –

–

–

129

interaction to the alternative spacetime coordinates of the storyworld (i.e. the world being told about); use of the distinction between figure and ground, alternatively described as located object versus reference object; encoding of regions, landmarks, and paths, as characterized by Landau and Jackendoff (1993); use of the distinction between topological (or inherent) and projective (or viewer-relative) locations, this principle being intimately related to the perspectival structuring of storyworlds that constitutes another ‘macro’ design principle in narrative contexts, i.e., perspective-taking (Herman 2002:chap. 8); reliance on the deictic functions of motion verbs; these verbs are located on a semantic continuum whose poles, in English, are come and go (but see Wilkins / Hill 1995) and whose distribution in narrative discourse helps map out the positions of storyworld entities as they move or are moved along more or less narratively salient paths; differential reliance on the rich what and sparser where systems of spatial cognition (Landau / Jackendoff 1993), the former based on object shapes, names, and kinds, and the latter based on much coarser geometric properties of objects.

Here I zoom in on motion verbs in particular to sketch an integrative (i.e. quantitative as well as qualitative) approach to processes of spatialization across narrative text-types. Motion verbs are particularly instrumental for the construction and updating of cognitive maps (Ryan 2003) for storyworlds. In English, these verbs are located on a semantic continuum whose poles are come and go (Brown 1995:108–124; 188–191; Landau / Jackendoff 1993; Zubin / Hewitt 1995). By encoding the directionality of movement, motion verbs express viewer-relative locations of entities being perceived by narrators, as well as paths taken by entities as they transition from place to place. In the natural-language narratives Gillian Brown studied, verbs and verbal phrases such as ‘come, arrive, walk in are used of entry into the space ... which is nearest the observer ..., whereas go, walked off / out and leave are used as characters leave that space’ (Brown 1995:190). Likewise, consider the fight between the monster Grendel and Beowulf in the Old English epic Beowulf. In this scene, the Beowulf poet uses verbs and verbal phrases to encode the directions of the two combatants’ movements along paths that lead to and away from Heorot.5 Relevant forms are highlighted in the following passages: 5

For a fuller account of motion-verb usage in this scene, taking into account the original Old English text and the specifics of Old English verb morphology, see Herman/Childs (2003:192–93). All of the examples that follow occur on p. 36 of the Donaldson translation listed in the references section (Donaldson 1993).

130

David Herman

There came gliding in the black night the walker in darkness. Then from the moor under the mist-hills Grendel came walking, wearing God’s anger. The creature deprived of joy came walking to the hall. From his eyes came a light not fair, most like a flame. He [Grendel] stepped closer, then felt with his arm for the brave-hearted man on the bed, reached out towards him, the foe with his hand. His heart was eager to get away, he would flee to his hiding-place, seek his rabble of devils. The giant was pulling away, the earl stepped forward. The notorious one thought to move farther away, wherever he could, and flee his way from there to his fen-retreat.

As these passages suggest, motion verbs contribute crucial semantic information concerning the participants’ emerging whereabouts in space—their spatial trajectories over the duration of the event-sequence being narrated. Furthermore, the motion verbs enable the poet to correlate stages of the conflict with the directions in which Grendel and Beowulf move along a distal-proximal axis stretching between the two characters (with the proximal end corresponding to Beowulf’s vantage-point). The emphasized verbs thus do more than simply provide descriptive backdrop for the narrated action. By configuring the storyworld spatially as well as temporally, they enable interpreters to identify an action structure in which characters’ movements relate to their intentional designs as the narrative unfolds. Initially, the emphasized verbal forms represent Grendel as located at the distal end of an axis whose proximal end is the vantage-point of Beowulf, who is feigning to be asleep within Heorot when the episode begins. The constructions encode movement along a single linear path cutting bidirectionally through space. As long as Grendel is approaching Beowulf’s (initially stationary) vantage point, the monster occupies the position of the aggressor and potential victor of the struggle that is about to ensue. But once he feels the strength of Beowulf’s grip, realizing that he has finally met his match, Grendel reverses course, seeking to pull away, to flee. Meanwhile, Beowulf begins to pursue the monster, stepping forward and thereby approaching the distal end of the axis formerly occupied by Grendel, suggesting that the warrior is now the aggressor and in fact the victor. In this way, specific behavioral sequences can be mapped onto the global mental representations built by tellers and interpreters of narratives—representations in terms of which what happens can be construed as (in part) the result of characters’ goal-directed actions and counter-actions. In the remainder of this chapter, I attempt to extend the research hypotheses about spatialization that have emerged from case studies like the analysis of Beowulf just outlined (cf. also Herman 2001a; 2003). Toward that end, after discussing how in general terms corpus linguistics constitutes


131

an important resource for narrative analysis (section 3), I explore patterns of motion-verb usage in a corpus of stories (section 4). This section develops a quantitative approach to verbs of motion, complementing qualitative studies of those verbs as prompts for building mental representations of motion events in storyworlds (section 4). My chapter also draws on research stemming from Leonard Talmy’s analyses of the semantic encoding of motion events (1975, 1985; Naigles et al. 1998; Naigles / Terrazas 1998; Slobin / Hoiting 1994), revealing an interplay between the semantic structure of motion verbs and processes of spatialization in narrative contexts (section 5). Section 6 concludes the chapter by exploring the broader implications of my analysis and indicating directions for future research. Throughout my discussion, I take my cue from linguists who use textual corpora to ground (i.e. substantiate statistically) claims about the overall distribution of verbal structures in language use, and also about those structures’ semantic and more broadly communicative functions. More precisely, I use a mixed-methods approach to test the explanatory limits of the hypothesis that different narrative text-types afford different sets of (preference-based) protocols for encoding motion events—to mention just one set of protocols among others supporting the process of storyworld (re)construction. Thus, in addition to exploring the role of quantitative methods in narratology, my chapter uses those methods to address a specific, two-part research question: Can variations in the functioning of story logic in fact be mapped onto differences among narrative genres (as suggested in Herman 2002)? And furthermore, if a correlation of this sort does obtain, how might it be attested—and more exactly characterized—through analysis of a corpus consisting of multiple narrative text-types? 3. Corpus Linguistics and Narrative Analysis As Biber, Conrad, and Reppen (1998) point out, one of the chief goals of the corpus-based approaches to language study is to identify and analyze ‘association patterns’, i.e. ‘the systematic ways in which linguistic features are used in association with other linguistic and non-linguistic features’ (Biber / Conrad / Reppen 1998:5; cf. McEnery / Wilson 1996:87–116; Knowles 1996).6 A particular linguistic feature might be studied in relation to lexical patterns or grammatical constructions, on the one hand, or in relation to stages 6

See Semino/Short/Wynne 1999 for a discussion of some of the advantages of the corpusbased approach over traditional, i.e. qualitative, methods for the study of narrative texts in particular. One advantage is that a corpus-based approach compels analysts to face difficult or ‘messy’ cases, and therefore instances that might constitute counter-examples to their basic research hypotheses.

132

David Herman

in the evolution of a language, on the other hand. Corpus-based analyses can also be used to study ways in which groups of linguistic features commonly co-occur in particular text-types or registers (Biber / Conrad / Reppen 1998:7), i.e. modes of discourse usage linked to particular situational or contextual parameters. Hence there are collocation patterns specific to (and partly definitive of) the register of coaching, the register of academic debate, the register of public auctions, and so on. The present chapter uses corpus-based methods to study association patterns among verbs of motion and narrative text-types. I examine several kinds of association patterns, each of which was targeted by means of a research question answerable by quantitative methods: Which verbs of motion are the most common across all narrative text-types in the corpus? Conversely, which genres feature the highest overall concentration of motion verbs? Further, which text-types feature the greatest variety of motion verbs, i.e. the largest number of different motion verbs? At a more granular level of analysis, do the text-types exploit motion verbs having a variable semantic structure (Talmy 1975; 1985), such that verbs with certain kinds of structure can be reliably correlated with certain narrative genres? In addition, within a particular narrative text, can fluctuations in the frequency of motion-verb usage be reliably mapped onto shifts between diegetic levels? This last question broaches an issue discussed by Biber, Conrad, and Reppen (1998), who note that ‘a crucial part of the corpus-based approach is going beyond quantitative patterns to propose functional interpretations explaining why the interpretations exist’ (Biber / Conrad / Reppen 1998:9). Reciprocally, the pursuit of evidentiary backing for functional interpretations can motivate the search for quantitative patterns in the first place. In any case, the quantitatively oriented questions listed in the previous paragraph are intermeshed in the following analysis with an analogous set of qualitatively oriented questions, or tentative functional glosses of association patterns found in the data: Does the distribution of motion verbs in my corpus reveal genre-based differences in spatialization processes? For example, do text-types featuring the highest overall concentration of motion verbs activate, on the one hand, mental models of an environment in which there is relatively dense network of motion events, whereas genres featuring a wider range of different verbs of motion trigger, on the other hand, alternative spatialization strategies—ones registering, in finer detail, the effects of motion events on the consciousnesses of narrators, characters, and character-narrators? Do mapping relationships between frequency distributions and shifts in diegetic levels feed into the processing strategies used to parse particular narrative texts into framing and embedded narratives in the first place?


133

Sections 4 and 5 below attempt to demonstrate more fully the interconnections between these two sets of questions—questions about how many and how often versus questions about how and why. But first, the remainder of the present section outlines the methodology used to code, quantify, and analyze the corpus used for this study. (For a fuller description of the corpus itself, see the Appendix to this chapter available at ). 3.1 The Corpus The corpus, which includes fictional as well as nonfictional works categorized into eight narrative text-types, totals approximately 212,000 words.7 More specifically, the corpus consists of two subtypes of spoken narrative, one affiliated with legend (or ‘tall tales’) and the other factual / historical: North Carolina ghost stories and oral Holocaust testimonies. The former were elicited during sociolinguistic interviews; the latter were downloaded from an online oral history project called ‘Voices of the Holocaust’. Corresponding to these two sub-genres of oral narrative are two parallel subtypes taken from the fictional discourse of the 19th and 20th centuries: realistic fictional narratives and psychological fictions with supernatural or quasi-supernatural subject matter (abbreviated as ‘Realistic 19’, ‘Psychol. 19’, ‘Realistic 20’, and ‘Psychol. 20’ in the data displays below). The corpus also includes, for purposes of contrast with the fictional written narratives, one 19th and one 20th century written nonfictional narrative (abbreviated as ‘Nonfiction 19’ and ‘Nonfiction 20’). (See the Appendix for a complete list of the texts used as exemplars of each generic category and a tally of all the motion verbs tabulated for the study.) With the exception of the ghost stories, which were tape-recorded, transcribed, and imported into the corpus as pre-existing electronic files, all of the narratives used for the analysis were retrieved from the internet and converted into .txt or ‘text-only’ files searchable by the CONC 1.76 concordancing program (see section 3.2.1 and the Appendix). The corpus is at present untagged; that is, items (words, phrases, clauses, etc.) are 7

Although relatively small by comparison with corpora described by Biber, Conrad, and Reppen (1998:281–87), the corpus used for the present pilot-study is comparable in scope to the one used by Semino, Short, and Wynne 1999 (cf. Semino/Short/Culpeper 1997), which contains approximately 250,000 words of text taken from 20th century British narratives. Nonetheless, further development of this project will require a larger corpus size—to avoid, for example, excessive influence by the style of a particular writer on the overall frequency distributions in the corpus. See my discussion in section 4 of Middlemarch as the textual exemplar of 19th century realistic fiction, and also the concluding section of my chapter. Research on the North Carolina ghost stories that formed part of my corpus (as discussed in what follows) was funded in part by NSF grant BCS-0236838; the author gratefully acknowledges this support.

134

David Herman

not marked with symbols indicating the grammatical category of graphic tokens corresponding to data-points in the corpus. 3.2 Methodology 3.2.1 Software Tools The corpus was compiled and analyzed by means of readily accessible software tools. Microsoft Word was used to create the .txt files required as input for CONC 1.76; Excel was used to create charts and other graphic displays of the data. CONC 1.76, a concordance-generating program available as freeware for Macintosh computers (see Appendix), was used to search the corpus for all tokens of the motion verbs being counted. 3.2.2 Descriptive Terminology, Method of Analysis, and Selection Criteria The following terms and concepts, adopted from corpus-based linguistic research, were used in conducting the analysis; my subsequent discussion draws on this nomenclature. – Lemma = a set of words having the same stem and / or meaning and belonging to the same major word class (i.e., part of speech), differing only in inflection and / or spelling (cf. Francis / Kucera 1982; Biber / Conrad / Reppen 1998:29). – Base form = the member of a lemma consisting of the stem alone without inflection. – Raw frequency counts = total number of occurrences of items in a text, regardless of the relative size of the text vis-à-vis other texts being analyzed. – Normalized or ‘normed’ counts = frequency counts that convert the number of occurrences of an item to a standard scale (here, 100,000 words), providing an accurate basis for comparison of texts of different sizes (Biber / Conrad / Reppen 1998:32–35; 263–64). In my corpus, texts included in the corpus ranged from 7,365 words (19th century psychological fiction) to 52,685 words (19th century nonfiction), making normalization an important heuristic tool. I used CONC 1.76 to generate automatically a word concordance and index for each text in the corpus. Using the indexing and concordancing functions of the program, which accommodates more narrowly specified searches than the relatively gross keyword searches supported for example by Microsoft Word, I tabulated lemmas for 132 different verbs of motion (i.e. 132 distinct base forms; see the Appendix). Because the corpus is untagged, I relied on interactive analytic techniques (e.g. manual checking


135

of lists of verbal forms) to a greater degree than would have been necessary for a tagged corpus, which would have facilitated more fully automated frequency counts.8 With two exceptions, only intransitive verbs of motion were counted: – Exception 1: the lemmas whose base forms are bring and take and which encode direction of movement along a distal-proximal axis (Fillmore 1997). – Exception 2: transitive verbs that take a reflexive pronoun (seated herself, stretched himself ). Transitive verbs encoding directionality of movement (e.g. chase, follow, pursue) were not counted. Tokens of motion verbs in fixed formulas (e.g. enter into an agreement, take care to, crossed his mind) were not counted. Also excluded were tokens associated with motion events not involving a clear path of motion (e.g. quivered, shuddered, trembled). Further, with get, only tokens that co-occurred with directional particles (up, back), locative adverbs (there, here), or appropriate nominal forms (home) were counted. Since my study targeted the lemmas for verbs of motion (not the base forms only), perfective, imperfective, infinitive, future, gerundive, and participial forms of the verbs were all counted. 4. Some Association Patterns in the Data: Measurable Trends and Functional Interpretations Table 1 presents the raw totals for the 20 most frequent verbs of motion found in the corpus (travel having been discounted because of its skewed distribution); Table 2 presents normalized totals for those verbs. Meanwhile, for purposes of comparison and contrast, Table 3 displays adjusted frequency rankings (i.e. frequency rankings based on normalized instead of raw counts) for these verbs in (1) the Brown corpus on which Francis and Kucera (1982) based their frequency analysis of English usage, and (2) the present corpus of narratives.9 8

9

For example, because the corpus is untagged, a form like run had to be checked for its grammatical status in context. To mention just a few of the grammatical possibilities, run can function as a noun (They had the run of the place) as well as a present-tense verb in relatively more standard dialects of English (We run as a form of exercise), a regularized past-tense verb in relatively less standard dialects (We run from that spooky place when we was scared), or a past participle supported by an auxiliary verb (Having run away, we had nowhere to go). As noted by Francis and Kucera (1982) and Biber, Conrad, and Reppen (1998:281), the Brown corpus contains about 1 million words of written American English. In contrast with the corpus used for the present study, not all of the texts included in the Brown corpus were narratives. Further, it should be noted that the 5,996 lemmas included in Francis and Kucera’s (1982) frequency analysis were not limited to verbs.

136

Psychol. 19

Realistic 20

Psychol. 20

Nonfiction 19

Nonfiction 20

81 50 0 24 1 16 1 2 6 3 0 0 1 0 0 0 0 0 1 0

Realistic 19

Base Form go come take get leave walk bring run turn fall travel* enter move arrive escape return cross approach reach slip

Oral Holocaust testimony

Ghost stories

David Herman

93 96 55 67 39 34 25 7 7 7 2 1 4 10 4 5 5 4 2 4

78 68 22 17 5 15 32 5 18 10 5 7 4 3 1 7 0 4 4 1

1 8 2 0 0 0 2 1 2 7 0 4 3 2 1 0 1 2 3 0

63 53 7 4 3 15 3 12 22 14 0 1 2 0 5 0 3 1 3 0

17 6 3 3 10 5 9 1 7 4 1 2 6 4 0 0 2 0 3 2

163 111 77 77 54 21 28 63 10 10 40* 3 5 0 19 15 12 11 6 15

196 53 52 31 31 37 5 9 7 5 5 5 25 27 5 6 2 0 2 0

(*The verb travel was thrown out because 40 out of a total of 53 tokens of the verb occurred in a single text, drastically skewing the overall distribution of this item.)


Realistic 19

Psychol. 19

Realistic 20

Psychol. 20

Nonfiction 19

Nonfiction 20

Base Form go come take get walk leave turn bring fall run move arrive reach enter escape cross approach return slip step travel*

Ghost stories

Table 1: Raw Totals for the 20 Most Frequent Motion Verbs in the Corpus

701.9 433.3 0 24 138.6 208 52 8.7 26 17.3 8.7 0 8.7 0 0 0 0 0 0 0 n/a

331.1 341.9 195.8 238.6 121.1 138.9 24.9 89 24.9 24.9 14.2 35.6 7.1 3.6 14.2 17.8 14.2 17.8 14.2 0 n/a

176.2 153.6 49.7 38.4 33.9 11.3 40.7 72.3 22.6 11.3 9 6.8 9 15.8 2.3 0 9 15.8 2.3 0 n/a

13.6 108.6 27.2 0 0 0 27.2 27.2 95 13.6 40.7 27.2 40.7 54.3 13.6 13.6 27.2 0 0 0 n/a

402.7 338.7 44.7 25.6 95.9 19.2 140.6 19.2 89.5 76.7 12.8 0 19.2 6.4 32 19.2 6.4 0 0 0 n/a

118.1 305.7 20.8 20.8 34.7 10 48.6 62.5 27.8 6.9 41.7 27.8 20.8 13.9 0 13.9 0 0 13.9 13.9 n/a

309.4 210.7 146.2 146.2 39.9 102.5 19 53.1 19 119.6 9.5 0 6 5.7 36.1 22.8 20.9 28.5 28.5 11.4 n/a

520.2 140.7 138 82.3 98.2 8.3 18.6 13.3 13.3 23.9 66.4 71.7 5.3 13.3 13.3 5.3 0 15.9 0 26.5 n/a

Totals 2573.2 1769.2 622.4 575.9 562.3 498.2 371.6 345.3 318.1 294.2 203 169.1 116.8 113 111.5 92.6 77.7 75 58.9 51.8

Table 2: Normed Totals (per 100,000 words) for the 20 Most Frequent Motion Verbs in the Corpus

Quantitative Methods in Narratology Base Form go take come get leave turn bring move run reach fall step return walk enter arrive approach cross escape slip

Frequency Ranking in Brown Corpus 52nd 58th 60th 70th 130th 159th 172nd 196th 204th 278th 395th 412th 413th 435th 437th 858th 996th 1189th 1396th 2169th (out of 5996 lemmas)

137

Frequency Ranking in Present Corpus 1rst 3rd 2nd 4th 6th 7th 8th 11th 10th 13th 9th 20th 18th 5th 14th 12th 17th 16th 15th 19th (out of 20 lemmas)

Table 3: Adjusted Frequency Rankings for the 20 Verb Lemmas in the Present Corpus versus the Brown Corpus (based on Francis / Kucera 1982)

As Tables 1–3 indicate, come and go function in both corpora as ‘semantic primitives’ in the representation of motion events (see, however, Wilkins / Hill 1995 for ethnographically motivated counterarguments); these verbs are among the most frequently occurring items across all text types, narrative as well as non-narrative, no doubt because of their semantic productivity, i.e. their highly generative capacity to encode directions of movement from projective or viewer-relative locations (Brown 1995; Fillmore 1997). Indeed, the high frequency rankings for these motion verbs across multiple discourse contexts provides independent support for the notion that narrative and other uses of language bear the impress of ‘embodied cognition’, i.e. the constraining as well as enabling effects of having a particular, localized, physically embodied vantage point on the world (Lakoff / Johnson 1980; 1999). Thanks to this sort of experiential grounding, motion events prototypically involve not just movement in space but movement-awayfrom or movement-toward some situated vantage point or ‘deictic center’ (Zubin / Hewitt 1995). Perhaps for the same basic reason, frequency counts also square with the overall preference noted by Linde and Labov (1975) in their study of informants giving accounts of their apartment layouts. Those informants preferred to provide dynamic, projectively filtered tours rather than static, ‘bird’s-eye’ maps of their apartments. Further, though a comparison of the second and third columns of Table 3 reveals some variation in the rank order of motion verbs across the two

138

David Herman

181 1257.4

Nonfiction 20

256 1636.2

Nonfiction 19

83 1127

Psychol. 20

353 797.6

Realistic 20

Psychol. 19

496 1766.1

Realistic 19

Raw Count 196 Normed Totals 1698.4


Ghost stories

corpora (e.g. the verb step has a higher overall frequency ranking in the Brown corpus than in my narrative corpus, whereas the reverse situation obtains for walk), the overall trend in frequency rankings is preserved across the two corpora. The rankings could be represented by two clines having roughly the same slope, again suggesting consistency in the strategies used to represent motion events across multiple contexts and discourse types. (It should be kept in mind, however, that the adjusted frequency rankings provided by Francis and Kucera 1982 are based on a set of lemmas not restricted to verbs. A direct comparison of verb frequencies in the two corpora is thus not possible.) But other trends in the data offset these large-scale association patterns. In other words, although general trends in the use of motion verbs are evident across text types (and are probably anchored in fundamental cognitive dispositions, basic ways of comprehending and communicating facts about events involving motion through space), a finer-grained analysis of the data reveals genre-specific patterning in motion-verb usage. To put this point in yet another way, by revisiting the data-set at a more granular level of analysis, different association patterns start to emerge; these association patterns suggest that narrative text-types, though anchored in the same broad parameters for spatialization, in fact rely on variable protocols for representing motion events. Story logic—more specifically, the logic of spatialization—can thus be thought of as a system of principles and parameters within which a given narrative text will occupy a particular set of coordinates. The principles must be implemented, in some manner or another, for a text or a discourse to be located within the system at all—to be interpretable as a narrative. But the parameters for variable realization of the principles determine precisely what place within the system (e.g. what generic label or text-type category) a given narrative can be assigned (cf. Herman 2004). Table 4 shows differences in the relative frequency of verbs of motion across the eight text-types included in my corpus. Figure 1 provides a graphic display of these same data, which are again presented both in the form of raw counts and as normed totals (i.e. per 100,000 words).

902 1712

685 1818

Table 4: Frequency of Motion Verbs across Eight Narrative Text Types


139

Figure 1: Frequency of Motion Verbs across Eight Narrative Text-Types

Figure 1 highlights two noteworthy aspects of the data. In the first place, the Figure indicates that, with one exception, the lowest overall frequencies for motion-verb usage can be found in 19th and 20th century psychological fiction. This finding harmonizes with my qualitatively based hypothesis that narrative genres can be correlated with preference-based systems favoring particular event-types over others (Herman 2002:35–37); according to this model, psychological fiction in particular favors (interior) states and inchoative processes over accomplishments or achievements, to use Vendler’s (1967) terms for analyzing verb semantics. In the storyworlds of psychological fiction, then, motion events (and the motion verbs encoding them) would be relatively dispreferred (but see section 6 below). As Figure 1 also shows, however, the text type with the lowest overall frequency of motion verbs was nineteenth-century realist fiction. But again a functional interpretation can be brought to bear on the data, which in turn support the interpretive gloss arrived at through qualitative methods. Specifically, as the Appendix indicates, the textual exemplar used for the category of 19th century realistic fiction was George Eliot’s Middlemarch. In this case, the feminist-narratological model proposed by Lanser (1991) proves relevant. Lanser’s hypothesis that women’s fiction reflexively comments on standard parameters for action—more precisely, her suggestion that women’s texts commonly feature female protagonists whose plights derive not so much from their actions as from their inability to become

140

David Herman

Nonfiction 20

333.4

113.9

148.63

Psychol. 20

274.8

Realistic 20

556.7

Psychol. 19

94.9

Realistic 19

56


60

Ghost stories

Number of Different Verb Lemmas (Raw Count) Normed Totals

Nonfiction 19

genuine agents—throws light on the frequency distributions shown in Figure 1 (and vice versa). Meanwhile, Table 5 and Figure 2 reveal another association pattern: namely, the number of different lemmas for motion verbs in each of the eight narrative genres. If one factors in the ‘anomalous’ distribution for 19th century realistic fiction, Figures 1 and 2 stand in a more or less strictly inverse relation. More precisely, the text-type with the lowest overall frequency of motion-verb usage, 19th century psychological fiction, features the largest number of different motion verbs, with 20th century psychological fiction having the second largest array of verbs.

14

29

42

41

43

48

121.3 103.3

Table 5: Number of Different Motion Verbs across Eight Narrative Text Types

Figure 2: Number of Different Motion Verbs across Eight Narrative Text-types


141

In turn—and to anticipate ideas that will be discussed in my next section, i.e. ideas pertaining to Talmy’s (1975; 1985) model for the semantics of motion events—one might assume a positive correlation between the number of different motion verbs used and the detail with which the manner component of motion events is represented. Hence another functional interpretation can be introduced: although psychological fictions disprefer motion events as such, favoring states and inchoative processes instead, when they do portray such events 19th and 20th century psychological fictions are likely to draw on a richer repertoire of verbs than would other narrative genres. A larger array of verbs is needed to encode how the events are being processed by the experiencing narrators, characters, or character-narrators, through whose conscious activity the events are being filtered. And in fictions of this sort, representation of the nature and scope of such conscious activity takes pride of place over the representation of the events themselves (again, however, see section 6). My next section reviews some further association patterns throwing a different light on the functional gloss just outlined, while also providing grounds for additional ways of interpreting the data. 5. Verb Frequencies and the Semantic Structure of Motion Events In this section, I draw on research stemming from Leonard Talmy’s analyses of the semantic encoding of motion events (1975; 1985; Naigles et al. 1998; Naigles / Terrazas 1998; Slobin / Hoiting 1994); this research explores how typologically different languages use different kinds of ‘conflation patterns’ to assign values to the semantic parameters for events involving motion. With my previous sections having focused on corpus linguistics as a resource for narrative analysis, part of my aim in the present section is to suggest ways in which, conversely, the study of narrative corpora bears on domains of linguistic research, including work on the semantic representation of motion events. Theorists building on Talmy’s typological theory have followed Talmy in positing different ‘conflation patterns’ for what he characterizes as verbframed vs. satellite-framed languages, e.g. Spanish vs. English. In Talmy’s account, all motion events consist of six basic components: the motion itself (or fact of motion), the figure, the ground, the path, the manner, and the cause. These components can be conflated with (= encoded in) the root of verbs of motion or else with the accompanying ‘satellite’—e.g. spatial prepositions, locative adverbs, verb particles, and so on. The pos-

142

David Herman

Psychol. 19

Realistic 20

Psychol. 20

Nonfiction 19

Nonfiction 20

Motion-Encoding Verbs (% of Total) Manner + Motion-Encoding Verbs (% of Total)

86

83.3

50.6

72.3

74

79.3

81.8

15.8

14

16.7

49.4

27.7

26

20.7

18.2

Table 6:


84.2

Ghost stories

Realistic 19

sible conflation patterns that result help account for typological differences between languages, yielding contrasts such as the one distinguishing verbfrom satellite-framed language systems. Talmy argues that, as part of a satellite-framed language, English verbs are likely to conflate the motion and manner components of motion events, with the path component being encoded in directional particles, locative adverbs, and other satellites. This semantic process is exemplified in a locution like The girl RAN INTO the house. In Spanish, by contrast, the main verb is more likely to conflate motion and the path of the event, while the manner of motion, if expressed at all, will surface as a gerund, participle, or prepositional phrase—as in La niña ENTRÓ a la casa CORRIENDO (gloss: ‘The girl entered the house running’). Evidence from my corpus lends only partial support to Talmy’s hypothesis. The distribution of motion verbs in the corpus does provide statistical support for Talmy’s claim that English is a satellite-framed language. Thus, of the 5 most frequently occurring verbs (go, come, take, get, walk), none encodes the path component of motion events; instead that component is encoded in directional particles, locative adverbs, or other satellites accompanying the verbs. In contrast with Talmy’s claims about what is typical for English, however, only one of the five most frequently occurring motion verbs (walk) conflates the motion and manner components. More generally, only five (= 25%) of the twenty most frequently occurring motion verbs (walk, run, fall, slip, step) conflate motion and manner. Table 6 displays the breakdown, by text-type, of verbs encoding just the fact of motion versus verbs encoding both the motion and the manner components. Figure 3 provides a graphic display of these same data.

Motion-Encoding vs. Manner + Motion-Encoding Verbs


143

Figure 3: Motion-Encoding vs. Manner + Motion-Encoding Verbs

In Figure 3, the motion-verb frequencies in 19th century psychological fiction seem to substantiate one of the hypotheses advanced in my previous section: namely, that psychological fiction is likely to foreground, more than other narrative genres, the manner component of motion events (but compare section 6). The frequency count for manner + motion-encoding verbs in 20th century psychological fiction supports that same hypothesis less robustly; however, the difference between the percentages of manner + motion-encoding verbs in 20th century psychological fiction and 20th century realistic fiction, which would seem to skew the association pattern at issue, is negligible (27.7% versus 26%) and may stem from the relatively small size of the corpus. Figure 3 shows that, in contrast with the other text-types, nearly half of all the motion verbs in 19th century psychological fiction are manner- as well as motion-encoding. But an even finer-grained analysis is required to reveal another way in which narrative structure interacts with spatialization processes in this case. As Figure 4 indicates, there is an association pattern between narrative levels and verb frequencies. The textual exemplar used for this generic category is Poe’s “The Fall of the House of Usher”, in which an anonymous homodiegetic narrator recounts the experience of visiting the house of Roderick and Madeline Usher; Roderick in turn serves as an intradiegetic character-narrator, recounting his unsuccessful attempt

144

David Herman

to ‘bury’ his sister. More than 80% of the motion verbs used in this text are found in the diegetic level corresponding to the first narrator’s tellling, with the motion- versus manner + motion-encoding verbs distributed proportionately across the diegetic and hypodiegetic levels. Thus, different spatialization processes may be triggered by cues located at different narrative levels, while the readers’ discernment of shifts in diegetic levels may in turn derive from such (quantifiable) changes in frequency distributions.

Figure 4: Motion-Encoding (me) and Manner + Motion-Encoding (m + me) Verbs across Narrative Levels in Poe’s “Usher” (19CPF)

6. Conclusions and Issues for Further Research The present study constitutes, of course, only a beginning. A larger corpus needs to be assembled to offset effects that may be a function of the particular texts chosen to exemplify the eight generic categories. What is more, additional text-types may need to be considered, in order to give a fuller picture of how (or even whether) narrative genres afford preferencebased systems for spatializing storyworlds.10 10

The analysis needs to be extended in another way as well. Specifically, differences between kinds of motion events should be factored into the association patterns under study. Thus, in psycholinguistic experiments, Naigles et al. (1998; cf. Naigles/Terrazas 1998; Slobin/Hoiting 1994) found that choices between path of motion verbs and manner of motion verbs in Spanish are affected by whether the events being referred to involve crossing a boundary separating spatial zones, and also by whether those boundaries are horizontal or vertical in orientation. With horizontal boundary events, Spanish speakers show a strong preference for path of motion verbs over manner of motion verbs; but


145

Overall, though, I hope to have demonstrated that the study of narrative is now at a stage where researchers can make substantial gains by aligning ‘phenomenological’ (i.e. qualitative, intuition-driven) and ‘empirical’ (i.e. quantitative, data-driven) methods more closely than hitherto. At issue is a cyclical process of discovery, testing, and re-discovery—a process that both enables and is enabled by an interdisciplinary, ‘mixed-methods’ approach to narrative analysis. Research stemming from the phenomenology of reading affords taxonomic distinctions and research hypotheses whose validity and reliability across readings and readers can be checked empirically, this being a testing procedure that may in turn lead to new taxonomies encompassing fewer or more numerous categories and distinctions, depending on the robustness of patterns revealed by empirical analysis. Conversely, because of their knack for describing how specific textual features cue particular processing strategies, narratologists using qualitative methods can provide a sounder basis for empirical, corpus-based studies of narrative structure and narrative interpretation. For example, the findings of the present study suggest that two of my original hypotheses may need to be qualified, or at least subjected to more thorough testing by means of a larger corpus. Specifically, the distributional patterns for motion verbs in 19th and 20th century psychological fiction supported with different degrees of robustness the hypothesis that (1) in contrast with a genre such as realistic fiction, psychological fiction in general disprefers motion events as such, favoring states and inchoative processes instead; and (2) that psychological fictions are likely to draw on a richer repertoire of verbs than realistic narratives, preferring to foreground the manner component of motion events over the fact of motion itself. Taken together, Figures 1, 2, and 3 suggest the need to revisit the assumption that the realistic and psychological narrative text-types can be correlated with different preference-based systems for encoding motion events. Given the similar overall distribution of motion-verb usage in 19th century psychological and realistic fictions (Figure 1); given the nearly equivalent number of different motion verbs used in 20th century psychological and realistic fictions (Figure 2); and given the almost identical patterning of sub-types of motion verbs in 20th century psychological and realistic fictions (Figure 3), additional exemplars of each text-type category need to be studied to test the validity of the basic distinction at issue. they show a weaker preference for path of motion verbs with a change in plane—i.e. when the motion event involves the crossing of a vertical boundary. In conducting their study, however, Naigles et al. (1998; Naigles/Terrazas 1998) used stimuli consisting of isolated (non-narratively presented) motion events. Thus, a question for further research is whether spatialization processes in narrative are also sensitive to two cross-cutting distinctions: boundary-crossing versus non-boundary-crossing events, on the one hand, and vertical versus horizontal boundaries between spatial zones, on the other hand.

146

David Herman

Furthermore, it is important to note the differential behavior of motion verbs in 19th versus 20th century narratives in this connection. Whereas the 19th century realistic and psychological text-types were most closely aligned in terms of overall distribution of motion verbs, in the case of the 20th century exemplars the realistic and psychological varieties converged with respect to the number of different kinds of verbs and also with respect to the verbs’ underlying semantic structure. Here the cyclical testing process mentioned above has begun to manifest itself: measuring my original, qualitatively based hypotheses against the distributional facts indicates that the scope of those hypotheses may need to be adjusted; at the same time, that discovery has in turn generated a new hypothesis about period-specific differences in spatialization strategies, suggesting the need to explore association patterns between motion-verb usage in the realistic and psychological text-types, on the one hand, and instances of those texttypes in different historical epochs, on the other hand. To sum up my argument, I paraphrase Kant: Past a certain threshold, the formulation and refinement of narratological taxonomies in the absence of empirical, quantitatively based testing may prove to be empty. But the empirical study of stories without qualitatively anchored— phenomenologically grounded—investigative hypotheses is blind. Hence, when it comes to analysis of spatialization processes in narrative contexts, integrating quantitative and qualitative approaches constitutes the most promising research strategy, as well as the best hope for illuminating the literally chrono-topic profile—the temporally emergent geometries—of storyworlds.11 References Bamberg 1997 Bamberg, Michael (ed.): “Special Issue: Oral Versions of Personal Experience: Three Decades of Narrative Analysis.” In: Journal of Narrative and Life History 7, 1997:1–415. Barthes 1977 Barthes, Roland: “Introduction to the Structural Analysis of Narratives.” In: Image-Music-Text. Translated by Stephen Heath. New York 1977: 79–124. Barthes 1986 Barthes, Roland: “The Reality Effect.” In: The Rustle of Language. Translated by Richard Howard. New York 1986:141–148. Biber / Conrad / Reppen 1998 Biber, Douglas / Conrad, Susan / Reppen, Randi: Corpus Linguistics: Investigating Language Structure and Use. Cambridge 1998. 11

I am grateful to the editors of this volume, particularly Chris Meister, for their helpful comments on an earlier draft of my chapter.


147

Bortolussi / Dixon 2003 Bortolussi, Marisa / Dixon, Peter: Psychonarratology: Foundations for the Empirical Study of Literary Response. Cambridge 2003. Brown 1995 Brown, Gillian: Speakers, Listeners and Communication: Explorations in Discourse Analysis. Cambridge 1995. Bruner 2003 Bruner, Jerome: Making Stories: Law, Literature, Life. Cambridge, MA 2003. Chatman 1990 Chatman, Seymour: Coming to Terms: The Rhetoric of Narrative in Fiction and Film. Ithaca 1990. Chomsky 1965 Chomsky, Noam: Aspects of the Theory of Syntax. Cambridge, MA 1965. Donaldson 1993 Donaldson, E. T.: Translation of Beowulf. In: Abrams, M. H. et al. (eds.): The Norton Anthology of English Literature. 6th edition, vol. 1, New York 1993:27–68. Emmott 1997 Emmott, Catherine: Narrative Comprehension: A Discourse Perspective. Oxford 1997. Fillmore 1997 Fillmore, Charles J.: Lectures on Deixis. Stanford 1997. Fludernik 1996 Fludernik, Monika: Towards a ‘Natural’ Narratology. London 1996. Francis / Kucera 1982 Francis, W. Nelson / Kucera, Henry: Frequency Analysis of English Usage: Lexicon and Grammar. Boston 1982. Gerrig 1993 Gerrig, Richard J.: Experiencing Narrative Worlds: On the Psychological Activities of Reading. New Haven 1993. Goddard 1997 Goddard, Cliff: “The Semantics of Coming and Going.” In: Pragmatics 7, 1997:147–162. Herman 2001a Herman, David: “Spatial Reference in Narrative Domains.” In TEXT 21.4, 2001:515–41. Herman 2001b Herman, David: “Narrative Theory and the Cognitive Sciences.” In: Narrative Inquiry 11.1, 2001:1–34. Herman 2002 Herman, David: Story Logic: Problems and Possibilities of Narrative. Lincoln 2002. Herman 2003 Herman, David: “Stories as a Tool for Thinking.” In: Herman, David (ed.): Narrative Theory and the Cognitive Sciences Stanford, CA 2003: 163–192. Herman 2004 Herman, David: “Toward a Transmedial Narratology.” In: Ryan, MarieLaure (ed.): Narrative across Media: The Languages of Storytelling. Lincoln 2004:47–75.

148

David Herman

Herman / Childs 2003 Herman, David / Childs, Becky: “Narrative and Cognition in Beowulf.” In: Style 37.2, 2003:177–202. Herman / Jahn / Ryan 2005 Herman, David / Jahn, Manfred / Ryan, Marie-Laure (eds.): The Routledge Encyclopedia of Narrative Theory. London 2005. Hockey 2001 Hockey, Susan: “Concordance Programs for Corpus Linguistics.” In: Simpson, Rita C. / Swales, John M. (eds.): Corpus Linguistics in North America. Ann Arbor 2001:76–97. Johnstone 2000 Johnstone, Barbara: Qualitative Methods in Sociolinguistics. Oxford 2000. Knowles 1996 Knowles, Gerry: “Corpora, Databases and the Organization of Linguistic Data.” In: Thomas, Jenny / Short, Mick (eds.): Using Corpora for Language Research. London 1996:36–53. Lakoff / Johnson 1980 Lakoff, George / Johnson, Mark: Metaphors We Live By. Chicago 1980. Lakoff / Johnson 1999 Lakoff, George / Johnson, Mark: Philosophy in the Flesh: The Embodied Mind and Its Challenge to Western Thought. New York 1999. Landau / Jackendoff 1993 Landau, Barbara / Jackendoff, Ray: “‘What’ and ‘Where’ in Spatial Language and Cognition.” In: Behavioral and Brain Sciences 16, 1993:217–65. Lanser 1991 Lanser, Susan Sniader: “Toward a Feminist Narratology.” In: Warhol, Robyn / Herndl, Diane Price (eds.): Feminisms: An Anthology. New Brunswick 1991. Linde / Labov 1975 Linde, Charlotte / Labov, William: “Spatial Networks as a Site for the Study of Language and Thought.” In: Language 51, 1975:925–939. Mateas / Sengers 2003 Mateas, Michael / Sengers, Phoebe (eds.): Narrative Intelligence. Amsterdam 2003. Naigles et al. 1998 Naigles, Letitia R. / Eisenberg, Ann R. / Kako, Edward T. / Highter, Melissa / McGraw, Nancy: “Speaking of Motion: Verb Use in English and Spanish.” In: Language and Cognitive Processes 13, 1998:521–559. Naigles / Terrazas 1998 Naigles, Letitia R. / Terrazas, Paula: “Motion-verb Generalizations in English and Spanish: Influences of Language and Syntax.” In: Psychological Science 9, 1998:363–369. Ochs / Capps 2001 Ochs, Elinor / Capps, Lisa: Living Narrative: Creating Lives in Everyday Storytelling. Cambridge, MA 2001. Ryan 2003 Ryan, Marie-Laure: “Cognitive Maps and the Construction of Narrative Space.” In: Herman, David (ed.): Narrative Theory and the Cognitive Sciences. Stanford, CA 2003:214–242.


149

Semino / Short / Culpeper 1997 Semino, Elena / Short, Mick / Culpeper, Jonathan: “Using a Corpus to Test a Model of Speech and Thought Presentation.” In: Poetics 25, 1997: 17–43. Semino / Short / Wynne 1999 Semino, Elena / Short, Mick / Wynne, Martin: “Hypothetical Thoughts and Words in Contemporary British Narratives.” In: Narrative 7, 1999: 307–334. Shklovskii 1990 Shklovskii, Viktor: Theory of Prose. Translated by Benjamin Sher. Elmwood Park, IL 1990. Short 2001 Short, Mick: “Epilogue: Research Questions, Research Paradigms, and Research Methodologies in the Study of Narrative.” In: Van Peer, Willi / Chatman, Seymour (eds.): New Perspectives on Narrative Perspective. Albany 2001:339–355. Slobin / Hoiting 1994 Slobin, Dan / Hoiting, Nini: “Reference to Movement in Spoken and Signed Language: Typological Considerations.” In: Proceedings of the Twentieth Annual Meeting of the Berkeley Linguistic Society. Berkeley 1994:487–505. Speelman / Kirsner 1990 Speelman, Craig / Kirsner, Kim: “The Representation of Text-Based and Situation-Based Information in Discourse Comprehension.” In: Journal of Memory and Language 29, 1990:119–132. Talmy 1975 Talmy, Leonard: “Semantics and Syntax of Motion.” In: Kimball, John P. (ed.): Syntax and Semantics. Vol. 4, New York 1975:181–238. Talmy 1985 Talmy, Leonard: “Lexicalization Patterns: Semantic Structure in Lexical Forms.” In: Shopen, Timothy (ed.): Language Typology and Syntactic Description. Vol. 3, Cambridge 1985:57–149. Toolan 2001 Toolan, Michael J.: Narrative: A Critical Linguistic Introduction. London 2001. Van Peer / Chatman 2001 Van Peer, Willi / Chatman, Seymour (eds.): New Perspectives on Narrative Perspective. Albany 2001. Vendler 1967 Vendler, Zeno: Linguistics in Philosophy. Ithaca 1967. Wilkins / Hill 1995 Wilkins, David P. / Hill, Deborah: “When ‘Go’ Means ‘Come’: Questioning the Basicness of Basic Motion Verbs.” In: Cognitive Linguistics 6, 1995: 209–259. Zubin / Hewitt 1995 Zubin, David A. / Hewitt, Lynne E.: “The Deictic Center: A Theory of Deixis in Narrative.” In: Duchan, Judith F. / Bruder, Gail A. / Hewitt, Lynne E, (eds.): Deixis in Narrative: A Cognitive Science Perspective. Hillsdale, NJ 1995:129–155.

The Lydgate Storyworld Alan Palmer (London) 1. Introduction The Lydgate storyworld is Lydgate’s mind in action. When we attempt to follow the mental functioning of Tertius Lydgate, one of the characters in George Eliot’s Middlemarch, we experience the whole Middlemarch storyworld from his perceptual, cognitive and ethical viewpoint. It is in this way that we see the whole of his social mind in action. In analysing the presentation of the consciousness of a character such as Lydgate, we need to consider not just the passages of text that present his inner speech in the speech modes of direct thought, free indirect thought and thought report; not just whether he is a flat or round character or which of the various intertextual stereotypes readers will apply to him; not just his position as an actant or function within the structure of the story; and not just his role as a focalizer or focalized object. All of those things are important and will contribute towards an understanding of how Lydgate’s mind works, but more is needed. The possible worlds theorist, Lubomír Doležel, asserts that, from ‘the viewpoint of the reader, the fictional text can be characterized as a set of instructions according to which the fictional world is to be recovered and reassembled’ (1988:489). The sets of instructions just listed are characteristic of traditional, classical narratological approaches towards presentations of consciousness in fiction. However, there are many others that are only now becoming visible. This essay is an attempt to describe and illustrate some of the sets of instructions that enable us to analyse in new and illuminating ways hitherto neglected aspects of the functioning of fictional minds. After spelling out the assumptions on which my arguments are based, I will talk in general terms about the notion of cognitive frames before explaining a concept that is central to this essay: the continuing consciousness frame. I will then discuss another pivotal concept, that of embedded narratives, before considering three subframes, as I term them, of the continuing consciousness frame: the relationship between fictional thought and fictional action; intermental thought, or shared, group or joint thinking; and doubly

152

Alan Palmer

embedded narratives: the versions of characters that exist within the minds of other characters. I will illustrate the three subframes by discussing, in particular, Lydgate’s relationship with what I call the Middlemarch mind (that is, the intermental or consensus view of the inhabitants of the town) as well as his relationship with his wife, Rosamond. Middlemarch is a novel about the inhabitants of an English industrial town in the early nineteenth century. The two main characters are Tertius Lydgate and Dorothea Brooke, two idealists who have to adjust, painfully and slowly, to the realities of the lives that they actually find themselves leading. However, the novel is characterized by a dense and complex interweaving of the stories of a large number of other people too. I have chosen to discuss this novel for three reasons. First, it is a canonical novel and, although it might be thought that there is little left to say about it, I hope to show that a fictional minds analysis can reveal fresh insights even into canonical ‘warhorses’. Also, the interweaving of a number of stories makes the novel especially amenable to the embedded narrative approach. Finally, a cognitive perspective on Middlemarch is particularly satisfying because, as I mention below, the language of the text is, to a surprising extent, explicitly cognitive in nature.

2. Five Assumptions This essay rests on five basic assumptions. I do not have space to justify them in detail here: this is attempted in Palmer (2004). However, I will spell them out now to make them explicit. 1. Such methodologies as the discourse analysis of thought presentation (stream of consciousness, interior monologue, free indirect discourse and so on), story analysis (the study of characters as actants, functions and so on), the concept of focalization or point of view and the study of characterization do not add up to a complete and coherent study of all aspects of the minds of characters in novels. 2. Narratology has neglected the whole minds of fictional characters in action by giving undue emphasis to private, passive, solitary and highly verbalized thought at the expense of all the other types of mental functioning. There is some evidence in Middlemarch of Lydgate’s inner speech (for example, ‘he thought to himself’ [110] and ‘he said to himself’ [111]), but there is much more evidence of the other sorts of mental functioning that I describe later on. 3. We should make use of what I call the parallel discourses on real minds, such as cognitive science (for example, Damasio [2000]), psycholinguistics (Pinker [1997], Wertsch [1991]), psychology (Elster [1999]), and

The Lydgate Storyworld

153

the philosophy of mind (Dennett [1991], Searle [1992]), to study the whole of the social mind in action in the novel. They are parallel discourses because they contain a very different kind of picture of consciousness from that provided so far by narratology, and so can be used to provide explanations that are fuller than those that are currently available as to how the reader is able to re-assemble Lydgate’s mind from the Middlemarch text. 4. The constructions of the minds of fictional characters by narrators and readers are central to our understanding of how novels work, because narrative is, in essence, the presentation of fictional mental functioning. It is not possible to follow the plot of Middlemarch without following the thought processes of Lydgate, Dorothea, Rosamond and the other characters in the novel. 5. When the existing narratological approaches referred to in 1 above are brought together within a new theoretical perspective, the study of fictional minds will then be established as a clearly defined and discrete subject area in its own right within the discipline of narratology. This essay is intended to give an indication of what this new subject area might look like, and how it can be applied to a specific fictional text.

3. Cognitive Frames Theorists such as David Herman (1997, 2002 and 2003), Monika Fludernik (1996), Mark Turner (1991) and Gilles Fauconnier (1997) have contributed a good deal to our understanding of the reading process by using the techniques of cognitive science. One concept that is of particular importance is that of the cognitive frame. The cognitive scientists Roger Schank and Robert Abelson have explained that we use ‘specific knowledge to interpret and participate in events we have been through many times’ (1977:37). These knowledge structures are called frames. They capture the essence of concepts or stereotypical situations, for example being in a living room or going to a restaurant, by clustering all relevant information for these situations together. This includes information about how to use the frame, information about expectations (which may turn out to be wrong), information about what to do if expectations are not confirmed and so on (Nebel 1999:324). Cognitive frames have default values that are used if no value is provided from the external world. These default assumptions tell us about what is likely to happen in particular situations in the absence of specific information. It is only when our assumptions are proved wrong that we have to improvise. We assume that a restaurant will serve food at a price unless we are informed otherwise. The default value of the food slot within the restaurant frame is: availability of a range of

154

Alan Palmer

food in exchange for payment. Marie-Laure Ryan’s principle of minimal departure (1991) is a description in different terms of the default values contained in the frames that we apply to fictional texts. Our assumption that the storyworld will not depart from the real world unless we are told otherwise is a default position. It applies in the case of Middlemarch, where there are no magical beings with supernatural powers, and where characters’ minds work in the ways that we expect them to work from our knowledge of real minds. In particular, cognitive frames are of great use in helping readers to fill gaps in storyworlds. As possible world theorists such as Lubomír Doležel (1998), Thomas Pavel (1986), Marie-Laure Ryan (1991) and Ruth Ronen (1994) have shown, storyworlds differ ontologically from the real world because they are incomplete. Every storyworld contains ontological gaps or ‘spots of indeterminacy’ in a phrase of Roman Ingarden’s (1973:246) that was also used by Wolfgang Iser (1978:170). These gaps constitute the difference between, on the one hand, the combination of the story and the discourse that constitutes the text; and, on the other hand, the storyworld. No discourse could ever be long enough to say in its story all that could be said about the whole storyworld. As a result, fiction is necessarily incomplete. The reader can cope with the gaps in the continuing consciousnesses of fictional minds because, in the real world, we experience gaps in other, real minds too. From an aspectual point of view, another mind is sometimes present to us (when we are with that person) and sometimes absent. Our real-world cognitive frame enables us to construct a continuing consciousness for the absent person, unless we suffer from an abnormal condition such as autism that causes ‘mind-blindness’. When we see people again we generally attempt to reconstruct what they have been doing since we last saw them in order to work out roughly how they are feeling now. We reconstruct their narrative. This process is very similar to the activities of readers. We bring to the reading process our real-world knowledge of how to fill gaps and construct narratives for actual people. Real absences equate to fictional gaps. Much of the language used in the discourse of Middlemarch invites the sort of cognitive reading that I describe below. Many of the points that I make about mind reading, for example, are, as you will see, explicitly made in the novel itself. The text refers several times to ‘other minds’ (for example, 401, 504 and 530). It also uses phrases such as ‘other people’s states of mind’ (536), ‘mental action’ (546), ‘social action’ (124), ‘the boundaries of social intercourse’ (64) and ‘consciousness of interdependence’ (64). In addition, the novel explicitly anticipates Bakhtin’s notion of the ‘word with a sideways glance’: ‘The vicar’s frankness seemed not of the repulsive sort that comes from an uneasy consciousness seeking to


155

forestall the judgement of others’ (119); and: ‘an uneasy consciousness heareth innuendoes’ (206).

4. Continuing Consciousness Frame Cognitive science can be applied to the specific issue of the construction of fictional minds by helping narratology to analyze the cues that enable readers to create the effect of characters’ mental functioning. Cognitive frames enable readers to comprehend the fictional minds that are contained in narrative texts. Specifically, we are able to follow the workings of characters’ minds by applying what I call the continuing consciousness frame. This key cognitive frame comprises the ability to ascribe consciousness to narrative agents. The reader uses existing or pre-stored knowledge of other minds in the actual world in order to process the emergent knowledge that is supplied by fictional mind presentations. The everyday work that we put into constructing other real minds prepares us, as readers, for the work of constructing fictional minds. Because fictional beings are necessarily incomplete, frames are required to supply the defaults that fill the gaps in the storyworld and provide the presuppositions that enable the reader to construct continually conscious minds from the text. One key default setting is the assumption that a consciousness will continue throughout the text until interrupted, as in life, by death or absence. Another is that characters will think and act in certain fundamental respects like real people. To make sense of a text, the reader collects together all of the isolated references to a specific proper name and constructs a consciousness that continues in the spaces between the various mentions of the character with that name. The reader strategy is to join up the dots. In particular, the reading process is very creative in constructing coherent and continuous fictional consciousnesses from what is often a bare minimum of information. We frequently finish novels with a strong sense of the individual personality of a particular character. However, if we were to take the trouble to count up the specific references to that character, we might be surprised at how little there is in the text on which we have based our vivid impressions. When I re-read Middlemarch recently, I was surprised to find that the famous scene in which Lydgate finds himself unexpectedly engaged to Rosamond occupies less than one page (208) in the Norton edition. It is the continuing consciousness frame that enables readers to generate so much information from so little source material. As readers we strongly prefer to read a text for maximum cognitive payoff. We always try to get as much information as we can from a text. David Herman has suggested that ‘current research indicates that we

156

Alan Palmer

would do well to ask, not just about story structure as such, but about the patterned, nonrandom ways in which readers and listeners tend to impute structure to certain strings of events presented in discourse’ (1999: 8). In my view, the processing strategies that are used by readers to infer characters’ inner lives are a central way in which structure is imputed to strings of events. Generally, events in the storyworld are of little importance unless they become the experiences of characters. We follow the plot by following the workings of fictional minds. Am I saying that this process of following characters’ fictional mental functioning is both a necessary and a sufficient condition for narrative comprehension? This is not an easy question to answer in those terms. My feeling is that it is too weak to say that it is necessary (after all, who would disagree?) and perhaps a little too strong to say that it is sufficient (because we need to have a good understanding of the physical features of the storyworld as well). What I would prefer to say is that it is the fundamental and principal way by which we understand narrative. A number of narrative theorists have referred to aspects of the continuing consciousness frame from within their own theoretical frameworks. Monika Fludernik (1996) puts the notion of experientiality at the center of her perspective on narrative. Mieke Bal explains the difference between the two editions of her book Narratology (1985 and 1997) in terms of a new and growing emphasis on subjectivity: ‘This attention paid to subjectivity is, indeed, the basic tenet of the theory presented in this book’ (1997:11). Marie-Laure Ryan (1991), Menakhem Perry (1979), Catherine Emmott (1997) and Ruth Ronen (1994) are among the other narrative theorists who have added substantially to our knowledge of how readers create storyworlds generally and also the characters that inhabit them. How can the conceptual framework described above be applied to a specific fictional mind? I begin by setting out below a schedule of the various passages in the Middlemarch text that relate to the mind of Tertius Lydgate. It includes not only the occasions when Lydgate is ‘on-stage’, but also, and most importantly, those occasions when he is being talked about by other characters. Although the schedule is a little impressionistic and could be presented in slightly different ways, it nevertheless gives a fairly accurate impression of Lydgate’s presence in the novel. Passage Passage Passage Passage Passage Passage Passage

1 2 3 4 5 6 7

(61–69): 8 pages (then a 9 page gap) (78–86): 8 pages (10 page gap) (96–130): 34 pages (49 page gap) (179–88): 9 pages (8 page gap) (196–209): 13 pages (27 page gap) (236–45): 9 pages (46 page gap) (291–93): 2 pages (27 page gap)


157

Passage 8 (299–316): 17 pages (4 page gap) Passage 9 (320–23): 3 pages (17 page gap) Passage 10 (340–44): 4 pages (57 page gap) Passage 11 (401–15): 14 pages (26 page gap) Passage 12 (441–73): 32 pages (9 page gap) Passage 13 (482–531): 49 pages (8 page gap) Passage 14 (539–40): 2 pages (5 page gap) Passage 15 (545–52): 7 pages (23 page gap) Passage 16 (575): 1 page Total: 212 out of 578 pages (a little over a third of the total) A number of points are worth making about this schedule. First, the passages featuring Lydgate tend to be shorter in length than the gaps between those passages. Only three of the sixteen passages are above seventeen pages in length. (These three comprise a half of Lydgate’s total narrative.) Ten of the other passages are under ten pages in length. By contrast, seven of the gaps are over twenty pages long and three are over forty-five pages long. Reading Lydgate’s narrative within the context of the whole text is an extremely gappy experience. During these gaps, the reader has to continue to apply the continuing consciousness frame. That is, we have to keep in mind that Lydgate continues to exist within the reality of the storyworld and will have in his mind such important issues as the hospital chaplaincy affair, his relationship with Rosamond, his money troubles and the receding prospects of further medical research. Second, because the Lydgate storyworld is aspectual in that it views the whole Middlemarch storyworld only from Lydgate’s subjective point of view, a good deal of the latter is missing from the former. Lydgate appears to have no or very limited knowledge of: Dorothea’s relationships with Casaubon, Ladislaw, her sister Celia and Sir James Chettam; the relationships between Mary Garth, the Garth family, Fred Vincy and Farebrother; Mr Featherstone; Raffles and the precise nature of Bulstrode’s secret; and the extent of the relationship between Rosamond and Ladislaw and Dorothea’s discovery of them in a compromising situation. Third, it is a very artificial exercise to try to consider Lydgate’s mind in isolation from the other minds in the storyworld. Lydgate’s narrative is inextricably embedded in the whole narrative of the novel. For example, to say that his narrative comprises a third of the whole novel is very misleading. This third includes large portions of the narratives of Rosamond (obviously), and also Farebrother, Bulstrode, Dorothea, Casaubon and Ladislaw. What has happened is that their narratives have been aspectually adjusted, and those characters are now seen mainly from his cognitive and ethical viewpoint.

158

Alan Palmer

Fourth, it is striking how Lydgate’s narrative is dominated by conversations in which he takes part and the discussions about him when he is not present. His mind is actively engaged with its social context. As the philosopher Gilbert Ryle explains, to ‘talk of a person’s mind … is to talk of the person’s abilities, liabilities, and inclinations to do and undergo certain sorts of things, and of the doing and undergoing of these things in the ordinary world’ (1963:190). This is precisely the picture that emerges from the schedule. There are only five passages of private thought (totalling fourteen pages). However, this label is rather inadequate because these passages are also profoundly social in nature. They frequently refer to past actions, are dominated by plans for future action (see section 6.1) and are informed by his mind’s dialogic relationships with other minds (see sections 6.2 and 6.3). Finally, notwithstanding the third point, these passages, embedded as they are within the larger Middlemarch narrative, add up to a coherent and continuous narrative in their own right. They would require the addition of only a little explanatory material to make an excellent short novel. The passages comprise the Lydgate storyworld because they consist of the whole Middlemarch storyworld as seen from Lydgate’s aspectual point of view. This view of Lydgate’s mind as a narrative that is embedded within the whole text brings us to the next section.

5. Embedded Narratives A key tool for the study of fictional minds is Marie-Laure Ryan’s notion of embedded narratives, which I am extending by using it to mean the whole of a character’s mind in action: the total perceptual and cognitive viewpoint, ideological worldview, memories of the past and the set of beliefs, desires, intentions, motives and plans for the future of each character in the story as presented in the discourse. The term embedded narratives is intended to convey the point that the reader has a wide range of information available with which to make and then revise judgments about characters’ minds. The relationship between the continuing consciousness frame and the notion of embedded narratives is this: the former is the means by which we are able to construct fictional minds; the latter is the result of that construction. Embedded narratives are the product of the application of the continuing consciousness frame to the discourse. When we examine embedded narratives, we see the storyworlds of novels from the limited and aspectual viewpoints of the inhabitants of those storyworlds. The results of an analysis of a single fictional mind can then be enmeshed into analyses of the other minds in the storyworld,


159

with their own embedded narratives, their own motives, intentions and plans. The combination of all the embedded narratives in a text forms the plot of the novel. A complete picture of an aspectual, subjectively experienced storyworld results. The storyworld is aspectual in the sense that its characters can only ever experience it from a particular perceptual and cognitive aspect at any one time. As the philosopher John Searle explains, ‘[w]henever we perceive anything or think about anything, we always do it under some aspects and not others’ (1992:156–57). The term embedded narrative is, in a sense, simply a label for an approach that has always been used by literary critics in practice, but which has not yet been sufficiently theorized. Some, but only some, of the material that is covered by the term has been categorized separately within narratology under the various headings that I mentioned earlier: consciousness representation, story analysis, focalization, characterization and so on. The usefulness of the embedded narrative label is that it encourages a detailed, precise, functional and inclusive approach towards the whole of a fictional mind in its social and physical context. It clarifies the process by which the reader constructs a series of encounters with a particular fictional mind into a narrative that is coherent and continuous. Currently, there is a hole in literary theory between the analysis of consciousness, characterization, focalization and story. Oddly, a good deal of fictional discourse is, in my experience, situated precisely within this theoretical hole. The theoretical gap between consciousness and characterization is particularly difficult to understand. For example, Rosamond’s refinement is ‘beyond what Lydgate had expected’ (109). It would be reasonable to infer from this statement a single mental event: say, Lydgate realizes in a self-conscious flash of intuition that this is how he feels about Rosamond. It is also his settled belief that this is so, and his belief is a state of mind that will exist over time, and will be true of his mind whether or not he is thinking about Rosamond at any given moment. It is also a reflection of the dispositions that are characteristic of his mind and therefore of his character or personality: his tendency to be arrogant, to be class conscious and to be attracted to women. These short, medium and long term elements of his mind relate to both consciousness and characterization and cannot be separated. Here is another among countless examples: ‘Lydgate had often been satirical on this gratuitous prediction, and he meant now to be guarded’ (199). Again, there is a complex and interesting balance between characterization (the past), current mental events and states (the present) and the kind of functional, purposive, problem-solving mental functioning that is oriented towards the future.

160

Alan Palmer

6. Three Subframes I will now discuss three of the subframes of the continuing consciousness frame. (1) The relationship between fictional thought and fictional action: the distinction between thought and action is not as clear-cut as narrative theorists have assumed. (2) Intermental thought, or shared, group or joint thinking: much of our thought is public and social in nature, and a good deal of our cognition, action and even identity is socially situated or distributed amongst other individuals. (3) Doubly embedded narratives: the fact that versions of characters exist within the minds of other characters. These three subframes utilize fundamental aspects of our real-world knowledge of the mental functioning both of ourselves and of others. They are certainly not the only subframes; others include non-verbal consciousness, dispositions (although I refer to them briefly), the emotions (a particularly important one) and non-consciousness (it is continually stressed that there are many aspects of his life that Lydgate has always taken for granted and never consciously thought about). Before I go onto a more detailed application of the subframes to Lydgate’s mind, here is a simple example from Evelyn Waugh’s Vile Bodies that can be used to illustrate all three: ‘The three statesmen hid themselves’ (1996:86). (It is worth mentioning here that this sentence would not even be considered as a presentation of consciousness at all within traditional narratological approaches.) This is a description of an action, but it goes further in identifying the accompanying mental processes than a statement such as ‘They stood behind the curtain’, which leaves more work for the reader to do in deciding why they are standing behind the curtain. The sentence can be decoded in consciousness terms as follows: the three agreed that it was in their interest to conceal themselves from someone, realized that it was possible for them to do so and decided together to take the action of hiding. In this way, the reader, as part of the process of understanding narratives, has to translate passages of action description into mind description in the manner of a ‘psychological’ novel. That is how the first subframe works. With regard to the second, the decision to hide is an intermental decision taken together by all three statesmen. As the psycholinguist James Wertsch has explained, ‘the notion of mental function can properly be applied to social as well as individual forms of activity’ (1991:27), and ‘the terms mind and mental action can appropriately be predicated of dyads [that is, pairs] and larger groups as well as of


161

individuals’ (1991:14). During the brief period during which they come to their joint decision, the three statesmen became an intermental unit. The third subframe, doubly embedded narratives, works like this: this decision to hide involves the creation within the embedded narratives of the statesmen of a doubly embedded narrative for other characters who may wish to find them and who, the statesmen anticipate, will not be able to because of the decision to hide. I am not saying that the minds that comprise a particular intermental unit necessarily fuse completely and become totally available to each other. Towards the end of Middlemarch, Dorothea realizes that she and Rosamond could never be together again with the same thrilling consciousness of yesterday within them both. She felt the relation between them to be peculiar enough to give her a peculiar influence, though she had no conception that the way in which her own feelings were involved was fully known to Mrs Lydgate (549).

There was an intermental moment between them and so, just for that brief moment, the two women formed an intermental unit. However, it was not a complete fusion. They now have different understandings of the event. The three statesmen may be hiding for different reasons and may misunderstand the reasons of the others and so on. (For more on groups, see Margolin [1996] and [2000].) In applying the three subframes to Lydgate’s mind, I will, after discussing thought and action in section 6.1, consider the second and third subframes together in sections 6.2 and 6.3 because intermental thought and doubly embedded narratives can be difficult to disentangle. Lydgate’s relationship with the intermental unit of Middlemarch, which I refer to as the Middlemarch mind (the subject of section 6.2), is negotiated in terms of the doubly embedded narratives that they both have of each other. Lydgate’s relationship with Rosamond (section 6.3) fails to become an intermental unit because the doubly embedded narratives that they have of each other differ so widely from their own internal narratives. In essence, these two sections are about how Lydgate’s mind works, how he thinks other minds work, how other people think his mind works, how he thinks other people think his mind works and so on. For a fascinating analysis of even deeper complexities of this sort, see Zunshine (2003). 6.1 Action The topic of action is a large one and so, for reasons of space, I will limit myself to making a few, very brief points here. (For more on action, see Palmer [2004].) The presentation of Lydgate’s mind in the discourse is

162

Alan Palmer

dominated in a number of different ways by the concept of action. His mind is generally described in terms of what he does. As Ryle says, when we talk about the mind we talk about the doing and undergoing of things in the ordinary world. The reader is able to see his mind in action, amongst other ways, in his medical work and specifically in his diagnoses of illness. It is also visible during the various conversations in which he takes part. For example, in our first encounter with him he is ‘listening gravely’ (60) while nonsense is being talked to him. This is dramaturgical action: evoking in a public audience a certain image or impression (Wertsch 1991:10). A good deal of thought report is what I refer to as contextual: the short unobtrusive clauses, phrases or even single words describing a character’s thought processes that are often combined with descriptions of actions. Much of the contextual thought report that refers to Lydgate’s mind is used to explain the purpose of his actions. These actions are often the speech acts that occur during conversations. The resulting combination of a description of an action and the reason for it as contained in the contextual thought report is very common: ‘Lydgate, not willing to let slip an opportunity of furthering a favourite purpose, ventured to say …’ (302). Also: ‘Lydgate’s ear had caught eagerly her mention of the living, and as soon as he could, he reopened the subject, seeing here a possibility of making amends for the casting-vote he had once given with an ill-satisfied conscience’ (342). Here is a more complex example: ‘He sat looking at her, and did not rise to pay her any compliments, leaving that to others, now that his admiration was deepened’ (110). Here, an action (looking), a non-action (not complimenting her), the decision that is the specific reason for the non-action (leaving it to others) and the general state of mind that is the basic reason for all three of these (admiration) are combined in a sentence that causally links action and consciousness together. Within the passages contained in the schedule set out above, there are five stretches of text that may be labelled ‘private thought’. However, this term is rather misleading. These passages are full of actions (‘He had quitted the party early’ [63]) and also of plans for future actions (‘He had come to Middlemarch bent on doing many things’ [64]). In the first and second passages of private thought (64 and 112–115), he is deciding not to marry and planning his future research. In the third (122–124), he is wondering how to vote on the chaplaincy question. In the fourth (404–408), he is trying to work out what to do about his growing money troubles and his inability to get Rosamond to understand them. In the final passage (509–511), he is thinking over past actions such as his marriage and his acceptance of a loan from Bulstrode, the joint actions of others (‘The general black-balling had begun’ [511]) and his own actions in response (‘And yet how was he to set about vindicating himself?’ [509]).


163

6.2 Lydgate’s Relationship with Middlemarch One of the most important characters in the novel is the large, well-defined intermental unit of the town of Middlemarch itself. It is referred to in a number of ways. Sometimes the descriptions are neutral: ‘his neighbours’ (96); and ‘the town’s talk’ (204). Sometimes they are rather arch constructions that are characteristic of the voice of the narrator: ‘the public belief’ (527); ‘all the world round Tipton’ (32); ‘in various quarters’ (314); ‘family party’ (240); ‘the laity’ (306); ‘in Middlemarch phraseology’ (511); and ‘the world—that is to say, Mrs Cadwallader the Rector’s wife, and the small group of gentry with whom he visited in the north-east corner of Loamshire’ (4). At other times, they betray the bitterness and frustration of the individual who is on the receiving end of the intermental consensus: ‘the petty medium of Middlemarch’ (129); and ‘Middlemarch gossip’ (240). In addition, the Middlemarch mind is sometimes presented in what may be called intermental free indirect discourse. The narrator expresses a view that, it soon becomes clear from the context, is in fact the view of the townspeople. For example: ‘And how should Dorothea not marry?’ (3); and ‘It was clear that Lydgate … intended to cast imputations on his equals’ (126). A particularly striking form of words that identifies Middlemarch as a group mind is this: ‘Middlemarch, in fact, counted on swallowing Lydgate and assimilating him very comfortably’ (105). This sentence beautifully frames the relationship between Lydgate and Middlemarch for the remainder of the novel. Middlemarch has a doubly embedded narrative of Lydgate as the idealistic young doctor who comes to the town and wishes to mould it in conformity to his wishes. However, he will be taught a lesson and will discover that it is he who has to change. A number of narrative theorists such as Menakhem Perry (1979) have drawn attention to the primacy effect. That is, the initial reading frames that are set up at the beginning of a text have long-lasting effects, and they tend to persist until the reader is compelled by the accumulating weight of contrary evidence to abandon them and set up new frames. Lydgate is introduced within an intermental frame. His identity is socially distributed before we meet him, and there are a number of discussions of him throughout the novel that continue the town’s exploration of his identity. It is striking that the early part of the novel contains far more information on the ‘Lydgates’ that exist in the minds of other characters than it does on the ‘Lydgate’ that emerges from direct access to his own mind. It is worth examining the beginning of this process in a little detail. The reader begins the initial, tentative and hypothetical construction of Lydgate’s embedded narrative by making use of the doubly embedded narratives of other characters. We first hear of him indirectly, while Lady Chettam and Mrs Cadwallader are discussing him. ‘Tell me about this new young surgeon,

164

Alan Palmer

Mr Lydgate. I am told he is wonderfully clever: he certainly looks it—a fine brow indeed’ (61). Mrs Cadwallader replies that: ‘He is a gentleman … He talks well’ (61). Lady Chettam agrees that he is ‘really well connected … One does not expect it in a practitioner of that kind’ (62). Mrs Cadwallader then notices that Dorothea Brooke ‘is talking cottages and hospitals with him … I believe he is a sort of philanthropist’ (62). So, we find out before we actually meet him that he is apparently young, clever, good-looking, a well-connected gentleman, someone who talks well, a sort of philanthropist and innovative and successful. These few words on the page are now transformed by the reader into an already pre-existing imaginary individual with a past that is part of the Middlemarch storyworld. His embedded narrative started twenty-five odd years ago, and the reader is now engaged in reconstructing it. This conversation is largely a consideration of Lydgate’s mind. Some features are obviously related to his mental life: being clever, philanthropic and successfully innovative. Others are slightly more indirect: ‘talking well’ is a description of behavior that implies a series of mental attributes such as cleverness, confidence, awareness of others and so on. Being a gentleman is yet more indirect, but presumably has implications for the way the mind works. In addition to these explicit qualities, it is also part of the competence of the reader to construct, this time by indirect means, some aspects of Lydgate’s mental life that are implicit in what we have been told. Let us say, for example: having the above qualities makes it likely that he is not only self-confident and ambitious, but also altruistic, imaginative and idealistic. The reader is using material on Lydgate’s social and publicly available mind that has been refracted through the conflicting world-views of the two different embedded narratives of Mrs Cadwallader and Lady Chettam. A version of his mind exists within their minds. Their minds are interacting with, conflicting with, interrogating and so on the constructions that they have formed of his mind. They disapprove of his being both a gentleman and a doctor and also of the fact that he is a doctor with ideas about the advancement of medicine. Although the characters do not explicitly speculate about the causal network behind Lydgate’s behavior, there is an implicit puzzlement over the motives that a gentleman would have for wanting to become a doctor. Lydgate’s own motivation becomes more explicit in the direct access to his mind later in the passage. Later events appear to show that the views of the two characters were fairly accurate in their discussions of Lydgate’s mind. In fact, to use a familiar but significant phrase, there are some respects in which he may not ‘know his own mind’. In the next chapter Lydgate is scornful about the possibility of losing his balance and falling in love, and we find out later that this is precisely what he does do.


165

The novel contains a number of scenes in which various groups of townspeople discuss Lydgate’s mind. An intermental consensus emerges from a variety of intramental perspectives and a joint doubly embedded narrative is constructed. These scenes result in an interesting dual perspective on the Middlemarch mind. Seen from one perspective, although these conversations create a consensus, they also contain intramental dissent. People disagree over, for example, the precise extent of Lydgate’s guilt in the matter of the bribe. However, seen from the point of view of Lydgate’s mind, the Middlemarch mind is fully shared, uniform and monolithic. And in a sense he is not necessarily wrong. Once the consensus has emerged, the reservations of some individuals around the edges do not count for very much. Sometimes the town’s views are related simply to the facts of the case and, interestingly, the speculation that Lydgate’s windfall is a loan from Bulstrode turns out to be factually correct. More often, though, the town is concerned with the ascription of motives: ‘The inferences were closely linked enough: the town knew of the loan, believed it to be a bribe, and believed that he took it as a bribe’ (504). The Middlemarch mind is always inclined to ascribe unworthy reasons to Lydgate’s actions: ‘After this, it came to be held in various quarters that Lydgate played even with respectable constitutions for his own purposes’ (314). The cognitive operations involved in the ascription of motives can become quite complex: ‘Thus it happened that on this occasion, Bulstrode became identified with Lydgate and Lydgate with Tyke; and owing to this variety of interchangeable names for the chaplaincy question, diverse minds were enabled to form the same judgement concerning it’ (126). A fully fledged narrative for Lydgate emerges from the network of ascription: ‘There was hardly ever so much unanimity among them as in the opinion that Lydgate was an arrogant young fellow and yet ready for the sake of ultimately predominating to show a crawling subservience to Bulstrode’ (312). At other times the consensus consists simply of abuse: ‘[T]hey agreed that Lydgate was a jackanapes’ (126). Throughout the novel, a dialogic relationship exists between the intramental Lydgate and the intermental Middlemarch. He is concerned with anticipating, reconstructing, arguing against the common view of the workings of his mind. For example, his mind is in a conflicted dialogue with the minds of the various groups and individuals concerned with the question of the chaplaincy of the hospital. Sometimes this relationship is expressed in general terms: ‘For the first time Lydgate was feeling the hampering threadlike pressure of small social conditions, and their frustrating complexity’ (124). At other times, a specific issue brings home to Lydgate the power of the intermental mind: ‘The affair of the chaplaincy

166

Alan Palmer

remained a sore point in his memory as a case in which the petty medium of Middlemarch had been too strong for him’ (129). Much of the conflict occurs within the context of his medical work: ‘He was impatient of the foolish expectations amidst which all work must be carried on’ (181). This ‘feud between him and the other medical men’ (188) is regularly explained in terms of mind-reading difficulties. Sometimes, Lydgate’s inability to develop a doubly embedded narrative that is sufficiently sophisticated to make his social relations easier is focussed on the group mind: ‘Lydgate’s remark did not meet the sense of the company’ (107); and: ‘But Lydgate had not been experienced enough to foresee that his new course would be even more offensive to the laity’ (306). At other times, it is focussed on an individual: ‘Lydgate had really lost sight of the fact that Mr Chichely was his Majesty’s coroner’ (108). However, the novel contains a balanced picture of Lydgate’s doubly embedded narrative of Middlemarch. In some ways, he performs quite well. There is plenty of evidence to show that, despite his undoubted arrogance, he does at least make an effort to reconstruct other minds and is quite successful in doing so. For example: ‘But would the end really be his own convenience? Other people would say so, and would allege that he was currying favour with Bulstrode’ (124). Later, he hopes that Dorothea’s actions can ‘clear [him] in a few other minds’ (530). He can also be quite sensitive in his constructions of individual doubly embedded narratives: ‘For Lydgate was acute enough to indulge him with a little technical talk’ (311); ‘That there might be an awkward affair with Wrench, Lydgate saw at once’ (180); and, ‘Lydgate was conscious of having shown himself something better than an everyday doctor, though here too it was an equivocal advantage’ (311). He is quite flexible and not overtly confrontational, despite feeling ‘some zest for the growing though half-suppressed feud’ (188). When he makes a successful diagnosis and another doctor does not, ‘Lydgate did not make the affair a ground for valuing himself or (very particularly) despising Minchin’ (311). Generally speaking, ‘We see that he was bearing enmity and silly misconception with much spirit, aware that they were partly created by his good share of success’ (314). When things become more serious during the Bulstrode loan crisis, he becomes very sensitive. He notices ‘a peculiar interchange of glances when he and Bulstrode took their seats’ (502); and, while he helps Bulstrode out of the room after he is accused, ‘It seemed as if he were putting his sign-manual to that association of himself with Bulstrode, of which he now saw the full meaning as it must have presented itself to other minds’ (504). Before going on to discuss Lydgate’s relationship with Rosamond in the next section it may be useful, as a contrast, to describe briefly Lydgate’s friendship with Farebrother, an inhabitant of the town who is not part


167

of the Middlemarch mind. Lydgate has a fairly full and accurate doubly embedded narrative of Farebrother. As their friendship develops, difficulties arise when Farebrother tries to help Lydgate. Initially, ‘Lydgate took Mr Farebrother’s hints very cordially, thought he would hardly have borne them from another man’ (315). (This sentence, by the way, is a good example of the inseparability of consciousness and characterization that I referred to earlier: it consists of mental events [cordial taking], states of mind [feelings for Farebrother] and character [general unwillingness to take hints].) Later, Farebrother offers more specific help. First, this is how it appears to Lydgate: ‘He knew as distinctly as possible that this was an offer of help to himself from Mr Farebrother, and he could not bear it’ (446). Next, this is how it appears to Farebrother: ‘Could this too be a proud rejection of sympathy and help? Never mind; the sympathy and help should be offered’ (492). What is significant here is that the empathy or mind reading between the two men is very accurate. Lydgate knows that Farebrother knows that Lydgate knows et cetera that the offer of help is embarrassing but has to be made. The mutual ascription of motives and states of mind is far more successful in this case than between Lydgate and his wife. 6.3 Lydgate’s Relationship with Rosamond The background to the following discussion is the default assumption that a marriage will be an intermental unit. However, the reality is that they both have deeply developed doubly embedded narratives of each other that are very different indeed from their actual narratives, and so, as a result, there is no evidence at all of any intermental thinking. In the novel that I quoted from earlier, Vile Bodies, the opposite is the case. Within a social group called the ‘Bright Young Things’ who are active in London in 1930, there is a good deal of intermental thought, but no doubly embedded narratives. They act together, but have no interest in each other’s inner lives. (See Palmer [2004] chapter seven for more detail.) Lydgate and Rosamond are interested: it is just that they are completely wrong about each other. For example, the narrator stresses that Lydgate and Rosamond lived in a world of which the other knew nothing, it had not occurred to Lydgate that he had been a subject of eager meditation to Rosamond … In Rosamond’s romance, it was not necessary to imagine much about the inward life of the hero, or of his serious business in the world (114).

Reading the engagement scene (208) carefully, it is clear that it comes about through a series of misunderstandings regarding what the other is thinking. Rosamond was ‘keenly hurt by Lydgate’s manner’ which was

168

Alan Palmer

simply the result of embarrassment. The long-term result is predictable: ‘Between him and her indeed there was that total missing of each other’s mental track, which is too evidently possible even between persons who are continually thinking of each other’ (405). As his money troubles grow, Lydgate decides to try to make their relationship intermental, to bring about a meeting of minds, and he fails. For example: ‘Perhaps Lydgate and she had never felt so far off each other before’ (409). He refers explicitly to the possibility of making the marriage an intermental unit, when he suggests that ‘›there are things which husband and wife must think of together‹’ (410). But soon after, ‘when he spoke again it was more in the tone of a man who forces himself to fulfil a task’ (410). During the crisis, ‘He did not speak to her on the subject, and of course she could not speak to him’ (522). The narrator concludes: ‘The beginning of mutual understanding and resolve seemed as far off as ever; nay it seemed blocked out by the sense of unsuccessful effort’ (524). It does not help the accuracy of Lydgate’s doubly embedded narrative of Rosamond that she is so adept at dramaturgical action. Even at an early stage in their relationship she is very accurate in her estimate of the kind of behaviour that will attract him. For example: ‘In two minutes he was in the room, and Rosamond went out, after waiting just long enough to show a pretty anxiety conflicting with her sense of what was becoming’ (180). Later, she is still careful to control Lydgate’s image of her: ‘Rosamond, however, was on her side much occupied with conjectures, though her quick, imitative perception warned her against betraying them too crudely’ (244). During the crisis, her private impression management becomes rather more perfunctory: ‘She received his kiss and returned it faintly, and in this way an appearance of accord was recovered for the time’ (413). However, her public impression management is studied and elaborate. During a party: Rosamond was perfectly graceful and calm, and only a subtle observation … would have perceived the total absence of that interest in her husband’s presence which a loving wife is sure to betray, even if etiquette keeps her aloof from him … In reality, however, she was intensely aware of Lydgate’s voice and movements; and her pretty good-tempered air of unconsciousness was a studied negation by which she satisfied her inward opposition to him without compromise of propriety (443).

The narrator frequently and explicitly draws attention to the fact that Rosamond constructs a doubly embedded narrative of Lydgate that owes far more to the emotional needs of her mind than to the actual workings of his. They meet and she is attracted: ‘Yet this result, which she took to be a mutual impression, called falling in love, was just what Rosamond had contemplated beforehand’ (80). ‘Rosamond, in fact, was entirely occupied not exactly with Tertius Lydgate as he was in himself, but with his


169

relation to her’ (115). Later, looking back, the narrator comments that ‘the Lydgate with whom she had been in love had been a group of airy conditions for her’ (457). When the problems arise, her reaction is predictable: ‘The thought in her mind was that if she had known how Lydgate would behave, she would never have married him’ (412). ‘In her secret soul she was utterly aloof from him’ (448). ‘Open-minded as she was [about the bribe allegation], she nevertheless shrank from the words which would have expressed their mutual consciousness, as she would have shrunk from flakes of fire’ (518). ‘Even this trouble, like the rest, she seemed to regard as if it were hers alone. He was always to her a being apart, doing what she objected to’ (523). Her solipsism even extends to the class implications of having Lydgate as a husband: ‘She was so intensely conscious of having a cousin who was a baronet’s son staying in the house, that she imagined the knowledge of what was implied by his presence to be diffused through all other minds’ (401). To summarise, the narrator comments that ‘she had been little used to imagining other people’s states of mind except as a material cut into shape by her own wishes’ (536). ‘It seemed that she had no more identified herself with him than if they had been creatures of different species and opposing interests’ (412). The narrator is kinder to Lydgate and acknowledges that he makes an effort to understand Rosamond’s mind. But there are difficulties: ‘Perhaps it was not possible for Lydgate, under the double stress of outward material difficulty and of his own proud resistance to humiliating consequences, to imagine fully what this sudden trial was to a young creature who had known nothing but indulgence’ (411). Nevertheless, ‘he had made many efforts to draw her into sympathy with him about possible measures for narrowing their expenses’ (448). For example, he suggests that ‘›We two can do with only one servant‹’ (448). Lydgate then acknowledges the distance between their narratives: ‘To many women the look Lydgate cast at her would have been more terrible than one of anger: it had in it a despairing acceptance of the distance she was placing between them’ (412). Regarding the relationship between Dorothea and Ladislaw, It was significant of the separateness between Lydgate’s mind and Rosamond’s that he had no impulse to speak to her on the subject; indeed, he did not quite trust her reticence towards Will. And he was right there; though he had no vision of the way in which her mind would act in urging her to speak (414).

The ascription is interesting here. He correctly assumes that she will speak to Ladislaw, but does not know about the thought processes that will lead her to do so. In the end, he renounces thoughts of an ‘ideal wife’ and begins to plan the adjustments necessary for living with the ‘real wife’. In my terms, he

170

Alan Palmer

has to ensure that his doubly embedded narratives will correspond more closely in the future to her actual narrative. As if to compensate, he fantasises about a virtual intermental unit: ‘he was beginning now to imagine how two creatures who loved each other, and had a stock of thoughts in common, might laugh over their shabby furniture, and their calculations how far they could afford butter and eggs’ (484). It is highly significant that the only references that I can find to Lydgate’s and Rosamond’s joint actions are these hypothetical or imaginary situations regarding making do with one servant, laughing over their furniture and calculating what they could afford. This married couple share no actual joint actions within the reality of their storyworld. 7. Conclusion When a phenomenon is identified which does not fit within a well-established paradigm, an anomaly is created. And, as Thomas Kuhn explained in The Structure of Scientific Revolutions: Discovery commences with the awareness of anomaly, that is, with the recognition that nature has somehow violated the paradigm-induced expectations that govern normal science. It then continues with a more or less extended exploration of the area of anomaly. And it closes only when the paradigm theory has been adjusted so that the anomalous has become the expected (Kuhn 1996:52–53).

The widespread and pervasive existence of fictional intermental thought constitutes an anomaly within the traditional narratological paradigm for the presentation of consciousness in the novel. The proof is the invisibility of intermental thought within the current theory. However, following a more or less extended exploration of this area and all of the other aspects of the whole of the social mind in action, the paradigm theory can be adjusted to take full account of the importance of fictional minds. Intermental thought will then be anomalous no longer; it will be expected. References Bal 1997 Bal, Mieke: Narratology: Introduction to the Theory of Narrative. 2nd ed., Toronto 1997. Damasio 2000 Damasio, Antonio: The Feeling of What Happens: Body, Emotion and the Making of Consciousness. London 2000. Dennett 1991 Dennett, Daniel C.: Consciousness Explained. Harmondsworth 1991.


171

Doležel 1988 Doležel, Lubomír: “Mimesis and Possible Worlds.” In: Poetics Today 9.3, 1988:475–96. Doležel 1998 Doležel, Lubomír: Heterocosmica – Fiction and Possible Worlds. Baltimore 1998. Eliot 1977 Eliot, George: Middlemarch. Ed. by Bert G. Hornback. New York 1977. Elster 1999 Elster, Jon: Alchemies of the Mind: Rationality and the Emotions. Cambridge 1999. Emmott 1997 Emmott, Catherine: Narrative Comprehension: A Discourse Perspective. Oxford 1997. Fauconnier 1997 Fauconnier, Gilles: Mappings in Thought and Language. Cambridge 1997. Fludernik 1996 Fludernik, Monika: Towards a ‘Natural’ Narratology. London 1996. Herman 1997 Herman, David: “Scripts, Sequences, and Stories: Elements of a Postclassical Narratology.” In: PMLA 112.5, 1997:1046–1059. Herman 1999 Herman, David: “Introduction: Narratologies.” In: Herman, David (ed.): Narratologies: New Perspectives on Narrative Analysis. Columbus 1999: 1–30. Herman 2002 Herman, David: Story Logic: Problems and Possibilities of Narrative. Lincoln 2002. Herman 2003 Herman, David: “Stories as a Tool for Thinking.” In: Herman, David (ed.): Narrative Theory and the Cognitive Sciences. Stanford 2003:163–92. Ingarden 1973 Ingarden, Roman: The Literary Work of Art: An Investigation on the Borderlines of Ontology, Logic, and Theory of Literature. Trans. by George C. Grabowicz. Evanston 1973. Iser 1978 Iser, Wolfgang: The Act of Reading. London 1978. Kuhn 1996 Kuhn, Thomas: The Structure of Scientific Revolutions. 3rd ed., Chicago 1996. Margolin 1996 Margolin, Uri: “Telling Our Story: On ‘We’ Literary Narratives.” In: Language and Literature 5.2, 1996:115–33. Margolin 2000 Margolin, Uri: “Telling in the Plural: From Grammar to Ideology.” In: Poetics Today 21.3, 2000:591–618. Nebel 1999 Nebel, Bernhard: “Frame-Based Systems.” In: Wilson, Robert A. / Keil, Frank C. (eds.): The MIT Encyclopedia of the Cognitive Sciences. Cambridge, MA 1999:324–326.

172

Alan Palmer

Palmer 2004 Palmer, Alan: Fictional Minds. Lincoln 2004. Pavel 1986 Pavel, Thomas G.: Fictional Worlds. Cambridge, MA. 1986. Perry 1979 Perry, Menakhem: “Literary Dynamics: How the Order of a Text Creates its Meanings.” In: Poetics Today 1.1–2, 1979:35–64, 311–361. Pinker 1997 Pinker, Steven: How the Mind Works. Harmondsworth 1997. Ronen 1994 Ronen, Ruth: Possible Worlds in Literary Theory. Cambridge 1994. Ryan 1991 Ryan, Marie-Laure: Possible Worlds, Artificial Intelligence, and Narrative Theory. Bloomington 1991. Ryle 1963 Ryle, Gilbert: The Concept of Mind. Harmondsworth 1963. Schank / Abelson 1977 Schank, Roger C. / Abelson, Robert P.: Scripts, Plans, Goals, and Understanding: An Inquiry into Human Knowledge Structures. Hillsdale 1977. Searle 1992 Searle, John R.: The Rediscovery of the Mind. Cambridge, MA. 1992. Turner 1991 Turner, Mark: Reading Minds: The Study of English in the Age of Cognitive Science. Princeton 1991. Waugh 1996 Waugh, Evelyn: Vile Bodies. Harmondsworth 1996. Wertsch 1991 Wertsch, James V.: Voices of the Mind: A Sociocultural Approach to Mediated Action. Cambridge, MA. 1991. Zunshine 2003 Zunshine, Lisa: “Theory of Mind and Experimental Representations of Fictional Consciousness.” In: Narrative 11.3, 2003:270–291.

‘Selective’ History Why and How ‘History’ Depends on Readerly Narrativization, with the Wehrmacht Exhibition as an Example1 Daniel Fulda (Köln)

This paper has three parts. In the first, I sketch certain basic principles for narratology of history in order to articulate the theoretical framework of the argumentation. Second, I ask what points of contact are possible between general narratology discussed in recent scholarship and the narratology of history. As will become clear, I am especially interested in the use of cognitivist approaches to bridge this gap. Finally I offer a case study. The most controversial act of telling history in the 1990s (in Germany), the so-called ‘Wehrmacht Exhibition’, will serve to test the explanatory power of a narratological analysis which conceives history as the product of a quasi-readerly (‘selective’) narrativization. 1. Some Basic Principles for a Narratology of History Within the field of historiography, ‘Narratology beyond Literary Criticism’ hardly figures as a novel theme. History writing has been conceived as a form of storytelling since Greek antiquity, which is to say, for almost as long as it has existed or there has been a concept of it (cf. Harth 1996: 834–836). Rhetorical theory thematizes and regulates history-writing as narratio through the 18th century (cf. Keßler 1982:42, 48f, 65–68, 71). Even after the emergence of history as a discipline in the 19th century, the representational work of the historian is consistently understood as a narration of the known past,2 although the writing of history becomes 1

2

I am very grateful to Joana Konova and Martin Tracey for translating this article into English. Cf. Mommsen (1905:10): ‘History is really nothing other than the clear perception of factual events, composed in part from the examination and sifting of the available evidence, and in part through the interweaving of the same in a way that accords with the knowledge of the people involved and the relations that exist to a narrative explicating cause and effect.’

174

Daniel Fulda

regarded as posterior to historical research and of lesser value than it (cf. Rüsen 1982:20f). To be sure, these ‘narratologies’ do not yet confront what historiography, at its present state of research, designates as narratively composed. For the crucial point of the recent discussion (which admittedly began some forty years ago) is not the fact that history narrates insofar as some history texts present an event, with a modest amount of detail and a certain vividness, as unfolding in time (and in so doing, does not reduce it to concepts and arguments). This understanding of narrative is already presupposed by the rhetorical concept of narratio (cf. Knape 2003:98). It is also presupposed by the commitment to narration that many historians use to dismiss treatments of history which they regard to be too philosophical or too sociological.3 On the contrary, the concept of narrative within contemporary theory of history is an epistemological concept. It concerns – the relation of narration to disciplinary status of historiography and, beyond this, – what makes past events count as ‘history’ in the first place. Narratology of history has its origin within Analytic philosophy in the sixties. It was elaborated, nevertheless, within very different disciplinary traditions,4 among them, the phenomenological hermeneutic of Paul Ricœur (1984 / 88), the ‘tropo-logic’ of Frank Ankersmit (1994a), and Jörn Rüsen’s attempt to mediate between science and the world of experience within his ‘Sinn-Historik’ (theory of making sense by historicization).5 Narration is understood here, first, ‘as a kind of explanation especially appropriate to the explanation of historical, as against natural, events and processes’ (White 1984:7).6 The foundation for its explanatory success is the intelligible logical consistency of the narration (cf. Danto 1965:233–256).

3

4 5 6

Reductions of history to philosophy were opposed above all by post-Hegelian historians of the mid-nineteenth century, cf. Droysen (1977:233): ‘The historian does not wrap speculative thoughts in factual realities; the facts that he recounts are themselves moments in his sequence of thoughts. ... The more the historian ... thinks in facts, the better he recounts.’ In the 1970s, the narrativity of historical writing was invoked primarily to oppose the use of social-scientific methods, cf. Mann (1979). Cf. Rigney’s overview of the use of the concept of ‘narrative’ (1991:594f). Cf. White’s magisterial survey of the discussion in the 1960s and 1970s (1984). For a collection of Rüsen’s essays in English, cf. Rüsen 1993. As the association of causal explanation and narration cited in footnote 2 shows, the historicism of the 19th century already pressupposes that the explanatory function of narration is discipline-specific, although it does not articulate this pressupposition explicitly. (Mommsen accordingly distinguishes ‘historical research’ from other disciplines on the grounds ‘that it is unable to bring its elements to properly theoretical development’; 1905:10.)

‘Selective’ History

175

In addition to that we have its synthesizing function. The modality of narrative makes possible a specific ordering and orienting perspective on the past and its genetic linkage to the present. Considered retrospectively, past happenings can acquire the processive logic of a coherent and purposive story. This experiential, non-law-based logic constitutes the form of scientific explanation appropriate for the humanistic and cultural disciplines. To be sure, narrative’s power to make plausible does not satisfy the standards of explanation within the natural sciences. Hayden White, for instance, is known for his thesis that the source of historiography lies in a purely ‘poetic act’, which itself draws more upon literature than science (White 1973:30). White refers principally to the emplotment as that which gives sense to historiographical representation. To narrate history thus entails following a particular pattern, investing a part of the past with a plot, which in turn constitutes its ‘deep structure’. With the help of this deep structure, the narrative motivates its own interpretation (i.e., the assignment of a meaning which is not contained in its factual elements alone; cf. White 1984:20–25). White’s uniformly poetological interpretation of historiographical narration has created quite a stir within the discipline of history as well as within the suddenly competent-feeling philologies—that is, it has at times provoked reductive defensive reactions and at times supported a ‘postmodern’ blurring of boundaries. Against the background of the other disciplinary traditions of continental Europe, which tend less toward the Anglo-Saxon dichotomization of scientific and hermeneutical, rhetorical, or cross-cultural methods, such exaggerations were soon corrected, and the scientific value of narrative’s literature-like method emphasized (cf. Ricœur 1984 / 88:vol. 1; Chartier 1997). A second basic feature of narration was less noticed in the general discourse on narration of the eighties and nineties: namely, that the past first becomes history through narrative structuring. In this vein, it is important to note that the term ‘history’ does not designate a structure which lies in the past itself, but rather represents a pattern for conceptualizing which provides us with the very capacity for seeing the ‘past’, along with our present and expected future, in that specific genetic context familiar to us. Past events are not essentially ‘history’, but rather become such when seen from a particular vantage point. Hans Michael Baumgartner was the first to elaborate the way in which narration constructs the perspective that transforms ‘bare’, amorphous events into structured history, distinguished by continuity and meaningful sequentiality. In his transcendental philosophy of history, the ‘basic feature of the narrative’ is identified as an ‘a priori scheme’ on which all reconstructions of history ‘are based—as the condition for the possibility of history’ (Baumgartner 1976:279).

176

Daniel Fulda

Baumgartner’s 1972 book was hardly noticed within the international discussion of history theory or within philological narratology,7 even though its thesis extends further than Hayden White’s emplotment. Baumgartner examines the narrativity not only of historiographical works, but also of the very concept of history that underlies such works. The cause of its neglect may lie, on the one hand, in a linguistic hurdle posed by the German language, and on the other, in Baumgartner’s transcendentalphilosophical method, which offers few points of departure to a text-oriented narratology. Baumgartner’s thesis was also lessened in value through the nearly simultaneous discovery of the historicity of our conception of history, that is, its formation in the second half of the eigthteenth century. This conceptual transformation comes to expression in German through the new ‘collective singular’, Geschichte, which for the first time synthesized all individual histories into one history.8 History as the specific structure of the past in its connection with the present and future does not follow automatically from the narration of stories about the past. Such (hi)stories (Geschichten) can be found in the earliest texts of western civilization. But these particular stories must be clearly distinguished from the narrative organization of all events which we know as ‘history’; in what follows, I denote history in this latter sense by capitalizing its letter ‘h’. To conceive the past in narratives—be it the individual past of one’s own life or be it in the larger sense of the past of certain peoples—is by no means to imply that the entire past in its relation to the present forms solely one unique History. In German, one single grammatical detail, namely the number and the article of the term Geschichte, indicates the crucial difference. Although there may be innumerable Geschichten (stories and histories) in which people have always conceived their pasts, on the other hand, the Geschichte (History) which covers the civilizations of the ancient world as well as the Chinese Middle Ages and the American Civil War, always exists exclusively in the singular and with the definite article. Reinhart Koselleck’s discovery that the totalizing conception of History did not always exist, but indeed has a history of its own, makes it impossible to proceed, as Baumgartner does, from an ahistorical (because a priori)

7

8

Thus, at a symposium sponsored by the Deutsche Forschungsgemeinschaft in 1980 entitled ‘Erzählforschung’ (cf. Lämmert 1982), only one contribution, by Hann-Jörg Porath, made reference to Baumgartner’s work, although Baumgartner himself chaired one of the symposium’s sessions. In the German-language standard work on narratology (Martinez/Scheffel 2002), Baumgartner’s work is passed over to the advantage of White’s Metahistory (ibid.:155–159). Cf. Koselleck (1975:647–663). For a brief account in English, see Koselleck (1985: 200f).


177

narrativity of History.9 If the mode of perception and thinking that is History is to be conceived in narrative terms, then this narrativity, for its part, must be historicized. The contradiction between an analysis of the concept of History that is, on the one hand, concept- and consciousnesshistorical, and on the other, transcendental-philosophical, can be resolved only through a demonstration that History’s specific came to be constituted over time (and was not always there), and that History was conceived from the beginning as a narrative structure. Building on this point, I advanced the thesis in the mid-1990s that the understanding of History took on its narrative structure around the year 1800, and so under the influence of the aesthetic thinking of the Goethe period (cf. Fulda 1996: 183–331, 411–443). The writing of History became possible at that time, because the sequence of events to be represented became understood as a Great story, and historiography was likewise organized on the model of literary texts. As early as 1978, Louis O. Mink pointed to the ‘storyformedness’ of our modern concept of History. He regarded it as the product of ‘a single unified story of the human past’ which emerged within the philosophy of history in the second half of the 18th century (cf. Mink 1978:135–141). From that time onward, ‘the idea that the past itself is an untold story has retreated from the arena of conscious belief and controversy to habituate itself as a presupposition in that area of our a priori conceptual framework’ (ibid.: 140). The narrative techniques of the historiography developed around 1800 made a decisive contribution to this habituation. In the exemplary work of historicism, Ranke’s Römische Päpste, written in the early 1830s, we can recognize ways of thinking and representational techniques formed in the literature of the Goethe period: e.g., the conversion of historical events into story form (often with plots like those of the Bildungsroman), the assigning of ideal tasks to important characters together with the withdrawal of the omniscient narrator, the immanent narrative explanation of historical processes, the symbolic concentration of the whole story in decisive moments, as well as the incorporation of seemingly trivial but nevertheless significant and / or vivid details (cf. Fulda 1996:344–410). History emerged as a discipline through the process of its aestheticization and, more particularly, its narrativization. Plot now expressed what was radically new within the notion, developed in the 18th century, of ‘the one

9

The historicity of the concept of ‘history’ itself is seldom considered within anglophone scholarship, perhaps because there has been no striking etymological transformation of this ‘collective singular’ in English. Megill (1995) is an exception is this regard; like the present author, he uses the majuscule H to foreground the concept of a continuous, all-encompassing ‘single History’.

178

Daniel Fulda

history’, because storytelling employed plot to lend coherence to the subject at hand.10 History (Geschichte) is a product of stories (Geschichten), more exactly, of those texts which narrate Geschichte / n (stories and history). History is less written down (wird aufgeschrieben) than it is continuously composed (immer erst erschrieben).11 2. Points of Contact between Narratology and a Narratology of History The words of the History to be composed resist the objectivistic error which contends that past events are in themselves ‘historical’ (in the narrativist sense of continuous, directed, and intelligible). The words do not conform as well to the perception of the past within a schema of narration. In order to emphasize that History provides a pattern for thinking that allows reality to be observed in a specific way, one must instead speak of ‘selective’ History—i.e., History which is constituted by being read (n.b., the origin of ‘selective’ from the latin legere ‘to read’). Without in any way pitting composition (erschreiben) and selection (erlesen) against one another, I wish today to place the latter in the foreground. The fact that, in recent years, scholars have discussed the narrative composition of History is understandable, given the increasing interest in historical writing. The extant discourse does not suffice, however, from the perspective of either narratology or the theory of history. The (narrativizing) ‘reading’ or ‘selection’ of the past’s representations is both more primary and more commonplace, since it is also practiced by those who merely receive representations of history yet do not produce it.12 Contra Hayden White, the narrativization of history 10

11

12

Peter Brooks also locates ‘the emergence of narrative plot as a dominant mode of ordering and explanation’ in the middle of the 18th century (1984:5f). Since that time, ‘Western societies appear to have felt an extraordinary need or desire for plots, whether in fiction, history, philosophy, or any of the social sciences, which in fact largely came into being with the Enlightenment and Romantics. As Voltaire announced and then the Romantics confirmed, history replaces theology as the key discourse and central imagination in that historical explanation becomes nearly a necessary factor of any thought about human society.’ Brook’s book is important because it accomplishes something rare in narratology: historicization of narrative structures. White (1999:1) says something similar, if not quite as pointedly: ‘our experience of history is indissociable from our discourse about it; this discourse must be written before it can be digested as history.’ Likewise, Angehrn speaks not only of a ‘writing’ but also of a ‘reading’ of history (1999). Unlike the concept of ‘selection’ proposed here, this reading does not denote the construction of history in an act of reception guided by schemata, but the concrete reading of really available texts, cf. Angehrn (1999:218f): History ‘is the subject of writing and reading: It materializes in a writing that constitutes the subject by speaking about it; and it is appropriated in a reading which deciphers and exposits documents and traces. ... History-writing is a reading grounded in a reading and a wanting to understand what is already written: the history already crystallized in texts and documents.’


179

is not only achieved through the ‘storification’ which historians undertake in their texts.13 To foreground the narrativizing constitution of History in such a ‘selective’ perception of the past is, moreover, to cohere with current trends in general narratology. As is well known, the classical structural analysis of texts has been supplemented for several years by questions about their communicative context or the cognitive presuppositions of narrativizing readings of them. The underlying question is: What guides the identification and understanding of systems of signs as narratives? As David Herman puts it, ‘What makes readers and listeners deem stories stories?’ (Herman 1997:1048) At present, answers to these questions are sought above all within cognitive science.14 It postulates that our knowing (here understood in a comprehensive sense to include perceptions and utterances) is structured by schemata, scripts, and frames (cf. ibid.:1047). What we perceive is perceived because our cognitive apparatus checks it against ‘internally stored’ schemata. These schemata process narrativity by means of a whole series of operations at different levels of textual comprehension. David Herman, for instance, focuses not only on the standard sequence of events in stories, or, in his own words, ‘the extent to which the form of the sequence facilitates script-aided recognition of a coherent action structure’ (ibid.:1053). He also focuses on the individual elements of each plot. At this level, it is a matter of schemata that use knowledge of the world to supplement what is signified by the texts read. This contributes to the narrativity of an utterance, when the ‘read signified’ can be united to the recipient’s prior knowledge of standard narrative elements and patterns: ‘Telling narratives is a certain way of reconciling emergent with prior knowledge’ (ibid.:1048). I would like to refrain here from any evaluation of this and related narratological efforts. The cognitive model sketched here crosses disciplinary boundaries: models of cognitive science and of psychology also provide basic theorems for philological narratology. In doing so, it requires philologists to make assessments that exceed their disciplinary competence. We encounter 13

14

Cf. White (1981:2f). By contrast, Monika Fludernik undertakes a narratological turn and understands ‘narrativization’ as ‘located in the dynamic reading process’ (1996: 34, cf. 313). For Fludernik, however, it concerns a meaning-building process, which happens in respect of texts: ‘Narrativization, that is the re-cognition of a text as narrative, characterizes a process of interpretation by means of which texts come to be perceived as narratives.’ Most philologists use cognitivist theoretical tools to frame narratological analyses or models, cf. Fludernik (1996:12), Jahn (1997). By contrast, David Herman employs narratology in the service of cognitive science. (cf. 2002:2). Our article pursues a third alternative: the use of cognitivist-theoretical theorems to refine the theory of history. Eder (2003:285) employs a similar approach for film analysis.

180

Daniel Fulda

something of a theory-architectonic ‘black box’ here and I am by no means sure that it effects philologists alone. In what concerns cognition’s contribution to narrativity, the experts of other disciplines must confine themselves to the construction of theories, since inner-psychic events do not admit of observation.15 At bottom one confronts here the old problem which attends all reception-oriented efforts: the necessity of recourse to ‘more or less plausible ideal constructs’ due to our inability to observe the actual reception (cf. Strasen 2002:206). One implication this has for theory, I submit, is that ‘narrative analysis’ is not to be directed in its entirety, as Herman would have it, to ‘the process by which interpreters reconstruct the storyworlds encoded in narratives’ (Herman 2002:5), but should analyze the structure of narrative artifacts as well as their reception. On the other hand, the advantage for interdisciplinarity of the ‘cognitive turn’ is clear: If narrativity is something attributed by the perceiving subject, then the object of this attribution is of secondary importance, be it artistic texts, films or other sequences of images, historical processes, psychic impressions, or other plot-based patterns.16 Those occupied with the connection between narrativity and history are obliged to consider a wide range of objects across many media, since history today is hardly ‘composed’ or ‘selected’ in verbal texts alone (cf. Sobchack 1996). Moreover, the localization of cognitive schemata in an interchange between experience and expectation does much to facilitate a theoretical grasp of the narrativizing construction of History: ‘Stored in the memory, previous experiences form structured repertoires of expectations about current and emergent experiences’ (cf. Herman 1997:1047). Moulded by experiences, cognitive schemata structure perceptions which then are themselves stored in the treasury of experience. This recursivity or interchange is entirely appropriate given the epistemological status as a pattern for thinking that History achieved 200 years ago: We perceive events as historical on the basis of ‘historical experiences’—which we experience, admittedly, in eminently mediated forms, including the form of readings—and construct further ‘historical experiences’ from this ‘knowledge’ of History. The object and mode of perception here mutually support one another, such that they can only be differentiated with explicitly theoretical effort (whereas the everyday understanding of History assumes that it has actually happened).17 15 16

17

For the perspective of a psychologist, cf. Echterhoff (2002:280). For the application of a cognitivistically-framed narratology of this kind to other media, cf. Eder 2003 (in respect of film), as well as Marie-Laure Ryan’s contribution to this volume. This recursivity gained momentum through the two processes from which modern historical thinking arose: first, through the change, described by Koselleck, in our expectations about changes in time which are caused by lived experiences and the changed perception of the


181

This abundance of conceptual overlap enables me to see, within cognitive narratology, a way to connect once again the theory of history with narratology in general (without suppressing the methodological difficulties that attend cognitive narratology). It is within this theoretical framework that it first becomes possible to describe History narratologically as both a pattern for reception and a product of reception. Cognitive narratology not only describes narrativity within the structure of artifacts but goes beyond this to investigate the way that such artifacts are cognitively processed, with the consequence that entire worldviews emerge within the consciousness of their recipients.18 In doing so, it looks at the epistemic function of narrating history—i.e., the function which Baumgartner had described from the perspective of transcendental philosophy and which invests history with its culture-forming power—in a new way.19 This alliance seems to me to be all the more urgent in respect of contemporary articulations of telling history. Today, many representations of history demand a considerable effort of narrativization on the part of the recipient; as History, these texts or other artifacts are not ‘finished’, and for this reason must be ‘selected’ as History. The causes for the low level of manifest narrativity are the modernist distrust of the traditional concept of History with its ‘Great Stories’ as well as a diversification of media and writing techniques which undermine the supremacy of the self-contained story-text as the prototype of historical representation. These trends are often construed as an affirmation that Postmodernity has dismantled unitary History.20 Against the background of cognitive narratology, however, reference to dismantled plots and the like21 does

18

19

20

21

present that follows from it; second, through the translation of a narrative model upon an historical event, which however was understood as objectively ‘poetic’. For Herman (2002:16) both ‘fictional and nonfictional narratives’ have ‘world-creating power’. There is as yet no application of cognitive narratology to historical discourse, but rather a pronounced exclusion of historiography from the field concerned with the cognitivist understanding of narration. Monika Fludernik accounts for this exclusion with the claim that historiography offers no representation of those experiences whose perception she maintains constitutes the core of narrativization (cf. 1996:328). From the perspective of history-theory, the dichotomization of historical knowledge and experience is surely untenable. A narratology of historical discourse also cannot accept Fludernik’s thesis that plot is not a distinguishing characteristic of narration (cf. 1996:311). For without plot, the very transformation that is the essence of history—i.e., the qualitative difference between the point of departure and endpoint of a process—can neither be explained narratively not even expressed. For an attempt to draw history-theoretical consequences from this, cf. Berkhofer (1995). Of her reading of ‘postmodern’ narratives, Elizabeth Deeds Ermarth concludes: ‘we are asked to give up plot and character, history and individuality, perhaps even ›meaning‹ as we have long conceived it’ (1992:212).

182

Daniel Fulda

not suffice. For narrativization emerges within the very ‘selective’ process ‘which enables readers to re-cognize as narrative those kinds of texts that appear to be non-narrative’ (Fludernik 1996:46). History, which can never be written as a whole anyway, appears then as an ‘internal story’—that is, it does not appear in a manifest, recorded form but rather in a virtual one.22 From this perspective, we can ask more specifically what achievements in narrativization are to be assumed when discontinuous representations are still understood as representations of History. 3. The Wehrmacht Exhibition. A Case Study After the Reunification, the single greatest event within German historical culture was without doubt the so-called ‘Wehrmacht Exhibition’.23 Developed by the private Institute for Social Research in Hamburg, it was displayed in 33 German and Austrian cities from 1995 to 1999. It attracted not only 800,000 visitors, but also stimulated deep and lasting discussions, in the public and the private spheres, and among both politicians and historians. Hollywood productions from Holocaust to Schindler’s List had no comparable impact, even if they attracted larger numbers of viewers. Its representation of history managed to draw everyone, and not merely those professionally engaged with the topic, into a debate that was conducted in the “Letters to the Editor” in newspapers as well as the German parliament (March 12, 1997).24 Applied to my question about the narrativizing construction of history, these external data already show the high degree to which history is a product of reception. Above all, the visitors themselves made visible, in working their way from document to document, and either standing, squatting, or bending as required by the various locations of the texts or pictures on display25—that History must always and as a matter of principle be selected (erlesen), that it can in no 22

23

24

25

Without entertaining history-theoretical questions per se, Manfred Jahn also seeks to extend narratological theory-construction to internal stories and their relationship to manifest external stories The exhibition came to be known as the ‘Wehrmacht Exhibition’, and is called by this title by both its supporters and opponents, although its theme was and is ‘only’ the conduct of the war on the eastern front and in Serbia. — The installation of the exhibition initially planned for New York was later cancelled. For the English-language catalogue prepared for that installation, cf. HISR 1999. The debates of the Federal Parliament (Bundestag) are recorded in Thiele (1997: 170–219). Cf. the pictures of the first exhibition in the catalogue for the second (HIS 2002:696, 714). As has been widely observed, the exhibition made considerable demands on the interpretative faculties of visitors. Cf. the discussions surrounding the installation of the exhibition in Bremen (Thiele 1997:95).


183

easy or simple way merely be read (gelesen). Extensive interviews have clarified the way that the exhibition was perceived.26 Taken alongside journalistic and other public reactions,27 these provide a unique corpus of documents for understanding the reception of the exhibition, which make it possible to establish empirically the theoretical turn sketched above to the narrativizing achievement of the recipient. Despite its immense importance for the contemporary culture of history, until now the Wehrmacht exhibition has not been analyzed narrologically. The strength of the reaction to the exhibition has been attributed to its thesis: to wit, that the war against the Soviet Union and in the Balkans was waged, as the title of the exhibition suggested, as a war of annihilation (Vernichtungskrieg), and that the putatively ‘clean’ Wehrmacht was an integral accomplice in the National Socialist crimes, whose victims numbered in the millions. The exhibition, however, was attacked less on account of this basic historical proposition than it was for the manner in which it was presented.28 Critics quickly targeted, above all, the massive use of historical photographs depicting executions, starving prisoners, abused Jews, and other victims of the German campaign. In 1999, a commission of historians convened by Jan Philipp Reemtsma, the Director of the Institute for Social Research, called for a thorough re-working of the exhibition, pointing, for one thing, to errors in its uses of sources and inattentiveness to detail (cf. Bartov 2000:86), and for another, to a general ‘domination by photographs’ (ibid.:88). According to the commission, this domination was problematic on three levels. As a visualization, the photos ‘dramatized’ the reconstructed crimes by means of an aesthetic designed for impact (ibid.:86), and in so doing caused ‘a powerful emotionalization of the audience’ (ibid.:88). Because the textual parts of the exhibition themselves did not elucidate contexts but instead had ‘a provocative character’ and were ‘doctrinally pointed’, the message of the images was not pinned down or supported discursively (ibid.). Images generally cannot capture the massive scale of the crime thematized in the exhibition, but only particular instances of it.29 26 27

28

29

Cf. the volume containing the articles by Naumann, Leh and Heer (all 1998). Cf. Prantl (1997), Thiele (1997), HIS (1999). With the exception of journalistic reporting, only one understandably partisan account exists in English, written from the perspective of the exhibition’s embattled lead historian, Hannes Heer, cf. Heer (1998a). According to the main thread of the discussion in Bremen, cf. Thiele (1997:69, 74, 123 et passim). When serious critics criticized the exhibition’s factual claims, it was on the grounds of their generality, cf. the criticism of Nazism expert Hans-Ulrich Thamer, that the exhibition ‘disobeyed the commandment to differentiate’ (ibid.:75). ‘Such disharmonies between what the text and the pictures say’, concludes the Historians Commission, ‘lead to a loss of credibility, in light of the great probative value which the photos themselves assume in the minds of the exhibition’s authors as well as its visitors’ (Bartov 2000:88). Picture-based representations must be framed by the stories that they contextualize if they are to mean anything historically.

184

Daniel Fulda

The much-discussed ‘impact of images’ (Wucht der Bilder),30 however, is not the decisive point. In my view, an analysis of the exhibition’s narrative structure promises a better explanation as to why a substantial part of its audience reacted dismissively towards it. The fact that it examines crimes committed by Germans may hardly have sufficed to provoke such dismissal, considering that the acknowledgement of German guilt is an established part, indeed a pillar, of contemporary historical culture. Moreover, when the new Wehrmacht Exhibition opened in November 2001—defending the same basic thesis, but having completly revised the techniques it used to display it—it aroused no resistance (save by certain far-right extremists). Accordingly, the form of presentation may have determined whether the events represented were or are accepted as parts of our History. Any form of presenting History, however, requires a narrative core. The exhibition begins with the formulation of its thesis.31 It not only points to the criminal quality of the German prosecution of war on the eastern front, but also refers to opponents of its thesis, to the regnant legend of the ‘clean Wehrmacht’ as well as to ‘German military history writing’ (HIS 1996:7). As a result of this confrontational juxtaposition, the exhibition cannot simply represent historical facts, but, in the words of Hannes Heer, has to ‘offer evidence’ (ibid). Heer and the other organizers of the exhibition stage a trial. They aim, in their own words, to overturn the ‘acquittal’ that the Wehrmacht gave itself after the war.32 The narrative mode of discourse without doubt also has a place within a court proceeding. Indeed, it is the narrative representation of a contested event itself that first makes the verdict possible. Therefore the narration before the court is directed entirely toward the judgment to be 30 31

32

On this widely-used formula, cf. Thiele (1997:166). The catalogue basically reproduces the same text and roughly half of the pictures displayed in the exhibition, albeit not with the same dimensions and in a different graphical sequence (cf. Hannes Heer in Thiele 1997:166, combined with Jureit 2004:11, 13). Indeed Heer later insists ‘the exhibition wishes to pass no belated and general judgment on an entire generation of former soldiers’ (HIS 1996:7). The juridical diction here and the thesis, formulated a few sentences earlier, ‘that the Wehrmacht participated in all crimes actively and as a single corporation’ makes Heer’s assurances seem rather hollow. There is no shortage of generalizations in the exhibition’s companion book: ‘One could not command sadism and a thirst for murder, the absence of human feeling and sexual perversion. Many of the troops brought them with them when they entered’ (Heer 1997:64). ‘At this time, the quality of human being in the Wehrmacht was no different in mentality than those in Himmler’s forces [i.e., the S.S.]’ (Heer/Naumann 1997:30). English edition: Heer/Naumann 2000.—The Institute for Social Research in Hamburg also now looks askance at the judicial efforts of the first exhibition as well as at the generality of the ‘global accusation’ it makes (cf. Jureit 2004:8f: ‘Here the exhibition employed a line of argument that did nothing more than to plead a case;’ it ‘used and presented its sources in a decidedly selective way’).


185

passed; it has its purpose not in itself, but rather in the conviction sought by the prosecution or the acquittal sought by the defense. Ancient rhetoric already recognized that narration in court proceeds, for these reasons, in a one-sided and selective manner; that it discards what does not suit its argumentative aim (cf. Arist. Rhet. 1416b, Cic. De oratore II, 330). Of course, any narrative is selective, yet it makes an enormous difference for the criteria governing its selection whether it involves an instrumentalized or an autonomous narrative (this distinction is of course an ideal-typical one).33 In a non-instrumentalized narrative, what is selected for inclusion is determined by its internal consistency, and not by reference to some preordained goal.34 By contrast, internal consistency, to which belongs the logical consistency or narrative coherence of the story that is recounted, is of less importance for instrumental narratives. Even in legal discourse, the emotions that it aims to produce are more important for having the intended influence over the person addressed (cf. Cicero: De oratore II, 176–216). The first Wehrmacht Exhibition and its accompanying catalogue provide a textbook example for this type of narration. That it does so is not a result of the fact that it concentrates on case studies, since every historian selects and shortens (not only with respect to the events, but also to the intended scope of his statement). The striking thing, by contrast, is the isolation of many individual crimes and horrors without explanation of their context, whether it be the context of military events, or the interdependence of actions by the Wehrmacht and actions by the Red Army or the partisans; or the murderous ‘division of labor’ between the Wehrmacht and other forces. When contexts are not indicated, no differentiations can be made, between the Wehrmacht and the SS or SD, but also in the Wehrmacht between zealous anti-Semites, sadists, accomplices, and those who, when possible, avoided murder (all of these deficits are criticized by the Commission of Historians; cf. Bartov 2000:81f). The first Wehrmacht Exhibition did not utilize the possibilities that narrative alone offers, unlike any other form of discourse, for differentiating as well as for associating. 33

34

Marie-Laure Ryan uses the distinction autonomous/illustrative. Cf. her contribution to this volume. Cicero already says as much about history-writing, as distinguished from judicial and political oratory (De oratore II, 63). For orators, the virtus of brevity (brevitas) requires that only that be told which serves the established end (cf. Knape 2003:102). By contrast, for historians nothing true (and relevant to the subject) may be omitted (De oratore II, 62). Relevance is of course a matter of perspective and is always judged from the present standpoint of those narrating. The methodology of the modern discipline of history recognizes this and chooses the coherence of the represented events as the criterion for historiographical selection: ‘the level of concentration must be determined by the relevance of individual facts to the connection that is being elaborated, in accordance with the requirements of the theme’ (Bernheim 1903:724).

186

Daniel Fulda

My thesis is then that the controversy surrounding the exhibition owes essentially to the fact that the structure of its discourse impedes the narrativization, which the viewer must undertake, in order to perceive the presented events as History. It was clear, even to the most unwilling viewer, that the exhibition concerned events in History. The association that Herman postulates with previous knowledge applies here; in this case, the previous knowledge in question is a historical one. The structure of its presentation and so of the events it represented was not narrative, but judicative, and for this reason hindered History-constituting narrativization on the part of the viewer. Narrativization was certainly not absolutely impossible. With considerable mental and historiographical investment, the viewer could make up the absent narrative contextualization for himself. First, however, this arrangement made a shared understanding easier for some—i. e., those who already had found the Wehrmacht somewhat guilty a priori.35 Those with reservations about the exhibition’s thesis—whether because, as an old soldier, the thesis indicted him or because he found it too unqualified—could easily say that it had not worked through history so much as stood in judgment of it. (This interpretation of the exhibition was further nourished when its organizers tried to stem criticism of it through legal action.36) My narratological analysis by no means denies that personal prejudices and political commitments played a role in the acceptance and rejection of the exhibition. The exhibition itself, however, offered them a point of departure in its decontextualizing and singularly instrumental narrative structure. The first Wehrmacht exhibition was ‘selective’ in the sense that its choice of stories was governed entirely by its argumentative goal and thus without concern for the contexts needed for historical understanding, but not in the sense that it proposed to the viewer through narrativization a ‘selective’ constitution of History. The controversial photos would merit mention here again since their deficiency in contextualization is part of that deficit in narrativity. This is its most striking part, but not the systematic core of the problem.

35

36

Among the visitors interviewed, one may find such reactions, for example, from the NVA officer portrayed by Klaus Naumann (cf. 1998:32–36) or from the SPD member discussed by Hannes Heer, who knew everything about the exhibition even before he visited it (cf. 1998b:97f). Both were, incidentally, reluctant to tell their stories. HIS (2002:711). The discursive paradigm of ‘passing judgment’ also governed the end of the first Wehrmacht exhibition: as scholarly criticism mounted, the Director of the Institute, Jan Philipp Reemtsma, convened a panel of historians which itself passed judgement on the exhibition in ordering a thorough revision of it (cf. ibid.:711–729, esp. 727f). Such verdicts are unusual within historical discourse; more often, one reasons from the premise that history goes on forever the conclusion that the study of history itself never ends.


187

One criticism occurs again and again in the interviews of visitors to the exhibition: namely, that the exhibition suppressed evidence for ‘the other side’: e. g., the ‘normal’ conduct of the war, the vast number of soldiers involved, the learning curve completed after the war, etc. (cf. Leh 1998:48–50, 73f et passim; Naumann 1998:23). At the level of personal experience, we see here a wholeness being reclaimed which is a distinctive feature of the modern, narratively-generated concept of History. What skeptical visitors of the exhibition were missing, even if they lacked a clear concept of it, was some measure of narrativity. This entirely becomes clear from the fact that they responded to it by telling autobiographical stories. Such a pattern of reaction was not restricted to the old soldiers alone among the interviewees. Many participants in public discussion and even the majority of speakers in the German parliament reacted according to it. They began, that is to say, to narrate, if not their own experiences, those of their fathers or other relatives.37 Narrative, as schema or complex of schemata, was at work here; it became palpable for us (or any rate its effects were palpable); narrative is the scheme which a recipient brings with him or her to organize historical experience. To be sure, many responded defensively, and in this regard, expressed themselves as the exhibition’s organizers had, in a judicative discourse employing flatly instrumental narratives (cf. Naumann 1998:37–39, Leh 1998:48f, 51–57 et passim, Heer 1998:82). But the tendency to narrate in these cases nonetheless attests to a narrative understanding of history, since the defensive stance responded to the discursive demand of the exhibition. As was clarified above on a theoretical level, reception-based narrative schemata lead, for one thing, to the recognition of perceived past events as parts of History. For another, they make possible retrospective couplings of new historical experiences with already extant ones. Moreover, they make possible the mediation of different levels of experience (in this case, between general history of the Second World War and personal or family experiences).38 37

38

Otto Schily (cf. Thiele 1997:181f), Freimut Duve (ibid.:190) and Christa Nickels (ibid.: 191–193), Alfred Dregger (ibid.:194) and Theo Waigel (ibid.:206f) reacted in this way to the debate in the German parliament. The ‘narrative-less’ construction of the exhibition, by contrast, makes no provision for stories brought to it by its visitors. To be sure, many of the 131 interviewees were asked to tell stories from their own lives. However, these stories were not recognized as something that a scholarly representation of history was warranted to pass along. For Hannes Heer, the stories of the old soldiers are ‘lies’ (1998b:111); in his view, scholars are not to incorporate the soldiers’s stories but rather to eradicate them. Such a policy, of course, judges and condemns not only suppressions, misrepresentations, and lies about war crimes but also entire ‘life stories’ (ibid.:93). Perhaps the difference in assessments of the first Wehrmacht exhibition derives from different views as to whether historians can and should play the role of judgmental therapists.

188

Daniel Fulda

The new Wehrmacht exhibition provides a cross-check case for my thesis. Its representational technique is also not narrative; it does not offer insight into a continuous sequence of events. Its deep structure is however narratively based, since it contextualizes at every turn.39 In each case, it situates the various crimes—which here too are represented exemplarily—within the larger course of the war. It reconstructs the underlying political and military decisions and gauges what amount of latitude both the criminals and those few heroes who consciously avoided crimes had for acting.40 The second exhibition’s representation of history, like its predecessor, aims to adduce evidence on behalf of something (namely, the overall and recurrent criminal quality of the German war against the Soviet Union), but it no longer does so for the tribunal of the present, but instead shows which norms for the prosecution of war, which were generally accepted at the time, were violated, and violated deliberately. In this sense, the exhibition and its catalogue primarily illumine the norms for the prosecution of war and specify which of the orders by Hitler and Wehrmacht leadership gave rise to the individual crimes on display. In this way, it gives the history its own value back. Put differently, the treatment of the theme is no longer directed toward some intended effect on the present, but instead aims at the widest possible ‘contextualization’.41 The historicizing principle is clearest in the last two stations in the exhibition, which sketch the way that West Germany and East Germany treated German war crimes and tell the ‘history’ of the first exhibition.42 The blanket and primarily moral accusation concerning the ‘great lie’ about the ‘clean Wehrmacht’ (HIS 1996:7) is reformulated here by a narration with considerably more nuance. To the extent that the exhibition documents its own history, it declines to assume the external 39

40 41

42

This is to distinguished from a superficial conception of narration whereby narration is the ‘telling of [self-contained] stories’. Alexander Pollak’s comparison of the two Wehrmacht exhibitions proceeds from such a conception. To the new Wehrmacht exhibition, Pollak attributes ‘an essentially fragmented structure without the strong narrative element underlying it that the first exhibition possessed.’ He does so because he overlooks the deep structure of the new exhibition, which is narrative and contextualizing as the first exhibition was not (2002:58). When he speaks, with regard to the first exhibition, of the ‘largely coherent narrative elements within the particular context of the temporally and spatially confined example’, Pollak describes well the purely instrumental use of narration, without however recognizing it as such (ibid.). Cf. the chapter ‘Handlungsspielräume’ (HIS 2002:579–628). Such is the stated purpose of the exhibition’s new designer, cf. Jureit (2004:9–11). The expansion of its themes serves to contextualize it further still: Whereas the first exhibition concentrated largely on the complicity of the Wehrmacht in the muder of Jews, the exhibition now examines the struggle against the resistance, the treatment of Soviet prisoners, and the starvation of the civilian population. Cf. the chapters ‘Nachkriegskriegszeit’ and ‘Kontroversen über eine Ausstellung’ in the catalogue (HIS 2002:637–730).


189

standpoint of the judicial authority and locates itself within the historical context. It is this reflexivity, that renders the narrativization complete. Every act of storytelling involves, to greater and lesser degrees, both historicizing and interest in the present, both contextualization and exemplarity.43 The problematic aspect of the first Wehrmacht exhibition, from the perspective of history theory, was that it crossed a boundary within this field of tensions. The exclusively instrumental use that it made of narrative elements, for the sake of promoting its preestablished argumentative goal, pushed its viewers backwards, to the time before the genesis of the modern concept of History and of modern historical writing in the late 18th century—a time when history metamorphosized from a repository of exempla to an entity composed of narration that is as autonomous as it is endless.44 This brings us back to the dependence discussed above of History upon narrative techniques. In closing, one possible misunderstanding may be precluded. Just as the past is not necessarily received as History, narrative contextualization does not figure as the only legitimate mode of encountering it. For narrative contextualization can easily lead to the mitigation of unwelcome details (the interviews mentioned above offer abundant evidence for this).45 Against this background, the Wehrmacht exhibition’s refusal of the usual narratives can emerge as a helpful even necessary dismantling—as an act of resistance against narrative domestication. So considered, even the photographs, which were criticized for their erratic impact and for being heavy on shock value, acquire a genuine meaning, since they resist historical explanation, which would always be narrative.46 If this is correct, an additional conclusion for the 43

44

45

46

Cf. Stierle (1973:360): ‘The narrative elaboration of every story is marked by its own specific imbalance. The imbalance derives from the individually different pragmatic context to which the story belongs. Whereas the paradigmatic conception of history comes closest to the domain of systematic texts in the exemplum, History on the syntagmatic level seeks to minimize its ›external determination‹.’ The following sentence clarifies what Stierle means by ‘external determination’ (Fremdbestimmtheit): ‘As long as history, as exemplum, is related to a moral-philosophical system, the narrative scheme appears ›determined externally‹ to a certain extent’ (ibid.). Cf. ibid.: ‘By dissolving its relation to the moral-philosophical system, History moves from the paradigmatic level to the syntagmatic level of an infinite connection ...’ Cf. Heer (1998:83, 86f, 99–100). Heer analyzes the storytelling techniques of the suppression here (84) and describes autobiographical storytelling as the discovery of bearable memories (91f). Beyond the Wehrmacht exhibition, the narrative behaviors of the witnesses to the era (and their offspring), are examined in empirical studies, allied to a certain theory of memory, by Welzer/Moller/Tschuggnall (2002). Cf. Lethen (2002:81f). Jörn Rüsen provides a theoretical discussion of how the Nazi crimes resist incorporation by historians into the sequence of history. He argues in support of integrating the metahistorical break of the genocide into the continuity of History, that

190

Daniel Fulda

theory of history may be drawn from the Wehrmacht exhibition: While narrativism explains the origin and perpetual reproduction of History, not everything that happens in it can be covered by narration. References Angehrn 1999 Angehrn, Emil: “Vom Lesen und Schreiben der Geschichte. Dekonstruktion und historischer Sinn.” In: Selbstorganisation 10, 1999:217–236. Ankersmit 1994a Ankersmit, F[ranklin] R.: History and Tropology. The Rise and Fall of Metaphor. Berkeley / Los Angeles / London 1994. Ankersmit 1994b Ankersmit, F. R.: “Kantian Narrativism and Beyond.” In: Bal, Mieke / Boer, Inge E. (eds.): The Point of Theory. Practices of Cultural Analysis. New York 1994:155–160, 193–197. Bartov 2000 Bartov, Omer et al.: Bericht der Kommission zur Überprüfung der Ausstellung “Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944”. 2000, see www.his-online.de / veranst / ausstell / bericht_kommission.pdf. Baumgartner 1972 Baumgartner, Hans Michael: Kontinuität und Geschichte. Zur Kritik und Metakritik der historischen Vernunft. Frankfurt am Main 1972. Baumgartner 1976 Baumgartner, Hans Michael: “Thesen zur Grundlegung einer transzendentalen Historik.” In: Baumgartner, Hans Michael / Rüsen, Jörn (eds.): Seminar: Geschichte und Theorie. Frankfurt am Main 1976:274–302. Berkhofer 1995 Berkhofer, Robert F., Jr.: Beyond the Great Story. History as Text and Discourse. Cambridge, Mass. / London 1995. Bernheim 1903 Bernheim, Ernst: Lehrbuch der Historischen Methode und der Geschichtsphilosophie. Mit Nachweis der wichtigsten Quellen und Hilfsmittel zum Studium der Geschichte. 3rd and 4th ed., Leipzig 1903. Brooks 1984 Brooks, Peter: Reading for the Plot. Design and Intention in Narrative. Oxford 1984. Chartier 1997 Chartier, Roger: On the Edge of the Cliff. History, Language, and Practices. Transl. by Lydia G. Cochrane. Baltimore / London 1997.

is, at once preserving it and overcoming it there (2001:254–258). By contrast, Frank Ankersmit pleads for a more fundamental revision of narrativist theory of history, on the grounds that ‘narrative integration has always been the historian’s main instrument for appropriating and domesticating the past’ (1994b:197).


191

Danto 1965 Danto, Arthur C.: Analytical Philosophy of History. Cambridge 1965. Droysen 1977 Droysen, Johann Gustav: Historik. Rekonstruktion der ersten vollständigen Fassung der Vorlesungen (1857), Grundriß der Historik in der ersten handschriftl. (1857 / 58) und in der letzten gedruckten Fassung (1882). Ed. by Peter Leyh. Stuttgart-Bad Cannstatt 1977. Echterhoff 2002 Echterhoff, Gerald: “Geschichten in der Psychologie: Die Erforschung narrativ geleiteter Informationsverarbeitung.” In: Nünning, Vera / Nünning, Ansgar (eds.): Erzähltheorie transgenerisch, intermedial, interdisziplinär. Trier 2002:265–290. Eder 2003 Eder, Jens: “Narratology and Cognitive Theories of Reception.” In: Kindt, Tom / Müller, Hans-Harald (eds.): What Is Narratology? Questions and Answers Regarding the Status of a Theory. Berlin / New York 2003: 277–301. Ermarth 1992 Ermarth, Elizabeth Deeds: Sequel to History. Postmodernism and the Crisis of Historical Time. Princeton 1992. Fludernik 1996 Fludernik, Monika: Towards a ‘Natural’ Narratology. London 1996. Fulda 1996 Fulda, Daniel: Wissenschaft aus Kunst. Die Entstehung der modernen deutschen Geschichtsschreibung 1760–1860. Berlin / New York 1996. Harth 1996 Harth, Dietrich: “Geschichtsschreibung.” In: Ueding, Gert (ed.): Historisches Wörterbuch der Rhetorik. Vol. 3, Tübingen 1996:832–70. Heer 1997 Heer, Hannes: “Killing Fields. Die Wehrmacht und der Holocaust.” In: Heer, Hannes / Naumann, Klaus (eds.): Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944. 8th ed., Frankfurt am Main 1997:57–77. Heer 1998a Heer, Hannes: “The Difficulty of Ending a War. Reactions to the Exhibition ‘War of Extermination: Crimes of the Wehrmacht 1941–1944’.” In: History Workshop Journal 46, 1998:187–203. Heer 1998b Heer, Hannes: “Landschaft mit Kratern.” In: Hamburger Institut für Sozialforschung (ed.): Besucher einer Ausstellung. Die Ausstellung “Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944” in Interview und Gespräch. Hamburg 1998:75–115. Heer / Naumann 1997 Heer, Hannes, Klaus Naumann: “Einleitung.” In: Heer, Hannes / Naumann, Klaus (eds.): Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944. 8th ed., Frankfurt am Main 1997:25–36. Heer / Naumann 2000 Heer, Hannes / Naumann, Klaus (eds.): War of Extermination. The German Military in World War II 1941–1944. New York / Oxford 2000. [Englisch edition of Heer / Naumann 1997]

192

Daniel Fulda

Herman 1997 Herman, David: “Scripts, Sequences, and Stories: Elements of a Postclassical Narratology.” In: PMLA 112.5, 1997:1046–1059. Herman 2002 Herman, David: Story Logic. Problems and Possibilities of Narrative. Lincoln / London 2002. HIS 1996 Hamburger Institut für Sozialforschung (ed.): Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944. Ausstellungskatalog. Hamburg 1996. HIS 1999 Hamburger Institut für Sozialforschung (ed.): Eine Ausstellung und ihre Folgen. Zur Rezeption der Ausstellung “Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944”. Hamburg 1999. HIS 2002 Hamburger Institut für Sozialforschung (ed.): Verbrechen der Wehrmacht. Dimensionen des Vernichtungskrieges 1941–1944. Ausstellungskatalog. Hamburg 2002. HISR 1999 Hamburg Institute for Social Research (ed.): The German Army and Genocide. Crimes Against War Prisoners, Jews, and Other Civilians in the East, 1939–1944. Transl. by Scott Abbott. New York 1999. [Englisch edition of HIS 1996] Jahn 1997 Jahn, Manfred: “Frames, Preferences, and the Reading of Third-Person Narratives: Towards a Cognitive Narratology.” In: Poetics Today 18.4, 1997:441–468. Jahn 2003 Jahn, Manfred: “‘Awake! Open your eyes!’ The Cognitive Logic of External and Internal Stories.” In: Herman, David (ed.): Narrative Theory and Cognitive Sciences. Stanford 2003:195–213. Jureit 2004 Jureit, Ulrike: “‘Zeigen heißt verschweigen’. Die Ausstellungen über die Verbrechen der Wehrmacht.” In: Mittelweg 36 13.1, 2004:3–27. Keßler 1982 Keßler, Eckhard: “Das rhetorische Modell der Historiographie.” In: Koselleck, Reinhart / Lutz, Heinrich / Rüsen, Jörn (eds.): Formen der Geschichtsschreibung. München 1982:37–85. Knape 2003 Knape, J[oachim]: “Narratio.” In: Ueding, Gert (ed.): Historisches Wörterbuch der Rhetorik. Vol. 6, Tübingen 2003:98–106. Koselleck 1975 Koselleck, Reinhart: “Geschichte.” In: Brunner, Otto / Conze, Werner / Koselleck, Reinhart (eds.): Geschichtliche Grundbegriffe. Historisches Lexikon zur politisch-sozialen Sprache in Deutschland. Vol. 2, Stuttgart 1975: 593–718. Koselleck 1985 Koselleck, Reinhart: Futures Past. On the Semantics of Historical Time. Transl. by Keith Tribe. Cambrigde, Mass. 1985. Lämmert 1982 Lämmert, Eberhard (ed.): Erzählforschung. Ein Symposion. Stuttgart 1982.


193

Leh 1998 Leh, Almut: “‘Die andere Hälfte der Wahrheit’. Was Zeitzeugen in der ‘Wehrmachtsausstellung’ vermissen.” In: Hamburger Institut für Sozialforschung (ed.): Besucher einer Ausstellung. Die Ausstellung “Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944” in Interview und Gespräch. Hamburg 1998:48–75. Lethen 2002 Lethen, Helmut: “Der Text der Historiografie und der Wunsch nach einer physikalischen Spur. Das Problem der Fotografien in den beiden Wehrmachtsausstellungen.” In: zeitgeschichte 29, 2002:76–86. Mann 1979 Mann, Golo: “Plädoyer für die historische Erzählung.” In: Kocka, Jürgen / Nipperdey, Thomas (eds.): Theorie und Erzählung in der Geschichte. München 1979:40–56. Martinez / Scheffel 2002 Martinez, Matias / Scheffel, Michael: Einführung in die Erzähltheorie. 3. Aufl. München 2002. Megill 1995 Megill, Alan: “‘Grand Narrative’ and the Discipline of History.” In: Ankersmit, Frank / Kellner, Hans (eds.): A New Philosophy of History. London 1995:151–173, 263–271. Mink 1978 Mink, Louis O.: “Narrative Form as Cognitive Instrument.” In: Canary, Robert H. / Kozicki, Henry (eds.): The Writing of History. Literary Form and Historical Understanding. Madison / London 1978:129–149. Mommsen 1905 Mommsen, Theodor: “Rede bei Antritt des Rektorats. 15. Oktober 1874.” In: Mommsen, Theodor: Reden und Aufsätze. 2., unv. Aufl. Berlin 1905: 3–16. Naumann 1998 Naumann, Klaus: “Kameraden oder Komplizen? Der Zwiespalt ganz normaler Berufssoldaten.” In: Hamburger Institut für Sozialforschung (ed.): Besucher einer Ausstellung. Die Ausstellung “Vernichtungskrieg. Verbrechen der Wehrmacht 1941 bis 1944” in Interview und Gespräch. Hamburg 1998:21–47. Pollak 2002 Pollak, Alexander: “Die Historisierung eines Tabubruchs. Von der umstrittenen Entmythologisierung des Bilds der ‘sauberen Wehrmacht’ zur versachlichten Dokumentation des Vernichtungskrieges: ein Vergleich der beiden Wehrmachtsausstellungen.” In: zeitgeschichte 29.2, 2002:56–63. Prantl 1997 Prantl, Heribert (ed.): Wehrmachtsverbrechen. Eine deutsche Kontroverse. Hamburg 1997. Ricœur, Paul 1984 / 88 Ricœur, Paul: Time and Narrative. Transl. by Kathleen McLaughlin and David Pellauer. Vol. 1–3, Chicago 1984–1988. Rigney, Ann 1991 Rigney, Ann: “Narrativity and Historical Representation.” In: Poetics Today 13.3, 1991:591–605.

194

Daniel Fulda

Roberts 2001 Roberts, Geoffrey (ed.): The History and Narrative Reader. London / New York 2001. Rüsen 1982 Rüsen, Jörn: “Geschichtsschreibung als Theorieproblem der Geschichtswissenschaft. Skizze zum historischen Hintergrund der gegenwärtigen Diskussion.” In: Koselleck, Reinhart / Lutz, Heinrich / Rüsen, Jörn (eds.): Formen der Geschichtsschreibung. München 1982:14–35. Rüsen 1993 Rüsen, Jörn: Studies in Metahistory. Ed. and introd. by Pieter Duvenage. Pretoria 1993. Rüsen 2001 Rüsen, Jörn: “Die Historisierung des Nationalsozialismus.” In: Rüsen, Jörn: Zerbrechende Zeit. Über den Sinn der Geschichte. Köln / Weimar / Wien 2001: 217–262. Sobchack 1996 Sobchack, Vivian (ed.): The Persistence of History. Cinema, Television, and the Modern Event. New York / London 1996. Stierle 1973 Stierle, Karlheinz: “Geschichte als Exemplum – Exemplum als Geschichte. Zur Pragmatik und Poetik narrativer Texte.” In: Koselleck, Reinhart / Stempel, Wolf-Dieter (eds.): Geschichte – Ereignis und Erzählung. München 1973: 347–376. Strasen 2002 Strasen, Sven: “Wie Erzählungen bedeuten: Pragmatische Narratologie.” In: Nünning, Vera / Nünning, Ansgar (eds.): Neue Ansätze in der Erzähltheorie. Trier 2002:185–218. Thiele 1997 Thiele, Hans-Günther (ed.): Die Wehrmachtsausstellung. Dokumentation einer Kontroverse. Dokumentation der Fachtagung in Bremen am 26. Februar 1997 und der Bundestagsdebatte am 13. März und 24. April 1997. Bremen 1997. Welzer / Moller / Tschuggnall 2002 Welzer, Harald / Moller, Sabine / Tschuggnall, Karoline: “Opa war kein Nazi”. Nationalsozialismus und Holocaust im Familiengedächtnis. In cooperation with Olaf Jensen and Torsten Koch. Frankfurt am Main 2002. White 1973 White, Hayden: Metahistory. The Historical Imagination in NineteenthCentury Europe. Baltimore / London 1973. White 1981 White, Hayden: “The Value of Narrativity in the Representation of Reality.” In: Mitchell, W. J. T. (ed.): On Narrative. Chicago 1981:1–23. White 1984 White, Hayden: “The Question of Narrative in Contemporary Historical Theory.” In: History and Theory 23, 1984:1–33. White 1999 White, Hayden: “Literary Theory and Historical Writing.” In: White, Hayden: Figural Realism. Studies in the Mimesis Effect. Baltimore / London 1999:1–26, 176–182.

The Narratological Fabric of the Gospels Ute E. Eisen (Gießen / Frankfurt) Probably no one will dispute that the gospels are stories. That justifies, in fact necessitates the questions: What are the characteristics of early Christian story (histoire), narrating (narration) and narrative (récit) in the gospels? In what follows these questions will be considered in two steps: 1. What are the gospels? What were the methodological questions that predominantly shaped analysis of the gospels in the 20th century? and 2. What constitutes narratology’s urgent contribution to analysis of the gospels? What do the gospels tell, and how do they tell it? My thesis is that narratology’s set of instruments is outstandingly well suited to unpack the structural moments, characteristics, and specifics of early Christian gospel narratives. This in turn demonstrates the urgent need for the application of these instruments to gospel study. 1. What Are the Gospels? The problem we face in analyzing the gospels is the fact that the term ‘gospel’ on the one hand describes a certain content, namely the Good News of Jesus as the Christ, but on the other hand it represents a literary genre, that of the four canonical gospels. This essay focuses on the question of the narratological constitutiva of the genre ‘gospel’. As far as scholarship is concerned, the analysis of the genre ‘gospel’ was strongly shaped by the methods of genre-, source-, tradition-, and redaction criticism, as will be briefly summarized below. As regards form or genre criticism: The genre ‘gospel’ was regarded for nearly a century as the sole original genre that Christianity had contributed to ancient literature.1 In recent times this consensus has been shaken; in particular, the close relationship of the genre ‘gospel’ to the genre ‘biography’ has been convincingly demonstrated. A new consensus is emerging. The ancient biography was a new genre derived, at the latest from the 1

Cf. Dormeyer (1993:4 passim).

196

Ute E. Eisen

first century BCE onwards, from those of historiography and encomium and had already ceased development in the third century CE. It can be shown that the gospels are marked by essential factors constitutive of the genre of ancient biography2: – The centering of the narrative on a principal person who is presented as a solidly established character with a predetermined fate. – A three-part structure: prehistory, public activity, and death of the protagonist.3 – And numerous recurring topoi, some of which I will address in the course of my remarks. The essential Christian writings in the genre ‘biography’ are the four canonical gospels (according to Matthew, Mark, Luke, and John). The belonging of the gospels to the ancient genre of biography yields a frame of reference within which the specific narrative form of the gospels and their literary significance within ancient narrative literature can be determined. Concerning source criticism: The debate (especially since the Enlightenment) over the question of the literary dependence of the gospels reached a certain conclusion in the form of the so-called Two Source Theory, which has enjoyed a broad consensus. According to this theory, the Gospel according to Mark is the initial text of the biblical gospels, insofar as it is the oldest gospel and was available to the Gospels of Matthew and Luke as a written source. In addition, these two gospels used a second source, which has been lost and can only be secondarily reconstructed: The Sayings Source Q, probably in written form, and oral, so-called special traditions that scholarship calls Special Material. That is one primary reason why the Gospels of Matthew and Luke are about one-third longer than the Gospel according to Mark. Gospel according to Mark

Saying Source Q

Gospel according to Matthew

Gospel according to Luke

Figure 1: The Two-Source-Theory

2

3

For this and what follows see especially Dormeyer (1999); Frickenschmidt (1997); Burridge (1992). From the point of view of narratology it is relevant that the biographies are not always told in this ‘order’ of events (prehistory, public activity, and death); cf. examples from Frickenschmidt (1997:351).

Narratological Fabric of the Gospels

197

The Gospel according to Luke also belongs to a two-part work, together with the Acts of the Apostles, which in a whole series of episodes describes the spread of the message of Jesus as Messiah, through his disciples (especially Peter and Paul), as far as Rome. The Gospels of Mark, Matthew and Luke are called ‘synoptic’, because they all follow the framework of the Gospel according to Mark and thus can easily be read in parallel and compared. The source-critical location of the Gospel according to John was far more complicated. Here, for a long time, the dominant opinion was that the fourth Gospel was unacquainted with the synoptic gospels, and therefore could not have made use of them as sources. Affinities in individual traditions were explained as the result of oral transmission. This consensus has been increasingly called into question, but a preliminary conclusion to the discussion is not yet in sight.4 As regards the history of oral tradition or tradition criticism: In addition, the question of the traditions redacted in the gospels and their (sub)genres was and is of central importance for gospel analysis. Among the liveliest questions here are: Which traditions can be traced to the historical Jesus? and: What are the religious-historical contexts to which individual early Christian traditions can be traced? and finally: How were these modified in the course of their transmission? Concerning redaction criticism: Beginning with the individual traditions, analysis culminates in the question of the compositional principles of the gospels, or the redactional reworking of the material by the authors of the gospels. This question is, for the most part, considered under the aspects of history and theology. On the basis of synoptic comparisons, word statistics, and tradition-critical hypotheses, scholars have been attempting to demonstrate the specific, fundamental historical and theological tendencies in the individual gospels and in their reworking of their materials and traditions. In gospels research, narrative-analytical categories have already entered the picture but have not yet become standard in the gospel interpretation,5 as an analysis of voice(s) in the individual synoptic gospels, questions of mood (showing, telling, focalization), the distinction between story time and narrative time, the sequential ordering, duration or rhythm, and frequency of the narrative. In addition, at the level of story, to date scarcely any narrative-analytical categories in gospel exegesis—such as the question of events and their hierarchizing (e.g., meta-events, à la Lotman),6 or the 4 5

6

This is the reason why I leave the Gospel according to John aside in this paper. Cf., for example, Merenlahti (2002). They have scarcely gained any foothold to date in German-language gospel research; for the state of scholarship cf. Eisen (2005). Cf. Eisen (2003:155–166).

198

Ute E. Eisen

characterization of the figures in the narrated world—have been applied. In this paper I will essay some aspects of an exemplary narratological analysis of the gospels.7 2. Aspects of the Narratological Fabric of the Gospels In a first step I will now describe aspects of narrating—narrating in the sense of Genettes narration—of the Gospel according to Luke, which together with the Acts of the Apostles constitutes a two-part literary work, and in what follows is called the Corpus Lucanum (2.1). In a second step I will present observations on the story (histoire) of the gospels and the narrative (récit) (2.2). 2.1 The Narrating of the Corpus Lucanum Both biblical books, the Gospel according to Luke and the Acts of the Apostles, have been written by the same author who has been called Luke in the tradition.8 The Lucan two-part work or Corpus Lucanum not only constitutes the most extensive story about Jesus and his disciples; in comparison with the Gospels of Mark and Matthew, it also has the greatest degree of complexity in its narrating. The Gospel according to Luke begins with a prologue (Luke 1:1–4), which in terms of narrative analysis can be called a narrator’s commentary that, here, exercises a mediating or connective function.9 The explicit narrator, who nevertheless remains anonymous, comments his narrative in ‘I’ form. He speaks as extradiegetic narrator to his extradiegetic narratee explicitly and by name, Theophilus: Seeing that many others have undertaken to draw up accounts of the events that have taken place among us, exactly as these were handed down to us by those who from the outset where eyewitnesses und ministers of the word; I in my turn, after carefully going over the whole story from the beginning, have decided to write an ordered account for you, most excellent Theophilus, so that you may learn how reliable the word is that you have received (Luke 1:1–4).

7

8

9

For a more comprehensive narratological analysis of the gospels and the Acts of the Apostles see Eisen (2005). The question of authorship must be kept open because the original manuscipts do not give indications for the author. The attribution to Luke, the physician and companion of Paul, occured only in the second century and is based on a complex reconstruction of biblical passages; cf. Schnelle (2002:284–288). In what follows I am using the scheme of narrative functions developed by Nünning (1996).


199

This prologue of the Gospel according to Luke has metanarrative functions. The narrative process itself is the major subject from a number of points of view. There is its relation to the ‘many’ who have already undertaken to write of the events (Luke 1:1). The narrator thus adopts an explicit relationship to the praxis of production in Christian groups and distinguishes his own undertaking from them in positive fashion. As his sources he names the proclamation of those who from the beginning have been eyewitnesses and servants of the word (Luke 1:2). From what follows, we learn how the undertaking of the narrator of the two-part work differs from that of the ‘many’—namely, the extent of his story and the method of his narrative. The narrator emphasizes that he has pursued everything carefully from the beginning (Luke 1:3). The combination of his ‘after carefully going over the whole story’ and ‘from the beginning’, indicates that the narrator, in contrast to his predecessors, has really researched all that have happened, that means the prehistory of Jesus’ activity as well as its post-history, after Jesus’ ascension into heaven (both is missing in the Gospel according to Mark, in the Gospel according to Matthew the ascension is not narrated). That also indicates the extent of his undertaking, which in fact surpasses not only the Gospel according to Mark but also the Gospel according to Matthew in the number of events narrated. In any case, the narrator begins in very detailed fashion with the circumstances of Jesus Christ’s conception and birth (Luke 1–2), and ends in a second book, the Acts of the Apostles, with the spread of the message, through Paul, as far as Rome (Acts 28). So, the narrator of the Corpus Lucanum goes far beyond the events narrated in the Gospel according to Mark and Matthew. In addition, the narrator emphasizes that he has written everything ‘in ordered account’ (Luke 1:3). I want to note here that the narrator, despite these announcements, also makes use of the narrative methods of pro- and analepsis. Overall, the spectrum of possibilities for shaping the order, the duration and pace, and the frequency within the narrative is exhausted. The message to the narratee of the narrative, Theophilus, is clear: The narrator is in every respect superior to those who have previously attempted this. To summarize: The narrator introduces his narrative, – – – –

first, it is more comprehensive second, it is told from the very beginning third, it is more precise and fourth, it is told in order.

In this way the narrative voice orders its narrative within a larger literary context, which it evaluates and in contrast to which it wishes to show its own method as superior.

200

Ute E. Eisen

In the prologue to the second book, the Acts of the Apostles, the narrator again speaks explicitly to his narratee, Theophilus: In my first book, Theophilus, I dealt with everything Jesus had done and taught from the beginning until the day he gave his instructions to the apostles he had chosen through the Holy Spirit, and was taken up to heaven (Acts 1:1–2).

Both prologues fulfill an appellative function, in that they speak to a narratee by name (Luke 1:3; Acts 1:1). The purpose of the undertaking, namely in relation to the explicit narratee, is also named in this context. The narrator addresses him directly in the second person singular and challenges him to recognize the ‘reliability of the word’ (Luke 1:4). Thus the expected attitude of receptivity is named. In this way the readers are drawn into the narrative process, with corresponding instructions for their behavior. The renewed address to his narratee at the beginning of the second book has here moreover a phatic function: the appellation serves to stabilize the channel of communication between narrator and narratee. At the same time, the program and authority of the narrator of the first book are recalled. The story of the first book is summarily told in a brief analepsis. Simultaneously, with the renewed address to the narratee, a close tie is established between the first and the second book. A further narratologically interesting phenomenon arises in the second book: The extradiegetic-heterodiegetic narrator of the Corpus Lucanum becomes in the Acts of the Apostles suddenly a homodiegetic narrator in the so-called ‘we’ passages (Acts 16:10–17; 20:5–25; 21:1–18; 27:1; 28: 16). Here he leaps, quite without introduction, into the action and, as a companion of Paul, becomes an eyewitness to the events by himself. The primarily extradiegetic-heterodiegetic narrative voice suddenly and without warning becomes homodiegetic, that is, a figure in the story. The extent of these passages, in comparison to the whole length of the Acts of the Apostles or, indeed, of the two-part work, is small; there are only four such brief narrative passages. Moreover, they are spread over thirteen chapters. However, the narrator achieves some major narrative effects by this. Firstly, the narrator shows that he took part in the events and in so far was an eyewitness. But whether this is authentic or not—a question that can never be answered with absolute certainty—this pretended eyewitness has the effect of giving the narrator additional authority. In the prologue to the gospel the narrator primarily formulated the authority of an historian (histor) for himself, and has demonstrated it in his narrative. Now, though still subordinated, the authority of an eyewitness is added to this as the narrative nears its end. This hierarchy corresponds to ancient convention, which gave great value on both authorities but greater value to the report


201

of the histor than to that of the eyewitnesses.10 Our narrator, however, claims both authorities for himself.11 It remains striking that the narrator in the ‘we’ passages is no more identifiable as a figure than in the prologue and in the other parts of the narrative. The Narrator appears in this ‘we’ as an anonymous companion, his primary characteristic being that he accompanies Paul on some of his journeys. That is the prime function of this ‘we’: it replaces the narrator in the diegesis. As a figure, he experiences the events as an eyewitness and thus lays claim to the authority of one who has seen something personally. This procedure was very successful. Even today this narrative is received in this way, namely as an authentic eyewitness account. With Genette, we can also call this narrative procedure ‘metalepsis’.12 2.2 Observations on the Story (histoire) and the Narrative (récit) of the Synoptic Gospels The basic story of the Gospel according to Mark is expanded by the Gospels of Matthew and of Luke. I want to show more clearly how this is done at the beginning and end of the narrative. Both the Gospels of Matthew and of Luke tell events around Jesus’ birth and infancy, something lacking in the Gospel according to Mark. The Gospel according to Mark opens with the sentence: ‘The beginning of the good news of Jesus Christ, the Son of God’ (Mark 1:1). However, that is the main information the Gospel according to Mark gives about the origins of its protagonist. It is different with the Gospels of Matthew and of Luke: Each of them is preceded by a fairly extensive story of birth and infancy. In these infancy narratives we encounter various literary topoi from ancient biographies,13 such as genealogies, and in the case of very prominent people also the circumstances of their conception and birth, for example in the case of Alexander the Great and Augustus.14 Further topoi are giftedness of the twelve-year-old protagonist, temptation stories, 10 11

12 13

14

Cf. Scholes/Kellogg (1966:242–248). And let it be said parenthetically that it was regarded as honorable among Hellenistic historians if an historian like Odysseus had borne ‘the spray and swell of the waves’ (Homer, Odyssee XII:219), still more: it should be an inescapable duty for every historian truly deserving of the name. Against this background it is not surprising that the ‘we’ narrator in the Acts of the Apostles is used primarily in narratives of sea voyages. See Cornils (2004). Frickenschmidt (1997:210–350) offers an exhaustive list of the topoi of ancient biographies, with abundant textual examples. Plutarch, Alexander 2f; Suetonius, Augustus 94; in both, a divine begetting is narrated. This aspect then begins to be positively rampant in the biographical novels (see Ps.Kallisthenes, Historia Alexandri Magni 12, or: Philostratos, Vita Apollonii I,4–5; cf. further Frickenschmidt 1997:243–244; 253–255).

202

Ute E. Eisen

character-summaries, etc. Many of these topoi appear in the birth and infancy narratives in the gospels as well as in ancient biographies. Strikingly, the birth and infancy narratives in the Gospels of Matthew and Luke are told differently: for one thing, in the selection of the events narrated and also in the way the same event is narrated. For example the announcement of Jesus’ conception and his birth: The Conception of Jesus is told in the Gospel according to Luke from the point of view of Mary (Luke 1:26–38) and in the Gospel according to Matthew from the point of view of Joseph (Matthew 1:18–25). At all in the Gospel according to Luke it is narrated in dramatic mood: that is in two scenes with psychonarration and direct speech and many details.15 In the Gospel according to Matthew it is not that dramatically narrated. There are more commentaries of the narrator and the birth of Jesus is only mentioned in single clause in the narrator’s account: ‘Until she had borne a son’ (Matthew 1:25).16 The Gospel according to Luke in contrast vividly relates the well-known account of the decree that went forth from Caesar Augustus and Jesus’ birth in a stable (Luke 2:1–7).17 15

16

17

‘In the sixth month the angel Gabriel was sent by God to a town in Galilee called Nazareth, to a virgin betrothed to a man named Joseph, of the House of David; and the virgin’s name was Mary. He went in and said to her, ›Rejoice, so highly favored! The Lord is with you.‹ She was deeply disturbed by these words and asked herself what this greeting could mean, but the angel said to her, ›Mary, do not be afraid; you have won God’s favor. Listen! You are to conceive and bear a son, and you must name him Jesus. He will be great and will be called Son of the Most High. The Lord God will give him the throne of his ancestor David; he will rule over the House of Jacob for ever and his reign will have no end.‹ Mary said to the angel, ›But how can this come about, since I am a virgin?‹ ›The Holy Spirit will come upon you‹, the angel answered, ›and the power of the Most High will cover you with its shadow. And so the child will be holy and will be called Son of God. Know this too: your kinswomen Elizabeth has, in her old age, herself conceived a son, and she whom people called barren is now in her sixth month, for nothing is impossible for God.‹ ›I am the handmaid of the Lord‹, said Mary, ›let what you have said be done to me.‹ And the angel left her’ (Luke 1:26–38). ‘This is how Jesus came to be born. His mother Mary was betrothed to Joseph; but before they came to live together she was found to be with child through the Holy Spirit. Her husband Joseph, being a man of honor and wanting to spare his publicity, decided to divorce her informally. He had made up his mind to do this when the angel of the Lord appeared to him in a dream and said, ›Joseph son of David, do not be afraid to take Mary home as your wife, because she has conceived what is in her by the Holy Spirit. She will give birth to a son and you must name him Jesus, because he is the one who is to save his people from their sins.‹ Now all this took place to fulfill the words spoken by the Lord through the prophets: The virgin will conceive and give birth to a son and they will call him Immanuel, a name which means ›God-is-with-us.‹ When Joseph woke up he did what the angel of the Lord had told him to do: he took his wife to his home and, though he had not had intercourse with her, she gave birth to a son; and he named him Jesus’ (Matthew 1:18–25). ‘Now at this time Caesar Augustus issued a decree for a census of the whole world to be taken. This census—the first—took place while Quirinius was governor of Syria, and


203

The Gospel according to Matthew contains the dramatic narratives of the coming of the Magi from the East, the flight into Egypt, and the slaughter of innocents by King Herod (Matthew 2), all of which are lacking in the Gospel according to Luke. From a narrative-theoretical perspective, then, it is no wonder that in the Christmas story read every year in church, at just this point narratives from the Gospels of Luke and Matthew are combined and read as one. With Jesus’ public activity in Galilee (Matthew 4:12; Luke 4:14), after his baptism (Matthew 3:13–17; Luke 3:21–22) and the temptation story (Matthew 4:1–11; Luke 4:1–13), the central section of the gospels begins. Regarding this central section I will only note that it is striking how differently the Gospels of Matthew and Luke treat their primary sources. While Matthew works newer material more generally into the whole, the Gospel according to Luke contains a major insertion within the framework of the Gospel according to Mark, in the form of a travel narrative (Luke 9:51–18:14). Travel narratives, incidentally, are also a topos in ancient biographies.18 Jesus’ activity is narrated in all the gospels as a journey through Galilee that ultimately leads to Jerusalem, and so to his death. It is especially in the central section of the gospels that the so-called passion predictions play an important role. Since more than once in the gospels the future suffering, death, and resurrection of the Son of Man is proleptically narrated by the lips of the reliable protagonist Jesus (Mark 8:31–33; 9:31; 10:32–34; Matthew 16:21–23; 17:12, 22–23; 20:17–19; 26:2; Luke 9:22, 43b–45; 17:25; 18:31–34). The high frequency overall places great weight on the passion narrative. Advance indications of the future fate of the hero are also a topos of ancient biographies. For example Plutarch’s advance notice of the senseless death of Pelopidas and Marcellus (Plutarch, Pelopidas 2), or of Aristides, ‘that he, who had at first been beloved, would later, because of his second name, become the object of distaste’ (Plutarch, Aristides 7). In contrast to the narrator’s speech in Plutarch, the so-called passion predictions in the versions in the Gospels of Mark, Matthew, and Luke are in the form of direct speech on the lips of the Son of God. They thus acquire greater weight; they are both certain about the future (Eberhard Lämmert) and at the same time

18

everyone went to his own town to be registered. So Joseph set out from the town of Nazareth in Galilee and traveled up to Judaea, to the town of David called Bethlehem, since he was of David’s House and line, in order to be registered together with Mary, his betrothed, who was with child. While they were there the time came for her to have her child, and she gave birth to a son, her first-born. She wrapped him in swaddling clothes, and laid him in a manger because there was no room for them at the inn’ (Luke 2:1–7). See Frickenschmidt (1997:272f).

204

Ute E. Eisen

dramatic in their mood of narration. They contain the central key words of the events to come. In Mark we read: The Son of Man must undergo great suffering, and be rejected by the elders, the chief priests, and the scribes, and be killed, and after three days rise again (Mark 8:31–33).

These prolepses are repeated in the Gospel according to Mark a second and third time (Mark 9:31; 10:32–34). In the third prolepsis, what is to come is narrated more fully than before: Now we are going up to Jerusalem, and the Son of Man is about to be handed over to the chief priests and the scribes. They will condemn him to death and will hand him over to the pagans, who will mock him and spit at him and scourge him and put him to death; and after three days he will rise again (Mark 10:32–34).

It is not only said that the Son of Man will be handed over to the high priests and scribes and condemned to death, but also details of his being arrested and the event after his death. In the final part of the whole narrative, in the so called passion narrative—part three in the ancient biography—these announcements are then scenically depicted, for example in the scene of the mocking of Jesus by the soldiers, who put a purple cloak on him and weave a crown of thorns and put it on him, and spit on him (Mark 15:16–20; Matthew 27:27–31; John 19:2–3). The passion narrative begins with the great crisis in Jerusalem, introduced by the oath of the high priests and the scribes and their decision that Jesus must be put to death (Mark 14:1–2; Matthew 26:1–5; Luke 22: 1–2). Oath is a topos which is also significant in ancient biographies.19 And the gospels have in common that in the passion narrative, and thus in the last section of the biography, they throttle back the pace of the narrative. This they also have in common with ancient biographies. While in the central section Jesus’ activity, lasting perhaps a year, his words and deeds, are told swiftly, and enriched with summaries. Different in the passion narratives: The few days in Jerusalem before Jesus’ execution are narrated in great detail and scene for scene (Mark 14–16; Matthew 21–28; Luke 19:29–24:53). But in the passion narrative also, the Gospels of Matthew and Luke differ in the dimension of the events they narrate, compared to each other and to Mark. Thus in the Gospel according to Mark the story of encounters with the Risen One is found solely in an external prolepsis that is not given 19

See Frickenschmidt (1997:317–320).


205

further narrative elaboration (Mark 16:7). Let me say parenthetically that the Gospel according to Mark with stories of appearances by the Risen One, as it appears in most Bible translations, is source-critical secondary, added by late manuscripts (Mark 16:9–20). The original Gospel according to Mark ends with the scene in which the women disciples find Jesus’ tomb empty (Mark 16:1–8). The women encounter ‘a young man in a white robe’—an angel. There follows an external prolepsis, in direct speech by the narrative figure of the angel: Do not be alarmed; you are looking for Jesus of Nazareth, who was crucified. He has been raised; he is not here. Look, there is the place they laid him. But go, tell his disciples and Peter that he is going ahead of you to Galilee; there you will see him, just as he told you (Mark 16:6f).

The angel’s speech is certain about the future, since in the narrative world of the Bible angels, as messengers of God, are absolutely reliable figures who make definitive prophecies. At the same time, a proleptical saying of Jesus (Mark 14:28) is resumed, namely, that Jesus is going before them to Galilee. This prolepsis, certain about the future, is followed by the closing narrative discourse: So they went out and fled from the tomb, for terror and amazement had seized them; and they said nothing to anyone, for they were afraid (Mark 16:6–8).

The ending of the gospel is thus apparently open, were it not for the words of Jesus and the angel, which demand faith on the part of the implicit narratee. A speech in the mouth of the reliable figure of an angel can at the same time be interpreted as an instruction for reading: Go back to the beginning in Galilee; there the story will start again from the beginning. This open end in the Gospel according to Mark is no longer found in the later written Gospels of Matthew and Luke. Here encounters with the Risen One are narrated scenicly. The resurrection itself is, significantly, not narrated in any gospel as a scene; only its effects are told: the finding of the empty tomb and namely in such a way that Jesus appears to his disciples and thus shows himself as one who has risen from the dead. These are, for example, the scenes everyone is familiar with, such as the one with the two disciples on the road from Jerusalem to Emmaus, who meet a traveler whom they only later—after his disappearance—identify as the risen Jesus (Luke 24:13–35), or the famous scene with the unbelieving Thomas, who must put his hands into Jesus’ wounds in order to believe (John 20:24–29). The Corpus Lucanum also goes beyond a scenic narrative of the appearances of the Risen One by reporting a further event. It tells of Jesus’ ascension into heaven; in fact, it does so twice, and in very different

206

Ute E. Eisen

versions, once at the end of the gospel (Luke 24:50–53) and again at the beginning of the second book, the Acts of the Apostles (Acts 1:9–11).20 Narratives of ascension to heaven were also a topos found in ancient biographies—only, however, in the case of very important historical figures, such as Caesar and Augustus (Sueton, Augustus 100; Caesar 88), Moses (Philo, Vita Mosis II.291), or Romulus, the founder of the city of Rome (Plutarch, Romulus 28).21 This also shows the kind of world-historical context in which the Gospel according to Luke places its protagonist. The Corpus Lucanum, as I have said, tells the story of Jesus’ ascension twice. The two stories diverge—something that has caused scholars repeatedly to raise the question of the original tradition. But a narrative-critical analysis yields a relatively simple explanation of the differences between the two narratives. The essential differences are clarified by the fact that the ascension story in the Gospel according to Luke has the primary function of leading the readers or hearers out of the story, while in the Acts of the Apostles it serves to lead them into it. The ascension story at the end of the Gospel according to Luke is silent: Then he took them out as far as the outskirts of Bethany, and lifting up his hands he blessed them. Now as he blessed them, he withdrew from them and was carried up to heaven (Luke 24:50–51).

The narrator reports Jesus’ gesture and blessing. In this way the ascension story in the Gospel according to Luke creates a distance between the figure and the readers or hearers and thus achieves an effect of closure as it ends the narrative. Readers and hearers can see the disciples, but they cannot hear them. The narrative is related with external focalization. Jesus’ gesture and his disappearing are seen from a distance. The event is viewed from without. In this way the readers and hearers are aided in leaving the narrated world. The lifting of the hands and the report of Jesus’ blessing are here an indication of dismissal of the readers and hearers, also, from the story, in the sense of a farewell blessing. In the Acts of the Apostles, by contrast, there is no blessing by Jesus: As he said this he was lifted up while they looked on, and a cloud took him from their sight. They were still staring into the sky when suddenly two men in white were standing near them and they said, ‘Why are you men from Galilee standing here looking into the sky? Jesus who has been taken up from you into heaven, this same Jesus will come back in the same way as you have seen him go there’ (Acts 1:9–11).

20 21

Cf. Parsons (1987). See in detail Frickenschmidt (1997:342–345); Parsons (1987:136 passim).


207

New here are a cloud carrying Jesus out of their sight, the fact of the disciples’ no longer seeing Jesus (Acts 1:9), their looking up to heaven, the appearance of two angels (Acts 1:10), and their announcement of Christ’s parousia (Acts 1:11). In addition, the two narratives differ in their narrative mood. In the Acts of the Apostles it is narrated in dramatic mood, with an internal focalization from the perspective of the disciples. In this way the narratee can follow the event through the eyes of the disciples and is thus brought into the event itself. The internal focalization here has the function, mentioned above, of opening the narrative by drawing the readers and hearers into the story. Here again, as in the annunciation to Mary and the discovery of the empty tomb, the narrator of the gospel according to Luke places the message on the lips of angels, as reliable figures equipped with high authority. With the words ‘that Jesus, as he has been taken up from you into heaven, this same Jesus will come back in the same way’, the ascension is narrated toward the future. In this way the opening of the narrative points proleptically to the end of the story.22 3. Conclusion These are only spotlights on a narratological analysis of the gospels. Let me make three final points. As regards story: A comparison of the gospels shows that we can observe a progressive tendency within the synoptic gospels to develop a broader and sometimes more detailed scenic style.23 In classic form criticism there was repeated discussion of the question: What was at the beginning of the tradition, the small form or the extended narrative? One of the most influential exegetes of the twentieth century, Rudolf Bultmann, proposed the thesis that the Markan story of the empty tomb in Mark 16 was a secondary narrative development of the credal formula: ‘Christ died for our sins in accordance with the scriptures, [and that] he was buried, [and 22

23

The prolepsis of the parousia of Christ, the Day of the Last Judgement (Act 1:11; 3: 20f; 17:31), is external, because it is not told at the end of the story of the Acts of the Apostles. In the Corpus Lucanum as a whole we can observe a tendency to tell the story of Jesus from creation to the day of the Last Judgment by means of analepses and prolepses. For example, the genealogy follows Jesus roots to Adam and God (Luke 3: 23–38), or Stephen’s speech tells the story of Israel from Abraham to the present (Acts 7:2–53). Thus it is not without reason that the ascension story is added to the Corpus Lucanum, or that in the second century apocryphal gospels will give a detailed account of Jesus’ descent into hell (Gospel of Nikodemus 17–27; Gospel of Bartholomäus 1:28–35). There is a tendency for the number of narrated events to increase; cf. Klauck (2002:126–128; 133 passim).

208

Ute E. Eisen

that] he was raised on the third day in accordance with the scriptures’ (1 Corinthians 15:3–4) that is, that narrative development is secondary to the original statement of faith.24 After what I have developed above, this proposal seems to me more persuasive than before. The Apostle Paul handed over the cited credal formula. He wrote one generation earlier than the Gospel according to Mark and he does not mention an empty tomb. Paul continues the credal formula: ‘[and that] he appeared to Cephas, then to the twelve. Then he appeared to more than five hundred brothers and sisters at one time … Then he appeared to James, then to all the apostles. Last of all, as to one untimely born, he appeared also to me’ (1 Corinthians 15:5–8). This addresses the question of what was at the beginning of the narrative: it was the experiences of Christians, which were narratively reworked, elaborated, and then were told and retold and expanded. Evaluation, with regard to the narrating: Within the gospel narratives, that in the Corpus Lucanum is more complex than expected. Elsewhere as well, in the gospels the possibilities for narration on various levels, with different degrees of involvement in the narrated events, are fully exploited. The spectrum extends from extradiegetic-heterodiegetic to intradiegetichomodiegetic narrators, the last of these fully involved in the events. As regards the narrative: At the level of the narrative itself we can observe a plenitude of narrative phenomena that can be grasped by the methodological instruments of modern narratology. At the level of time and mood we encounter very different phenomena, with differing intensity.25 Only a broadly-conceived and detailed comparison of the gospels can quantify these preliminary observations. In addition, there needs to be a narrative-analytical comparison with ancient biographies. Only then can generally valid conventions for ancient narrative be formulated. In conclusion, we may say that narrative analysis of texts enriches theology in that it transcends the often dominant and frequently fruitless question of the historicity of the texts. Narrative analysis underscores that these texts are developing narratives that tell the story of Jesus as 24 25

Bultmann (1921/1995:308–310). Let me also say in advance that the Gospels are written in a simple Greek that cannot compete with high-level Greek. An example of this—relevant for narratological analysis—is the fact that in the New Testament the narrative of words is done primarily through reported speech and in direct speech. Indirect speech is far less frequent than in classical Greek or, most especially, in the Latin authors. In this and other regards the popular or colloquial nature of the language of the Gospels is very evident. Only the Gospel according to Luke and the Acts of the Apostles, that is, the Corpus Lucanum, constitute something of an exception; they are written more in an ‘educated Greek’. But it is not only the Gospels that have received bad grades from scholars for their language. They share that fate with their companions in the genre, namely certain ancient biographers, for example Cornelius Nepos and Suetonius (cf. Frickenschmidt 1997:32–34).


209

the Messiah and his effects on the figures of the narrated world. The best thing about this is that the different gospels contain something for different taste, and there is not one authorized narrative, but four of them: the Gospel according to Mark, of Matthew, of Luke and of John. The Gospel according to Mark is certainly the most ‘rational’ gospel, and is suited to those who do not put much value on an excessive legendary ornamentation. In addition, in my opinion it is high time that the classic methods of gospel exegesis, which have always been oriented to literary scholarship, be expanded to include narratology in the standard analysis. The time has come, for biblical studies26 as well as for the study of antiquity.27 References Bal 1986 Bal, Mieke: Femmes imaginaires: L’ancien testament au risque d’une narratologie critique. Utrecht 1986. Bal 1988a Bal, Mieke: Murder and Difference. Gender, Genre, and Scholarship on Sisera’s Death. Bloomington 1988. Bal 1988b Bal, Mieke: Death and Dissymmetry: The Politics of Coherence in Judges. Chicago 1988. Bultmann 1921 / 1995 Bultmann, Rudolf: Die Geschichte der synoptischen Tradition. 10th edition, Göttingen 1995. Burrigde 1992 Burrigde, Richard A.: What Are the Gospels? A Comparison with the Graeco-Roman Biography. Cambridge 1992. Cornils 2004 Cornils, Anja: “La métalepse narrative dans les Actes des Apôtres: un signe de narration fictionnelle?” In: Pier, John/Schaeffer, Jean-Marie (eds.): Métalepses. Entorses au pacte de représentation. Paris 2004 [forthcoming]. De Jong 1989 De Jong, Irene J. F.: Narrators and Focalizers. The Presentation of the Story in the Iliad. Amsterdam 1989. De Jong 1991 De Jong, Irene J. F.: Narrative in Drama. The Art of the Euripidean Messenger-Speech. Leiden 1991.

26

27

For New Testament Studies cf. Merenlahti (2002); Marguerat/Bourquin (1999); for Old Testament Studies the narratological analysis is already established, cf. for example Utzschneider/Nitsche (2001:150–186); Bal (1986; 1988a; 1988b). For Classics cf. Schmitz (2002:55–75); de Jong (1989; 1991; 1994; 2001).

210

Ute E. Eisen

De Jong / Sullivan (eds.) 1994 De Jong, Irene J. F. / Sullivan, J. P. (eds.): Modern Critical Theory and Classical Literature. Leiden 1994. De Jong 2001 De Jong, Irene J. F.: Narratological Commentary on the Odyssey. Cambridge, UK 2001. Dormeyer 1993 Dormeyer, Detlev: Das Neue Testament im Rahmen der antiken Literaturgeschichte. Eine Einführung. Darmstadt 1993. Dormeyer 1999 Dormeyer, Detlev: Das Markusevangelium als Idealbiographie von Jesus Christus, dem Nazarener. Stuttgart 1999. Eisen 2003 Eisen, Ute E.: “Boundary Transgression and the Extreme Point in Acts 10:1–11:18.” In: Schaberg, Jane / Bach, Alice / Fuchs, Esther (eds.): On the Cutting Edge: The Study of Women in Biblical Worlds. Essays in Honor of Elisabeth Schüssler Fiorenza. New York / London 2003:154–170. Eisen 2005 Eisen, Ute E.: Die Poetik der Apostelgeschichte. Narratologische Studien. Freiburg, CH 2005. Frickenschmidt 1997 Frickenschmidt, Dirk: Evangelium als Biographie. Die vier Evangelien im Rahmen antiker Erzählkunst. Tübingen 1997. Gospel of Bartholomäus “Gospel of Bartholomäus.” In: Schneemelcher, Wilhelm (ed.): Neutestamentliche Apokryphen in deutscher Übersetzung. Bd. I: Die Evangelien. 6th edition, Tübingen 1990:424–440. Gospel of Nikodemus Gospel of Nikodemus. In: Schneemelcher, Wilhelm (ed.): Neutestamentliche Apokryphen in deutscher Übersetzung. Bd. I: Die Evangelien. 6th edition, Tübingen 1990:395–418 Homer (Odyssee) Homer: Odyssee. Griechisch und Deutsch, mit Urtext, Anhang und Registern. Transl. by Anton Weiher, introd. by A. Heubeck. 11th edition, München et al. 2001. Klauck 2002 Klauck, Hans-Josef: Apokryphe Evanglien. Eine Einführung. Stuttgart 2002. Marguerat / Bourquin 1999 Marguerat, Daniel / Bourquin, Yvan: How to Read Bible Stories. An Introduction to Narrative Criticism. Translated by John Bowden. London 1999. Merenlahti 2002 Merenlahti, Petri: Poetics for the Gospels? Rethinking Narrative Criticism. London / New York 2002. Nünning 1997 Nünning, Ansgar: “Die Funktionen von Erzählinstanzen: Analysekategorien und Modelle zur Beschreibung des Erzählerverhaltens.” In: Literatur in Wissenschaft und Unterricht 30, 1997:323–349.


211

Parsons 1987 Parsons, Mikeal C.: The Departure of Jesus in Luke-Acts. The Ascension Narratives in Context. Sheffield 1987. Philo (De vita Mosis) Philo: De vita Mosis (lib I–II). Philonis Alexandrini opera quae supersunt. Ed. by Leopold Cohn and Paul Wendland. Berlin 1902 (Reprint 1962). Philostratus (Vita Apollonii) Philostratus: The Life of Apollonius of Tyana. With an English Translation by F.C. Conybeare, 2 vols., London / Cambridge,MA 1969 Plutarch (Alexander, Aristides, Pelopidas, Romulus) Plutarch: Plutarch’s Lives. With an English Translation by Bernadotte Perrin, 11 vols. London 1914–1926. Ps.-Kallisthenes (Historia Alexandri Magni) Ps.-Kallisthenes: Leben und Taten Alexanders von Makedonien. Der griechische Alexanderroman nach der Handschrift L. Ed. and transl. by Helmut van Thiel. 2nd edition, Darmstadt 1983. Schmitz 2002 Schmitz, Thomas A.: Moderne Literaturtheorie und antike Texte. Eine Einführung. Darmstadt 2002. Schnelle 2002 Schnelle, Udo: Einleitung in das Neue Testament. 4th edition, Göttingen 2002. Scholes / Kellogg 1966 Scholes, Robert / Kellogg, Robert: The Nature of Narrative. London / Oxford / New York 1966. Suetonius (Augustus, Caesar) Suetonius: Die Kaiserviten / De vita Caesarum. Lateinisch-Deutsch. Ed. and transl. by Hans Martinet. 2nd edition, Düsseldorf / Zürich 2000 Utzschneider / Nitsche 2001 Utzschneider, Helmut / Nitsche, Stefan Ark: Arbeitsbuch literaturwissenschaftlicher Bibelauslegung: Eine Methodenlehre zur Exegese des Alten Testaments. Gütersloh 2001.

Narrative Discourse and Identities Michael Bamberg (Worcester, MA) 1. Introduction At the outset of my contribution to this volume, I need to stress that my research interests lie in identity; more precisely, in the identity constructions of children and adolescents. In other words, I am analyzing narratives in order to trace how children in their transformations to young adults bring off claims about themselves that result in something like a sense of self and therefore, something that can be claimed to be relevant to one’s ‘identity’. The approach with which I am working is part of the attempt to explore identity formation processes from the perspective of ‘the natives’; that is, with as little preconceptions as possible, particularly preconceptions that come from a perspective informed by a notion of maturity in terms of what it means to be an adult at a particular socio-historical time-place coordinate.1 I will outline the type of work I do in terms of ‘narrative research’ further below, but first I would like to fend off the potential misunderstanding that my work directly lends itself to a ‘better understanding’ of narrative. In other words, the approach I am embracing in my pursuit of adolescents’ identities and the work I am doing with narratives does not directly contribute to the field of narratology. Having presented this strong disclaimer, I will nevertheless try to take this opportunity to position my own approach in such a way that it can be read in contrast to a trend in current narratological theorizing. This trend, which I—admittedly somewhat polemically—have called the ‘cognition-über-alles’ position, is on the verge of becoming the dominant attempt to lend to narratology a seemingly more scientific habitus. Having 1

This type of ethnographic/ethnomethodological analysis is explicitly anti-deductive. If we already knew the central concerns of the life-world of ‘the natives’, we would go in with clear hypotheses, test them, and not waste our (and the natives’) time with expensive (in the sense of time-consuming) qualitative research. However, it should also be clear that there is no totally ‘presupposition-free inquiry’. Ethnographic research, particularly if ethnomethodologically informed, has to be open and become increasingly reflexive of the ideological positions that are exposed in this type of inquiry.

214

Michael Bamberg

entered the field of cognition myself at the heights of the ‘cognitive revolution’ in the mid-seventies, and having embraced the cognitive approach to language and narrative enthusiastically for more than a decade, I have become increasingly dissatisfied with its straight-jacketing controls when it comes to exploring the lives of ‘real’ people, and the telling of ‘real’ stories in ‘real’ contexts.2 Originally, the turn to cognition, which resulted in the field of cognitive science, had a clear liberating force: The image of the ‘cognizing person’ was not only no longer at the mercy of outside stimuli or forces, but also, the ‘cognizing person’ is viewed as constructing meaningful relationships by taking information in, working it over, and ‘putting it out’ (in verbal and non-verbal actions). More importantly, viewing the person and his / her central organizing apparatus, the mind, simultaneously as the producer and interpreter of meaningful entities, became an approach that opened up a radically new way of doing inquiry into the human faculties as ‘competencies’. However, as I will briefly outline below, this ‘revolution’ came also with considerable costs. The world of practices (formerly conceived of as ‘behavior’), in particular human discourses, becomes an add-on. Talk as the everyday business of interaction in this purview becomes one of the many aspects of what humans can do because they are endowed with competencies and parameters, and these competencies have become more and more central to what seemingly needs to be explained—particularly by developmental research. As a consequence, actual talk-in-interaction as well as narratives-in-interaction become applications and deviations, all the way up to ‘distortions’ of what the ‘actual’ mind is able to accomplish—particularly in experimental or institutionally augmented settings. In order to study what the mind is able to do in such situations that are relatively far removed from the everyday, ‘explanatory approaches’ are called for that show how mind and brain interact in the production of meanings. Therefore, the empirical domain to conduct this type of research can no longer be a description of the everyday, of talk, and of narrativesin-interaction, the way they are negotiated in daily routines. In contrast, explanations are gleaned (by glimpses of the ‘actual’ mind) in idealized, experimental conditions, or even better, in controlled simulations, ‘revolutionizing’ the empirical domain for narrative investigations.

2

This appeal to ‘something real’ is not supposed to contrast with something that is ‘not real’, but rather to view people (in my case young adolescents) involved in everyday interactions, sharing accounts on topics that are relevant to them. It will become clear further below that this orientation is concerned more with ‘small stories’ (Bamberg 2004b), the way they surface in everyday interactions—in contrast to full-blown life stories (elicited in research interviews or therapeutic sessions) or written biographies.

Narrative Discourse and Identities

215

Before I elaborate further on the potential costs of our turn to cognition and cognitive science, let me foreshadow briefly an alternative, one that a number of psychologists and scholars in communication theory and sociology who became increasingly disillusioned with the limitations of cognitive science have been working on for the last two decades. This is the orientation I will outline more fully in the second and third section of this chapter. This approach focuses more strongly on the action orientation of language in ‘communities of practices’. With this orientation, we decidedly analyze what people do when they talk and what they do when they tell stories. Starting with practical talk-in-interaction and narratives as embedded in such talk, presents the attempt to break free from the constraints of the ‘cognition-über-alles’ position with its inherent costs. Thus, the turn to discourse counters the previous turn to cognition; however, it does not claim that cognition is dead or redundant. Nor does it replace the ‘cognition-über-alles’ position with a newly formed ‘discourse-über-alles’ position, but rather, it complements and sets straight the former with an opening to an empirical realm where cognitions emerge out of discourse as well as discourses from cognition. In other words, the approach to narratives as discourse and performance, the way I will elaborate below, does not explain cognition ‘away’, but knocks it off its hegemonic ‘überalles-position’ and puts it from its head onto its feet, where cognition can become a product of discursive, story-telling practices. I should, however, mention that there are attempts that seek to connect and complement the two views that I present here as contrasts and in discordance (narrative as cognition versus narrative as discourse), particularly in the work represented by Herman (2002, 2003a, 2003b). And clearly, as with my own academic biography from cognition to discourse in my work with narratives, one could highlight more strands that point to an underlying coherence. However, in the hope to contribute to more clarity, I have chosen to structure this chapter in terms of two contrasting positions. For this reason, I will first (sections two to four) sketch a few thoughts as to why narratologists at this point in their long-standing history may be attracted by and turn to cognition. Thereafter, in sections five through eight, I will explicate my own work with narratives as an approach that attempts to analyze narratives-in-interaction in order to see what people actually do when they narrate. How it will be possible to integrate this approach into what is the main concern of narratologists, that is, moving closer to a definition of ‘narrative’ (or at least closer to a delineation of what ‘narrative’ can be), may not become clear instantly. However, I hope that this chapter will stimulate discussions toward that end.

216

Michael Bamberg

2. Why Narratologists Might Want to Turn to Cognition To start with, literary studies and narratology are in good neighborhood. The cognitive revolution has swept across the social sciences to the point that even social psychology, the former stronghold of social behavioral research, is in the firm grip of cognition. Nowadays, what is social in psychology is studied by ‘getting inside the head’ (Taylor / Fiske 1981), so that we can experimentally investigate how social phenomena are represented in the individual mind, or, as Greenwood (2004:239) calls it, to explore ‘cognition directed toward other persons and social groups’. The study of emotions, personhood, and even ‘the world’ has been successfully subjected to a cognitive orientation that views the human mind as the central and universal organizer of information—or, in less agentive terms, the place where information about self and the world is centrally organized (Hogan 2003, 2004; Taylor / MacLaury 1995; Wierzbicka 1999). However, what is in it for literary studies and narratology in particular to jump onto this very powerful band-wagon, unless it is simply attempting to reach for the mere proximity to what commonly counts as ‘science’ and ‘scientific’? In my opinion, there are at least two compelling reasons: the first stemming from narratologists’ preoccupations with and strong privileging of the literate over the oral, and the second, from the hope finally to link what traditionally has been divided into more or less two separate centers of concern, the author and the reader. Both are ‘good reasons’, in the sense that they reflect orientations to expand shortcomings of a traditionally more textually oriented narratology. However, as I will try to argue, both are simultaneously coming with great costs; costs that may blind alternative and potentially more productive ways to expand traditional narratology and connect it more closely with the ‘narrative turn’ in the social sciences.

3. Cognition and the Oral-Literate Distinction The oral-literate distinction has been widely discussed, and it is generally assumed that the development of writing systems has had some major impact on our self-construal and the ways we make sense of (social) others in modern times. Writing handles best the developmental organization of bounded categories in the form of events; it creates a beginning, a middle, and an end. In writing, we have become capable of making lists, charting changes, categorizing everyday experiences, developing a new form of memory, and ensuring the transmission of memories between generations (Goody 1977; Goody / Watt 1968; Ong 1982). Olson has pointed out that


217

writing in ontogenesis facilitates the attribution of belief and emotion states to others, both of which are said to be central in children’s construals of mind and intentionality, and these in turn are developmental keys en route to learning how to read and write (Olson 1997a, 1997b). But why? And how? There seems to be no clear agreement on how to answer these questions, nor is it clear whether there is a clear boundary between non-literate and literate epochs within the European development of literate cultures. Similarly, what kinds of oral practices tie children optimally into learning to read and to write is another wide open issue demonstrating the oral-literate continuum (Heath 1983). However, agreement seems to exist on the categorical distinction that written texts contrast sharply from oral speech in terms of their openness and contextual limitations. While oral texts are limited to the immediate situation of the interlocutors, this ‘narrowness of the dialogical situation’ explodes in writing (see Josselson’s and Freeman’s discussions of Ricœur; Freeman 2004; Josselson 2004). While the oral is fleeting, the referential and ideational fixity of writing orients more clearly toward intentions ‘behind’ the text that are to some degree now inscribed or fixated by writing. While meanings are loosely situated in oral dialogue, they can be negotiated and ultimately surface in oral encounters, though in a fleeting sort of existence. In writing, however, they seem to be more overtly and directly accessible. Again, we may wonder: Why? Why is it that the written text seems to be superior and simply a better candidate for the investigation of what ‘really’ seems to be at stake in the construction of meaning and its interpretation? Only rarely has the question been raised as to whether the oral origins of narrating (socio-genetically as well as onto-genetically) have had any consequences on transferences into other medial representations. Wolf (2002:36f) briefly touches on this question, only to dismiss oral storytelling as a special case within prototype theoretical considerations to narrative and to use (written) fairy tales for his demonstration of a ‘narrative prototype’. Fludernik (1996, 2003) more explicitly claims to privilege ‘spontaneous conversational storytelling’ (1996:13); that is, oral versions of non-fictional storytelling, however, only to revert and give privileged status in her analysis to fictional stories.3 An additional question, raised 3

Fludernik (2003) argues that the model she has developed in (1996) ‘takes its inspiration from natural narrative, arguing that natural narrative is the prototype of all narrative’ (248), but it should be clear that her goal is very different from my own. While my interest in narratives is concerned with how people use them, hers is to work up a definition of narrative that can be applied to ‘all types of narrative texts’, including ‘the two least researched areas of narrative texts—pre-eighteenth-century narrative (medieval and early modern) and postmodernist narrative’ (ibid.).

218

Michael Bamberg

recently by Freeman (2004), is whether written transcripts of oral narratives have implications in the sense that predilections stemming from traditional narrativity leak into the analysis. Along these lines it should not come as a surprise that discourse (oral talk) itself is modeled as a text, and its referentiality is declared to be its central ingredient. Discourse is the exchange of referentially denoted information, the way it is represented in the individual mind, encoded by culturally available semiotic means (usually in terms of a linguistic code), and subsequently encoded by the reader / interlocutor. Discourse is ‘cognitive discourse’, exchange between ‘talking heads’. In the worst scenario it is the mere exchange of information; in a somewhat better world, it is the negotiation (between interlocutors) of cognitive models. And in an even better world, it is a negotiation that includes a constant updating of such models (see Herman 2002). How we, as information processors, text producers, interactants, ended up with our mental models in the form of (more or less) ready-made competencies, ready for exchange and updating in performance, is the issue I will pick up on with my alternative proposal below.

4. Cognition as ‘Distributed’ between the Author and the Reader Classical structuralist narrative theory takes the (written) text as given and investigates the structural features of the text (Nünning / Nünning 2002). From here it moves in two possible directions: one is toward the author and tries to answer the question of how the text came into being; the other works from the text toward the reader and attempts to answer the question of how the text is interpreted. Author-oriented approaches typically are interested in aspects of the author’s life, his or her biography or spirit as it is breathing in the text. Psychological, in particular psychoanalytic, interpretations have their place in this orientation. Reader-oriented approaches are relatively young (Iser 1974, 1978). They developed during roughly the same period in which the cognitive turn took its grip in psychology; that is, during the sixties and seventies, paralleled by very similar assumptions. While reception theory was primarily guided by the question of how the reader interacts with the text (and in this sense what the reader brings to the text in terms of expectations), cognitive theorizing in psychology was turning to comprehension issues of a similar but broader range, that is, asking the question of how the human mind picks up patterns and enriches them with schematic information (from expectations and memories) into meaningful units. Developments in artificial intelligence, a sub-domain of cognitive science, promised exciting developments in the simulation


219

of such comprehension processes and resulted in advances such as story grammars and machine translation projects. It is worth noting that these two directions of author-oriented and reader-oriented text studies rarely were able to connect within the field of literary studies and its sub-discipline, narratology. This, so it seemed, could productively change by more fully embracing the cognitive turn and transporting cognitive theorizing more explicitly into literary studies and narratology. The text in cognitive theorizing is less the starting point for pattern-seeking, but rather the connective tissue for and between author and reader—or in broader terms, between speaker and hearer. Concepts borrowed from frame- and prototype-semantics (Fillmore 1975, 1982; Lakoff 1977, 1987; Rosch 1975, 1978; to mention a few) provided the links between mental configurations of representations that are able to supplement the cues given in text and communication with additional, supplementary information. For instance, verbs such as buying, selling, putting up for sale, purchasing, or auctioning, all can be said to trigger aspects of a more holistic scenario (or ‘gestalt’) of the ‘financial transaction scenario’ (Fillmore 1982; Herman 2002:164). Language processors of the form of the human mind (or artificial, though intelligent, systems) automatically fill in the other, unsaid, aspects of the scenario to a fuller understanding of who is involved, including contextual aspects of how the transaction took place. The choice of specific lexical / textual items and devices highlights the particulars of cognitive scripts or scenarios (such as Schank / Abelson’s ‘restaurant script’ 1977) that are taken to be culturally shared and as such contributing substantially to human understanding and sense making. Against this background of cognitive theorizing, it becomes intelligible that the study of narratives as spoken and written texts is always the study of texts as deviations from the prototypes that are assumed to be shared by speakers / writers and audiences / readerships. Actual narrative texts are the imperfect copies or performances of idealized, but ‘psychologically real’, representations of the idealized speaker, writer, hearer, or reader. In this sense, the narrative as a cognitive category, it is argued, is as ‘natural’ as the category ‘birds’ or ‘furniture’, from where we, as contextual, cultural beings, derive—through frequent exposure and ‘experience’—the categories that are central (prototypical) to us, such as ‘robins’ and ‘chairs’ for Northern Americans. What used to be construed as two different orientation points in traditional narrative theorizing has become the central unit of cognitive narrative research. Empirical research has developed a number of different means to approximate our ‘natural category’ (the culturally shared prototype) of storyhood. These means were sophisticated ways to test for story comprehension and story recall (prompted and unprompted),

220

Michael Bamberg

appreciation and goodness-judgments of goals, motives, or emotional tone, as well as comprehension studies of non-literal statements and non-typical stories. And some of us would like to take this as the definition of what ‘story’ means, so we can ‘measure’ deviations from it, and / or see how much of this central category applies to narratives told in everyday conversations and narratives in other modalities, such as film, music, as well as across the different arenas of its application, such as court rooms, medicine, history, psychoanalysis and the like. Let me stress that there is nothing wrong with this type of argument and the type of research that follows up on it. To be clear, research that demonstrates effects that can be interpreted in terms of some form of ‘psychological reality’ of prototype categories is a clear progress vis-à-vis traditional checklist inventories, since it is based on some kind of fuzziness of the assumed category boundaries and open to some form of cultural, contextual processes of formation. However, if the argument is maintained that these categories guide not only decision making processes in experimental conditions, but (all) our activities in everyday categorizations and interactions, this position is elevated into one that places ‘cognition-überalles’, that is, it becomes a predilection with consequences. Ochs / Capps (2001) have listed five practical implications that the hegemony of cognition has had on the privileging of narrative dimensions in the social sciences: (i) with regard to the dimension of ‘tellership’, conventional (cognitive) narrative analysis has privileged ‘one active teller’ in contrast to ‘multiple active co-tellers’; (ii) high ‘tellability’ has been over-explored at the expense of low ‘tellability’; (iii) detached ‘embeddedness’ from surrounding talk and activity has been emphasized over a more contextual and situational ‘embeddedness’; (iv) a more certain and constant ‘moral stance’ has been assumed as the default case in contrast to a more uncertain and fluid one; and (v) with respect to linearity and temporality, the closed temporal and causal order has been privileged over a more open temporal or spatial ordering. These—in my opinion unfortunate—tendencies, though not in any way caused by the turn to cognition in narrative theorizing, nevertheless seem to come in the wake of an otherwise productive inquiry into the cognitive dimensions of narratives.

5. Narratives-in-Interaction as Vehicles to Fashion Identities Again, at the outset of this section, let me reiterate that the purpose of my work and my intellectual involvement with narratives is not to find out or contribute to a better understanding of what narratives are. In addition, and this may come even more as a surprise, my primary interest


221

in narratives is not even in what the narratives that I am analyzing are about. Form and content are of secondary relevance. They only become relevant as far as they assist the analyst in figuring out for what it is they are used. In other words, the analyses of form and content of narratives in identity research are heuristics in the effort to analyze how self and identity come to existence. Of course, this shift in emphasis requires some staking out of the terrain of investigation. Therefore, this section is devoted to clarify the general approach I am embracing. First, I will sketch the discursive approach (within the larger framework of ‘discursive psychology’) that is laying the foundation for my interest in what best is characterized as ‘identity negotiation’—or even better as ‘identity confrontations’, events in which conversationalists encounter interaction-trouble and need to manage and fine-tune their resources in order to come across in alignment with institutional and interpersonal demands.4 In this section, ‘the discursive approach to narrating’, I will argue that narratives are ‘built on-line’; they are fashioned in order to build and work one’s way through challenging circumstances. Then, in the next section entitled ‘positioning analysis’, I will lay out an analytic framework that is able to take care of this type of ongoing relationship work that narratives are said to accomplish. In a third section, I will summarize the analysis of a ‘small story’ (Bamberg 2004b) to orient the intended reader toward the application of this type of approach to narratives-in-interaction. 6. The Discursive Approach to Narrating Grounding narratives in interaction, I follow tenets of a social constructionist approach that applies ideas from ethnography, discourse analysis, and ethnomethodology to psychological issues and concepts (Edwards / Potter 1992; Potter 1996; Potter / Wetherell 1987). This type of approach typically is concerned with identifying the rhetorical and argumentative organization of discourse the way it is used to fashion self- and identity-claims. This, for us as discourse analysts, implies paying close attention to the way speakers’ accounts are rhetorically and argumentatively organized, which we only can do (as analysts) by closely following the interactive subtleties and rhetorical finessing that are part of the daily expression of attitudes, 4

It is assumed that ‘interactional trouble’ is more the norm than the exception; in particular, this is so when it comes to claims about self and identity. The kind of relational maneuvering of claiming a positive social value for oneself that Goffman called ‘facework’ (Goffman 1967), even if participants cooperate in sustaining its enactment, always requires us to place ourselves ‘on the line’. Face can either be lost or saved.

222

Michael Bamberg

evaluations, and assessments. As such, analyzing narratives-in-interaction operates in close proximity to discursive approaches that examine evaluative expressions as parts of interactive, social, and cultural practices, which implies the close scrutiny of how such expressions are put to use, as opposed to speculating about the mental or attitudinal objects that they putatively reflect (Edwards / Potter 1992; Potter 1996; Potter / Wetherell 1987). Rather than seeing attitudes, values, or self-claims as cognitively given, either / or, and slow to move, attitudes are seen as talk’s business, as partial and shifting devices (or ‘topics’) that spring up in a constantly shifting interaction that occasions and makes use of these devices, and then moves on (Antaki in press; Edwards / Potter 1992). As a result, applying this type of discursive approach to narrative analysis in identity research is fully interested in the inconsistencies, contradictions, and ambiguities that arise in interactions. And narratives are taken to be primary territories where co-conversationalists seek and find ways to mitigate the interactive trouble and fashion a portrayal of themselves in ways that are interactively useful. Rather than seeing narratives as intrinsically oriented toward coherence and authenticity, and inconsistencies and equivocations as an analytic nuisance, the latter are exactly what are most interesting. They offer a way into examining how storytellers are bringing off and managing their social identities in contexts (Bamberg 2004a). Seen this way, such instances no longer appear as contradictions or inconsistencies, but rather as openings into which the analyst can delve and see how such multiple attending and rhetorical finessing is used to work up identity claims that do appear as complex, reportable, and authentic, and not too obvious, challengeable, or immature (see Korobov / Bamberg 2004). It is in this sense that participants in interactions constantly adjust their actions to what is created ‘in the moment’. As Sigman puts it: ‘communication is not always or primarily the execution / enactment of prototypes or scripts; certain problematic situations both emerge and are resolved through ongoing communication’ (Sigman 1995:9). It also should be clear at this point that this type of work with discourse neither advocates nor denies the pre-existence of previous knowledge, experience, or personality traits; it simply circumnavigates the necessity of having to explain interactive manoeuvres within a cognition-über-alles approach. Communicative competence of participants in interactions along these lines becomes the ‘competence to monitor the progress of an interaction and fashion … turns to effect remedial steps if it heads in the wrong direction’ (Sanders 1995: 118). And narratives are no exception. In contrast to the ‘biographic method’ of the German sociological tradition (e.g., Fischer-Rosenthal / Rosenthal 1997; Fuchs-Heinritz 2000;


223

Rosenthal / Fischer-Rosenthal 2000) that is interested in analyzing elaborate self-accounts in the form of life stories, I am (more) interested in ‘small stories’, the ones that are told in mundane encounters and everyday circumstances. While biographical life stories are typically elicited by use of particularized interrogation techniques in institutional settings (research or therapy), I am starting from the more general assumption that stories in principle are rhetorical tools for point or claim making, irrespective of whether they are ‘revealing’ personal and private issues about the speaker and irrespective of whether they thematize whole lives or a singular incidental event or happening. In entering the narrative realm the point or claim that is under construction becomes contextualized in the form of exemplary actions by exemplary characters that are appropriated (from a speaker’s point of view) to ‘act out’ and to make currently relevant the claim the speaker intends to convey for the here-and-now of the conversation. This principle holds whether the speaker talks about him- / herself, his / her life, or about others. However, inserting the self of the teller into the story line opens the door to the possibility of an ‘I’ that has been or even still ‘is’ in flux, is open to interpretation, and can be viewed from different angles. The conversational point of presenting different ‘I’s’ at different times and places, subjected to different character constellations, can be highly effective in constructing a particular understanding of ‘me’ as speaker in particular conversational contexts. In other words, the sequence of I-positions in the story-world are the means to bring off a claim with regard to ‘this is the way I want you to understand me’, here and now: The I as a character who has emerged in the story-world is made relevant to the me as the speaker in the here-and-now. This differentiation between the self as character in the story and the self as speaker (animator and / or author) is extremely important, because we all too often tend to collapse them too quickly in our analyses. However, although there is no principled difference between drawing up characters in a story world, in which the self of the speaker figures as character, from drawing up story worlds in which he / she is not, I would like to concede that the former usually has more at stake in terms of anticipating and preventively fending off potential objections by the audience. In sum, narratives, irrespective of whether they deal with one’s life or an episode or event in the life of someone else, always reveal the speaker’s identity. The narrative point-of-view from where the characters are ordered in the story world gives away—and most often is meant to give away—the point-of-view from where the speaker represents him- / herself. By offering and telling a narrative, the speaker lodges a claim for him / herself in terms of who he / she is. In narratives in which the speaker talks about or even

224

Michael Bamberg

thematizes him- / herself, this is neither more nor less the case. However, constructing a self as a character in the story world and entering this construction as a claim for the self of the speaker, requires ‘additional’ rhetorical work in order to be heard ‘correctly’. It is this ‘additional rhetorical work’ that elevates ‘personal narratives’ into the realm of interesting data, and not the fact that speakers are revealing something that counts as more intimate or ‘personal’. It is along these lines that I would like to argue that narratives told in everyday interactions always lodge claims about the speaker’s sense of self, and in their attempts to convince and make these claims intelligible, speakers incorporate counter claims vis-à-vis what they think could constitute possible misunderstandings.

7. Positioning Analysis For the purpose of analytic work with narratives, I had begun to apply in some of my previous work the concept of ‘positioning’ (Bamberg 1997, 2003; Talbot et al. 1996). This concept has gained current relevancy in theorizing identity and subjectivity, where ‘positions’ are typically seen as grounded in master narratives but opening up and conserving some territory for individual agency. Elaborating on Butler’s (1990, 1995) notion of performing identities in acts of ‘self-marking’, I have tried to advance a view of positioning that is more concerned with self-reflection, self-criticism, and agency (all ultimately orientated toward the possibility of selfrevisions). In so doing, I suggest that we clearly distinguish between the ‘being positioned’ orientation, which is attributing a rather deterministic force to master narratives, and a more agentive notion of the subject as ‘positioning itself’, in which the discursive resources or repertoires are not a priori pre-established but rather are interactively accomplished. ‘Being positioned’ and ‘positioning oneself’ are two metaphoric constructs of two very different agent-world relationships: the former with a world-to-agent direction of fit, the latter with an agent-to-world direction of fit. One way to overcome this rift is to argue that both operate concurrently in a kind of dialectic as subjects engage in narratives-in-interaction and make sense of self and others in their stories. In taking this orientation, the ‘who-am-I?’ (identity) question does not presuppose a unitary subject as the ground for its investigation. Rather, the agentive and interactive subject is the ‘point of departure’ for its own empirical instantiation (Butler 1995:446) as a subject that is constantly seeking to legitimate itself, situated in language practices, and juggling several story lines simultaneously. The analysis of how speakers actively and agentively position themselves in talk starts from the assumption that the


225

intelligibility of their claims is situationally and interactively accomplished. However, since this intelligibility is the result of what is being achieved, and therefore inherently oriented to, we begin our actual narrative analysis by paying close attention to the ways in which the represented world of characters and event sequences is drawn up. Here we attempt to spot descriptions and evaluations of the characters and analyze the time and space coordinates in the way that these relate to social categories and their action potential. From there we move into a closer analysis of the way these referential and representational aspects of story construction are assembled in their sequential arrangement among the participants of the conversation. The assumption that governs this step is that particular descriptions and evaluations are chosen for the interactive purpose of fending off and mitigating misinterpretations. The descriptions and evaluations rhetorically function to convey how speakers signal to their audience how they want to be understood. In working from these two levels of positioning (one with respect to the content of what the story supposedly is about, the other with respect to the coordination of the interaction between speaker and audience), we are better situated to make assumptions about the ideological orientation within which the speakers are positioning a sense of self; that is, as signalling complicity in order to mark off segments that can be countered. The analysis of the first two positioning levels is intended to lead progressively to a differentiation of how speakers work up a position as complicit with and / or countering dominant discourses (master narratives). It is at this juncture that we come full circle by showing how subjects position themselves in relation to discourses by which they are positioned. In other words, analyzing talk in interaction along these lines enables us to circumvent the aporia of two opposing subject theories, one in which the subject is determined by existing narratives, the other in which the subject is the ground from which all narratives are invented. Ironically, this way of analyzing talk-in-interaction for the purpose of gaining an understanding of how interactants establish a sense of self (in stories-in-interaction) resembles closely what in developmental theorizing is termed ‘microgenesis’ (see Bamberg 2004a). This approach focalizes the momentary history of human sense-making in the form of emergent processes. It assumes that developmental changes (such as learning or better understanding) emerge as individuals create and accomplish interactive tasks in everyday conversations. The interactive space between the participants, whether situated in interviews or other social locations, is the arena in which identities are micro-genetically performed and consolidated and where they can be micro-analytically accessed. Here I am borrowing from developmental (Bamberg 2000; Catan 1986; Riegel 1975; Werner

226

Michael Bamberg

1957; Werner / Kaplan 1984; Wertsch / Stone 1978), conversation-analytic (Schegloff 1982; Sacks 1995; Sacks / Schegloff / Jefferson 1974), and ‘communities of practice’ approaches (Eckert 1989, 2002; Hanks 1996; Wenger 1998) to analyze the sequential and relational structure of narratives-ininteraction, for the purpose of inquiring not only into the developing sense of self and others, but also into what is shared as the cultural background of sense-making. This does not imply that such ‘senses’ of self, other and generalized other (culture) do not exist previously to or outside the discourse situation. However, for the analysis of narratives-in-interaction, I am suggesting the bracketing out of these categories so that we can be open to the analysis of what the participants make currently relevant in the interactive setting. In entering this orientation from a socio-linguistic and ethnomethodological vantage point, I am proposing considering and analyzing narratives as brought off and carefully managed in the social realm of interaction rather than as texts that come in the form of stories.

8. How to Use Narratives-in-Interaction to Analyze Identities In this section I will elaborate my aforementioned approach to the ‘narrative analysis of identities’ and give an illustration in the form of a brief example. The story I want to analyze is a very short account about a male 11th grade student, who is said to talk a lot about his gayness (near his locker), and who is further characterized as associating more with girls than other boys. This account, which altogether does not entail much of a plot development, stems from a 15-year-old boy and is situated in the context of a group discussion with an adult male moderator and five other male age mates. It will become clear that a good assessment of what the story is about can only be made if we are able to take into account why the story was shared, which requires an investigation into how the story is interactionally grounded, and how it is jointly accomplished by the participants of the interaction. The discussion topic at the start of this excerpt is whether there are any gay boys at their school. James, who in turn 4 had already established to be better informed than Ed about the current status of gay boys at their school, in turn 6 claims to actually know a few gay boys at their school. However, midstream he self-repairs his claim to this kind of knowledgeable authority by downgrading it to ‘just’ ‘having seen’ them. One possible explanation for downplaying the quality of his relationship with gay schoolmates may be to fend off being heard as ‘too close’ to them; that is, as someone who has ‘gay friends’ and possibly even is gay himself. However, he is challenged by Ed and Alex in their subsequent


227

turns (7 and 8), though not for ‘having gay friends’ (or being gay). But instead, Ed and Alex ‘notice’ that James does not have clear criteria for recognizing others as gay—as if James did not know what he was talking about. James, in turn 9, responds by seeking clarification (‘how do I know they’re gay?’). He displays ‘not understanding’ Ed’s and Alex’s challenges, and treats them as if they were groundless. From here the conversation could go into a number of different directions. For instance, a potential dispute could evolve about typical gay characteristics. However, when Ed upholds his challenge (turn 10), James responds with a turn-initial ‘well’ (a general shifter of frames that also signals the intention of holding the floor for an extended turn) and shifts focus from ‘plural gays’ to an unspecified ‘singular he’, supposedly a member of the ‘gay category’. This ‘he’ is further specified as an 11thgrader, and his name is explicitly not mentioned. The rhetorical device of explicitly not mentioning his name is a clever way of displaying sensitivity and discreteness, and thereby indexing the interactive business at hand as not gossiping or any form of ‘bashing’ a particular person. However, at the same time, these very same devices foreshadow and gear up the audience’s expectations toward something that is highly tellable and gossipy. Ed’s and Josh’s demands (in turns 12 and 14, respectively) to hear names bespeak exactly this. However, instead of giving names, James (in turn 13) moves further into descriptive background details; namely that he has class with mostly 11th-graders, and thus—in contrast to the other five boys in the ongoing conversation, who all are 9th-graders—may be more knowledgeable of the boy he has introduced in turn 11 and left unspecified thus far. So, the interactional setting in which the storied account is grounded is the following: James, who seemed to have successfully laid claim to knowing better and more about the gay population at their school toward the beginning of this excerpt, is challenged for not being able to distinguish gays from non-gays. This seems to force James to respond by setting the scene for what orients toward a more elaborate account in the form of a story. He introduces a specific character, presumably a gay 11th-grader, opening up audience expectations for what is to come next as a sequence of descriptions and evaluations (most likely of the character in question) that clarifies why and how he (James) actually is able to make accurate judgments on gay issues. In other words, with his subsequent story James is expected to reclaim the authority on gay issues that had been questioned.

228

Michael Bamberg

Excerpt: How do I know they’re gay? 1 Ed there are some gay boys at Cassidy 2 Moder do they do they suffer in eh at your schools do they are they talked about in a way / / 3 Ed / / I don’t think there are any I don’t think there are any openly gay kids at school 4 James ah yeah there are 5 Ed wait there’s one there’s one I know of 6 James actually I know a few of them I don’t know them but I’ve seen them 7 Ed how can you tell they’re gay 8 Alex yeah you can’t really tell 9 James no like how do I know they’re gay 10 Ed Yeah 11 James well he’s an 11th grade student the kid I know I’m not gonna mention names 12 Ed alright who are they (raising both hands up) 13 James okay um and I’m in a class with mostly 11th graders 14 Josh and his name is (rising intonation) 15 James ah and and ah and um a girl who is umm very honest and nice she has she has a locker right next to him and she said he talked about how he is gay a lot when she’s there not with her like um so that’s how I know and he um associates with um a lot of girls not many boys a lot of the a few of the gay kids at Cassidy The actual story unfolding in turn 15 is not a typical event or plot story, but rather consists of two pieces of further descriptive information. First, a description of the 11th-grader: He is said to talk a lot about his being gay and to hang out at school more with girls than with boys. These pieces of information arguably provide evidence for the alleged person’s membership of the category ‘gay’, and in this sense can be said to relate the point the audience may be waiting for. The second piece of information is more subtle and also more interesting, although it does not seem to be directly relevant to why James actually relates this story, this is, to show that he actually can differentiate gays from non-gays. However, this piece of information makes the story more tellable: James presents the information about the gay boy as ‘second-hand knowledge. He uses ‘reported speech’ (here in the form of ‘indirect speech’, i.e., as a summary quote) to recreate the action in question (= having seen gays in their school) through the talk of someone else who is held socially accountable. He skillfully introduces an overhearing (though nameless) witness, who is characterized as female, honest, and nice, and as having her locker right next to the boy whose


229

reputation is at stake in this account. It is this girl who is presented as overhearing the speech actions of the boy that give rise in the unfolding story to the characterization ‘gay’. And supposedly this girl has reported this information back to James. In sum, James’s attempt to regain his credibility and authority (on ‘gay issues’) rests on his presentation of an overhearing eyewitness and relaying the crucial information as hearsay. And by placing his reputation as knowledgeable in the hands of this witness and her reputation, he is able to successfully ‘hide’ behind this eyewitness. Thus, the question arises, how does he manage to come across as believable in spite of the fact that he himself does not have any first-hand knowledge—at least not in this particular case? James seems to be accomplishing several activities at the same time: First, when openly challenged not to be able to differentiate gays from non-gays, he successfully (re)establishes his authority. He lists a witness’ account and rhetorically designs this witness as reliable. This witness is ‘honest’ (in contrast to ‘a liar’) and ‘nice’ (in contrast to ‘malicious’ or ‘notoriously gossiping’). In addition, giving details such as ‘her locker next to his’ contribute further to the believability of James’s account. Furthermore, the characterization of the boy as talking ‘a lot’ about his gayness, makes it difficult to (mis)-interpret the girl’s (and James’s) accounts as potential misreadings. Second, introducing his witness as a girl (note that James could have left the gender of this person unspecified), and in addition as one who did not talk directly to the gay boy, further underscores how James wants to be understood: In line with his corrective statement in turn 6 (‘just having seen gay boys, not really knowing them’), to have a close confederate who is also close to the gay boy (and speaking with him ‘a lot’) could make this confederate hearable (again) as in close relationship with a ‘gay community’. Thus, designing this confederate as a girl, who is not even being addressed by the gay boy when he talks about his sexual orientation, makes it absolutely clear that there is no proximity nor any other possible parallel between this boy’s orientation and James’s. A girl is a perfect buffer that serves the role to demarcate the difference in the sexual orientations of James and the gay classmate. Third, James’ staging of the ‘fact’ that this boy ‘associates with a lot of girls and not boys’ (except with a few other gay kids at school) at the very end of his story, is very telling. Had James mentioned this at the beginning—that is, as his abstract and orientation for why he is sharing his account—he could have easily been heard as too quickly buying into the typical (stereo-typical) view of gays. And this could have resulted in further challenges from Ed and Alex as just talking ‘from the top of his

230

Michael Bamberg

head’ and not really knowing. However, placing this generalized statement at the end of his very detailed account, and giving it the slot of the coda, he uses this typicalization to finish the storied account and orient the conversation toward why it is that gays hang more often out with girls, and this is what actually happens in the talk that follows. It may be fair to say that the more general group-level ascriptions of the boy as hanging out with girls and gays is more likely to be heard as stereotypical if followed by his carefully scripted account of how he actually knows about particular gay boys at their school. In other words, this way of strategically sequencing his ‘evidence’ allows James to epitomize the group of gays by having captured the individual in relation to the aggregate; and in turn helps James to move himself back into the group of ‘his peers’ by drawing a boundary between ‘them’ and ‘us’. In sum, James’s story is doing multiple things: When openly challenged that he doesn’t know how to differentiate gays from non-gays, his story enables him to re-establish his identity as knowledgeable and reliable; furthermore, it helps him to fend off coming across as gossiping and being heard as prejudiced, that is as antigay or homophobic. However, most important, his story allows him to carefully fashion himself as heterosexual and straight. It is in this sense that his story reflects masculine norms and a sense of heteronormativity. However, as I would like to argue, this sense of a heteronormative self—just like his sense of self as a cool authority in ‘gay issues’, a non-gossiper, and as someone who is not homophobic and prejudiced—are all active accomplishments of the participants who in concert put these norms to use. They are achieved by the way this story is situated and performed within this very local setting. Thus, it is the situation that determines the logic or meaning of the norms being circulated, and not the boys’ cognitions or previously established concepts that they seem to have acquired elsewhere and now ‘simply’ bring to their interactive encounters. And it is in this sense that the boys (as members of the social category ‘boys’) are both producing and being produced (or ‘acquired’—see Hall 2004) by the routines that surround and bring off these kinds of narratives-in-interaction. And although our particular ‘small story’ in a strict sense is the response to the challenges by Ed and Alex in turns 7 and 8, it answers a number of other identity challenges that are hearable in the way the story is made to fit into the ongoing negotiation. It should be stressed that this particular local ‘small story’ as an exercise in maneuvering through the challenges of gossiping, homophobia, and heteronormativity is simultaneously a practice of negotiating competing ideological positions. It is in situations like this that children and adolescents, but also adults in the form of a life-long process, draw on multiple subject positions; positions that can be used to


231

be complicit or to counter existing master narratives (Bamberg, 2004b). Practicing ‘small stories’ are indispensable stepping-stones in the identity formation process of the person.

9. Concluding Remarks My contribution to this volume has been intended as a question-asking chapter rather than one that lays out clear-cut orientation guidelines for narratologists. It emerged from my puzzlement with why narratologists have become embraced and increasingly seem to be embracing cognition, while there are other (better) alternatives available. As I have stated repeatedly, for social scientists whose interest lies in people’s identities, the question of what narrative really is (that is, the definition of ‘narrative’ as a literary or oral category), is not relevant. I am working with what people tell us, but equally important, with how they tell their stories. The story that I briefly analyzed (as an example in the previous section of this chapter) may not even count as a ‘narrative’ to some of my readers. But that is beside the point. Narratives-in-interaction are not particularly privileged speech genres. They happen. And the analysis of these ‘happenings’ does not provide a deeper or better window into people’s lives. It is one of many. However, narratives are ‘interesting’ and ‘telling’ devices, since they usually enable speakers to arrange their claims in a ‘more organized’ fashion: Speakers, with their narratives, react to previous pieces of the interaction, and orient, with their temporal and spatial layout of the narrative, to the future course of talk-in-interaction. How speakers are entering the floor and are managing to hold the floor by successfully blocking off interruptions or objections, and how they constantly monitor how they will be heard, gives us better insights into how several simultaneous positions by a singular speaker are brought off and managed in synchrony. This is what I attempted to demonstrate with James’s ‘How do I know they’re gay?’ narrative. And as I hope I was able to show, this narrative is only to some degree about ‘how I know’—and to an equal, if not larger, degree about James’ self-claims as a non-gossiper, as not prejudiced vis-à-vis homosexuals, and as a ‘normal heterosexual’. The focus of telling the story is on the creation of ‘normalcy’ and to claim this normalcy for the moment of this narrative-in-interaction. The narrative is rhetorically designed to do this job for him. Neither does it reflect that James ‘is’ normal, nor was the narrative brought off ‘because’ James ‘is normal’. I think of this mundane insight as a beginning for continued work with narratives; it is definitely toward the end of more and better analyses of identities-in-the-making. But also, I hope, it has some potential

232

Michael Bamberg

to energize discussions around issues of what narratives are and how it is possible that we can do such interesting things with them. Let me conclude with some final thoughts on ways in which the concept of positioning may help illuminate and re-conceptualize such notions as ‘focalization’ and ‘evaluation’. As should have become clear from the aforementioned, working with oral ‘narratives-in-interaction’ as presented in this chapter sympathizes with Fludernik’s suggestion to ‘scrap the concept of focalization in its traditional configurations’ (1996:346) and to develop alternatives that start from and can be applied to ordinary, every-day stories, the way they occur in the form of ‘small stories’. In contrast to Fludernik, however, I would like to suggest that the project of ‘recontextualizing narrative’ should not rest on the notion of consciousness nor cognition, but on ‘action’—or better, on what is being accomplished between co-conversationalists in terms of their strategic management of positioning selves vis-à-vis others and vis-à-vis dominant subject positions in the form of master narratives. The concept of ‘evaluation’, as developed by Labov (1972) and expanded by Polanyi (1989), can equally be better accounted for in terms of ‘positioning strategies’. Rather than assuming that there are some internal mechanisms (operating at the time of the actual experience or at the time of telling the experience) that cue the speaker into his / her evaluation (and cue the listener into the point for the telling), positioning emphasizes the interactional accomplishment of ‘doing evaluating’. Furthermore, the linguistic (and supra-segmentational) devices that actually result in what can be read as ‘evaluative stance’ are not only all over the delivery of the text, but more relevant, they point to previous and subsequent speech as well as to larger aspects of the context; they function as ‘contextualization cues’ (Gumperz, 1982). ‘Positioning’ should incorporate what ‘focalization’ and ‘evaluations’ are supposed to accomplish. Sorting out the linguistic and supra-segmentational performative means that index what’s going on in the represented world of the story (the way characters are positioned vis-à-vis one another), and adding them to the layer of discursive means that index the relational work that is being accomplished between the co-tellers / audience (where tellers and co-conversationalist gain their interactional identities), we are better equipped to understand how tellers become positioned vis-à-vis themselves with claims they hold to be true and relevant above and beyond the local conversational situation and in relation to dominant ideological subject positions. This should not be misunderstood as wanting to do away with ‘evaluative perspective’ and ‘focalization strategies’. However, both can become incorporated in the more general strategic orientation of speakers to convey one’s sense of self in the face of challenges to one’s


233

identity positions—may they be concrete and actual in the form of challenges by interlocutors in the communicative situation, or may they be more abstract, imagined and anticipated in the form of demands by competing discourses that always seem to position us. References Antaki (in press) Antaki, Charles: “The Uses of Absurdity.” In: van de Berg, Harry / Wetherell, Margaret / Houtkoup-Steenstra, Hanneke (eds.): Analyzing Race Talk: Multidisciplinary Perspectives on the Research Interview. Cambridge, UK (in press). Bamberg 1997 Bamberg, Michael: “Positioning Between Structure and Performance.” In: Journal of Narrative and Life History 7, 1997:335–342. Bamberg 2000 Bamberg, Michael: “Critical Personalism, Language, and Development.” In: Theory & Psychology 10, 2000:749–767. Bamberg 2003 Bamberg, Michael: “Positioning with Davie Hogan – Stories, Tellings and Identities.” In: Daiute, Colette / Lightfoot, Cynthia (eds.): Narrative Analysis: Studying the Development of Individuals in Society. London 2003:135–157. Bamberg 2004a Bamberg, Michael: “‘I know it may sound mean to say this, but we couldn’t really care less about her anyway.’ Form and Functions of ‘Slut-Bashing’ in Male Identity Constructions in 15-Year-Olds.” In: Human Development 47, 2004:331–353. Bamberg 2004b Bamberg, Michael: “Talk, Small Stories, and Adolescent Identities.” In: Human Development 47, 2004:366–369. Butler 1990 Butler, Judith: “Performative Acts and Gender Constitution: An Essay in Phenomenology and Feminist Theory.” In: Case, Sue-Ellen (ed.): Performing Feminisms: Feminist Critical Theory and Theatre. Baltimore 1990: 270–282. Butler 1995 Butler, Judith: “Collected and Fractured: Responses to Identities.” In: Appiah, Kwame Anthony / Gates, Henry Louis (eds.): Identities. Chicago 1995:439–447. Catan 1986 Catan, Liza: “The Dynamic Display of Process: Historical Development and Contemporary Uses of the Microgenetic Method.” In: Human Development 29, 1986:252–263. Eckert 1989 Eckert, Penelope: Jocks and Burnouts: Social Categories and Identity in the High School. New York 1989.

234

Michael Bamberg

Eckert 2002 Eckert, Penelope: “Demystifying Sexuality and Desire.” In: Campbell-Kibler, Kathryn/ Podesva, Robert J./ Roberts, Sarah/ Wong, Andrew (eds.): Language and Sexuality: Contesting Meaning in Theory and Practice. Stanford, CA 2002:99–110. Edwards / Potter 1992 Edwards, Derek / Potter, Jonathan: Discursive Psychology. London 1992. Fillmore 1975 Fillmore, Charles: “An Alternative to Checklist Theories of Meaning.” In: Berkeley Linguistic Society 1, 1975:123–131. Fillmore 1982 Fillmore, Charles: “Frame Semantics.” In: The Linguistic Society of Korea (ed.): Linguistics in the Morning Calm. Seoul 1982:111–137. Fischer-Rosenthal / Rosenthal 1997 Fischer-Rosenthal, Wolfram / Rosenthal, Gabriele: “Narrationsanalyse biographischer Selbstrepräsentationen.” In: Hitzler, Ronald / Honer, Anne (eds.): Sozialwissenschaftliche Hermeneutik. Gießen 1997:133–164. Fludernik 1996 Fludernik, Monika: Towards a ‘Natural’ Narratology. London 1996. Fludernik 2003 Fludernik, Monika: “Natural Narratology and Cognitive Parameters.” In: Herman, David (ed.): Narrative Theory and the Cognitive Sciences. Stanford, CA 2003:243–267. Freeman 2004 Freeman, Mark: “The Matter of the Text.” In: Narrative Inquiry 14, 2004: 29–43. Fuchs-Heinritz 2000 Fuchs-Heinritz, Werner: Biografische Forschung. Eine Einführung in Praxis und Methoden. 2nd edition, Opladen 2000. Goffman 1967 Goffman, Erving: “On Face Work.” In: Goffman, Erving (ed.): Interaction Ritual. New York 1967:5–45. Goody 1977 Goody, Jack: The Domestication of the Savage Mind. Cambridge 1977. Goody / Watt 1968 Goody, Jack / Watt, Ian: “The Consequences of Literacy.” In: Goody, Jack (ed.): Literacy in Traditional Societies. Cambridge 1968:27–68. Greenwood 2004 Greenwood, John D.: The Disappearance of the Social in American Social Psychology. Cambridge 2004. Gumperz 1982 Gumperz, John: Discourse Strategies. Cambridge 1982. Hall 2004 Hall, Rogers: “Attaching Self and Others to Social Categories as an Interactional and Historical Achievement.” In: Human Development 47, 2004:354–360. Hanks 1996 Hanks, William F.: Language and Communicative Practices. Boulder, CO 1996.


235

Herman 2002 Herman, David: Story Logic. Problems and Possibilities of Narrative. Lincoln, NE 2002. Herman, David 2003a Herman, David: “Introduction.” In: Herman, David (ed.): Narrative Theory and the Cognitive Sciences. Stanford, CA 2003:1–30. Herman, David 2003b Herman, David: “Stories as a Tool for Thinking.” In: Herman, David. (ed.): NarrativeTheory and the Cognitive Sciences. Stanford, CA 2003:63–192. Heath 1983 Heath, Shirley Brice: Ways with Words. Cambridge 1983. Hogan 2003 Hogan, Patrick C.: The Mind and Its Stories. Narrative Universals and Human Emotion. Cambridge 2003. Hogan 2004 Hogan, Patrick C.: Cognitive Science, Literature, and the Arts. Cambridge 2004. Iser 1974 Iser, Wolfgang: The Implied Reader. Baltimore, MD 1974. Iser 1978 Iser, Wolfgang: The Act of Reading: A Theory of Aesthetic Response. Baltimore, MD 1978. Josselson 2004 Josselson, Ruthellen: “The Hermeneutics of Faith and the Hermeneutics of Suspicion.” In: Narrative Inquiry 14, 2004:1–28. Korobov / Bamberg (in press) Korobov, Neill / Bamberg, Michael: “Positioning a ‘Mature’ Self in Interactive Practices: How Adolescent Males Negotiate ‘Physical Attraction’ in Group Talk.” In: British Journal of Developmental Psychology (in press). Labov 1972 Labov, William: “The Transformation of Experience in Narrative Syntax.” In: Labov, William: Language in the Inner City. Philadelphia 1972:354–396. Lakoff 1977 Lakoff, George: “Linguistic Gestalts.” In: CLS 13, 1977:236–287. Lakoff 1987 Lakoff, George: Women, Fire, and Dangerous Things: What Categories Reveal About the Mind. Chicago 1987. Nünning / Nünning 2002 Nünning, Ansgar / Nünning, Vera: “Von der strukturalistischen Narratologie zur ‘postklassischen’ Erzähltheorie: Ein Überblick über neuere Ansätze und Entwicklungstendenzen.” In: Nünning, Ansgar / Nünning, Vera (eds.): Neue Ansätze in der Erzähltheorie. Trier 2002:1–33. Ochs / Capps 2001 Ochs, Elinor / Capps, Lisa: Living Narrative. Creating Lives in Everyday Storytelling. Cambridge, MA 2001. Olson 1997a Olson, David R.: “On the Relations Between Speech and Writing.” In: Pontecorvo, Clothilde (ed.): Writing Development: An Interdisciplinary View. Amsterdam 1997:3–20.

236

Michael Bamberg

Olson 1997b Olson, David R.: “The Written Representation of Negation.” In: Pragmatics and Cognition 5, 1997:235–252. Ong 1982 Ong, Walter: Orality and Literacy: The Technologizing of the Word. London 1982. Polanyi 1989 Polanyi, Livia: Telling the American Story. Cambridge, MA 1982. Potter 1996 Potter, Jonathan: Representing Reality: Discourse, Rhetoric, and Social Construction. London 1996. Potter / Wetherell 1987 Potter, Jonathan / Wetherell, Margaret: Discourse and Social Psychology: Beyond Attitudes and Behaviour. London 1987. Riegel 1975 Riegel, Klaus F.: “Toward a Dialectical Theory of Development.” In: Human Development 18, 1975:50–64. Rosch 1975 Rosch, Eleanor: “Cognitive Reference Points.” In: Cognitive Psychology 7, 1975:531–547. Rosch 1978 Rosch, Eleanor: “Principles of Categorization.” In: Rosch, E. / Lloyd, B. B. (eds.): Cognition and Categorization. Hillsdale, NJ 1978:27–48. Rosenthal / Fischer-Rosenthal 2000 Rosenthal, Gabriele / Fischer-Rosenthal, Wolfram: “Analyse narrativ-biografischer Interviews.” In: Flick, Uwe / v. Kardorff, Ernst / Steinke, Ines (eds.): Qualitative Forschung. Ein Handbuch. Reinbeck 2000:456–468. Sacks 1995 Sacks, Harvey: Lectures on Conversation: Volumes I and II. Oxford 1995. Sacks / Schegloff / Jefferson 1974 Sacks, Harvey / Schegloff, Emanuel A. / Jefferson, Gail: “A Simplest Systematics of the Organization of Turn-taking for Conversation.” In: Language 50, 1974:696–735. Sanders 1995 Sanders, Robert E.: “A Neo-Rhetorical Perspective: The Enactment of Role-Identities as Interactive and Strategic.” In: Sigman, Stuart J. (ed): The Consequentiality of Communication. Hillsdale, NJ 1995:67–120. Schegloff 1982 Schegloff, Emanuel A.: “Discourse as an Interactional Achievement: Some Uses of ‘uh huh’ and Other Things That Come Between Sentences.” In: Tannen, Deborah (ed.): Analyzing Discourse: Text and Talk. Washington, DC 1982:71–93. Schank / Abelson 1977 Schank, Roger C. / Abelson, Robert P.: Scripts, Plans, Goals and Understanding: An Inquiry into Human Knowledge Structures. Hillsdale, NJ 1977.


237

Sigman 1995 Sigman, Stuart J.: “Introduction: Toward Study of the Consequentiality (Not Consequences) of Communication.” In: Sigman, Stuart J. (ed.): The Consequentiality of Communication. Hillsdale, NJ 1995:1–14. Taylor / MacLaury 1995 Taylor, John A. / MacLaury, Robert E.: Language and the Cognitive Construal of the World. Berlin 1995. Talbot et al. 1996 Talbot, Jean / Bibace, Roger / Bokhour, Barbara / Bamberg, Michael: “Affirmation and Resistance of Dominant Discourses: The Rhetorical Construction of Pregnancy.” In: Journal of Narrative and Life History 6, 1996: 225–251. (Also at ) Taylor / Fiske 1981 Taylor, Shelley E. / Fiske, Susan T.: “Getting Inside the Head. Methodologies for Process Analysis in Attribution and Social Cognition.” In: Harvey, John H. / Ickes, William / Kidd, Robert F. (eds.): New Directions in Attribution Research. Hillsdale, NJ 1981:459–524. Wenger 1998 Wenger, Etienne: Communities of Practice. Cambridge 1998. Werner 1957 Werner, Heinz: Comparative Psychology of Mental Development. New York 1957. Werner / Kaplan 1984 Werner, Heinz / Kaplan, Bernard: Symbol Formation. Hillsdale, NJ 1984. Wertsch / Stone 1978 Wertsch, James / Stone, Addison: “Microgenesis as a Tool for Developmental Analysis.” In: Quarterly Newsletter of the Laboratory of Comparative Human Cognition 1, 1978:8–10. Wierzbicka 1999 Wierzbicka, Anna: Emotions Across Languages and Cultures: Diversity and Universals. Cambridge 1999. Wolf 2002 Wolf, Werner: “Das Problem der Narrativität in Literatur, bildender Kunst und Musik: Ein Beitrag zu einer intermedialen Erzähltheorie.” In: Nünning, Vera / Nünning, Ansgar (eds.): Erzähltheorie transgenerisch, intermedial, interdiszipliär. Trier 2002:23–104.

Psychotrauma, Narration in the Media, and the Literary Public—and the Difficulties of Becoming Interdisciplinary1 Harald Weilnböck (Berlin) 1. Introduction In what ways and under what conditions can narration contribute to the therapeutic process of coming to terms with traumatic experiences? How might we improve our concept of narration, be it in the context of therapy, the public domain, creative writing, or the reception and production of media output, in order to describe more effectively and understand better the processes of narrative interaction that we know to alleviate emotional suffering and reduce social tension? Why are narration and its counterpart, the equally narrative act of reception performed by conversation partners, readers, and the audience of the media, therapeutic in the sense that narration helps prevent further emotional suffering and aggression? On closer consideration, it is clear that the question of the relationship between psychotrauma, narration, and therapy expresses nothing more (and nothing less) than the central concern of bourgeois artistic and literary production since Goethe and, more particularly, since Schiller’s theory of the aesthetic education of humanity: how can art improve the lot of humanity, individuals, and communities alike? My approach to this question is based on the study of psychotrauma, part of the wider field of psychotherapy. Psychotraumatology is the science of mental injuries and how they can be healed. The study of psychotrauma draws on the findings of developmental and cognitive psychology, especially child and infant psychology, and thus has a scientific, experimental basis. However, the study of psychotrauma also draws on qualitative so1

The fifth section of this article originally included a narratological close reading of a sequence of interaction between participants in the conference itself. Since the editors decided that this section was not appropriate for inclusion here, I have omitted it from this article. Instead, the close reading will appear in Forum: Qualitative Social Research () in autumn 2004.

240

Harald Weilnböck

cial psychology and psychotherapy studies in which methods of narrative analysis are employed, so it is also based on textual analysis.2 This implies working under the theoretical assumptions of what in German is generally called Handlungstheorie, probably best described as an approach to cultural phenomena that is based on the study of action and interaction.3 This approach is much less a theory of texts than a theory of how the human mind interacts with texts. Methodologically, it encompasses textual analysis and qualitative reader-response research. The above position has an unusual effect on the way in which I work in that my thoughts seem to speak with two voices. It is as if I had two different intellectual selves—which is probably not that uncommon among Germanists well-acquainted with the Faustian experience of ‘zwei Seelen, ach!, in meiner Brust.’ One of the inner voices seems to belong to a literary theorist, the other to someone engaged in research on psychology and human interaction. From time to time, these voices conflict because the analysis of interaction not only treats narration as a text with content and form but also, and primarily, focuses on the specific affect-related psychological functions that narration has for its author and readers. The psychologist’s voice in me says that if people tell stories to other people, whether orally or in writing, they do not do so because they want to create a particular form or implement certain textual features in order to stimulate cognitive responses in the recipient. From what we have learnt from developmental and infant psychology, we know that the human mind simply does not work in this way. We have only to think of young children when they tell stories, when their eyes widen and they start breathing heavily, finding it difficult to draw enough air into their lungs to express the narratives they want to express. The narratives of these children do something more than produce content and form or transmit units of cognitive information. Narrating children have feelings, strong feelings—and thoughts at the same time. And they manage to bring those feelings across to their listeners. If you listen to a child telling you a story, you end up having some of the same feelings that the child had while narrating. These feelings and affects migrate, so to speak, from the child to you: they are transferred. If this does not take place, there is something seriously wrong with the way in which you and the child interact with one another, for the entire affectrelated psychological development of the child is dependent on this kind of 2

3

See Fischer/Riedesser 1998. On the applications of psychotrauma studies in the cultural sciences, see Mauser/Pietzcker 2000. Hirsch (1998) focuses on interaction-based microtraumas and object-relation traumas, and Weilnböck (2001) discusses the use of the concept of trauma in literary theory. For examples of the analysis of literature from a psychological perspective, see Pietzcker (1992), Raguse (1994), and Weilnböck (2004a, 2005).

Psychotrauma and Narration

241

narrative exchange and affect transference.4 As we intuitively have known since the tales about Casper Hauser, children cannot develop properly without telling narrative stories. The admirably lucid and helpful introduction to literary narratology by Martinez and Scheffel can be seen as representing my literary self. When I read their definition of narration as a process in which ‘someone tells a certain something to someone else’, and when I see them proceed to concentrate on the ‘certain something’ of narration, my literary self is content (Martinez / Scheffel 1999:9). But the psychologist in me protests at once; he wants to focus on the ‘someone’ and ‘someone else’ of narration as well. And he wants to do textual analysis at the same time. It is here that my difficulties begin, because, for some reason that is not yet entirely clear to me, the psychologist’s focus on the affect-related psychological functions of narration seems a dangerous methodological addition rather than a useful one. It threatens the voice of the literary theorist in me. And here the conflict begins. If we ignore this inner conflict for a moment, the main question can be stated as follows: how can the study of psychotrauma be of assistance to literary theory? In this context the concept of the subjective traumatic experience is of crucial importance. Mental injuries and psychotraumatic effects are caused not by the event itself, even if it leads to physical injury, but by the subjective impact of the event. In this respect, we can conceive of trauma as a person’s subjective trauma narrative about an event (or about a continuing situation of interaction-related microtraumas).5 A subjective trauma narrative of this kind clearly has everything that, from a psychological point of view, narratives would appear to require. On the level of its content, there are subjects and objects (in this case, perpetrators, bystanders, and victims), and there are intentions and actions (i.e. processes of interaction and consequences of actions). On the level of the narrative act itself, the subjective trauma narrative creates a process of interaction with a listener. Narration is thus a key concept of psychotherapy, which can reasonably be expected to be of benefit to the study of narratological questions. In order to arrive at a more precise understanding of the potential benefits of the conceptual interface between narratology and psychotherapy, 4 5

See Dornes 1993, Pantoja 2001, and Bohleber/Drews 2001:304–346. This definition goes beyond and is more specific than some of the more recent psychological definitions of the narrative, which assume that mental activity in general is best described in terms of the constant production and modification of mental narratives (see Geißler 2002:37–84, and Pantoja 2001). These more recent concepts are of great significance for literary theory and the cultural sciences since they indicate that the ability to form narratives is present at a very early stage in development.

242

Harald Weilnböck

we can begin by asking how trauma narratives work. It is clear that they are a particular kind of narrative with certain characteristics. They are horrifying because of the experience narrated in them, and they are always incomplete and disfigured narratives: when the point is reached at which the mental injury becomes subjectively unbearable and turns into an unspeakable horror (‘namenloser Schrecken’), the narrative collapses and the story can no longer be told in its entirety. The traumatic part of a person’s life experience loses its narrativity (see Marie-Laure Ryan’s article in this book).6 This does not mean, however, that the traumatized person ceases to speak; shock and speechlessness do not last significantly longer than the traumatic situation itself. After that, in one form or another, narration takes over.7 From a narratological point of view, then, trauma is the presence of an uncompleted narrative, a narrative process gone astray. At this point, however, the literary theorist in me voices his dissent: how can we conceive of an incomplete narrative in the first place? Is such a thing even possible? We are compelled to treat every literary text as a complete one, no matter how disjointed and fragmented it may seem—who are we to say whether someone’s narrative is less complete than it could be, let alone should be?8 Psychologists adopt an entirely different position. Their work is based on the assumption that narratives are by their very nature continuously changing works in progress and always bear the potential for further development. The voice of the psychologist in me, in fact, describes my respective role as follows: I interact with incomplete narratives, helping the narrator to make the narrative more complete so that its suppressed and / or forgotten parts loose their negative impact on the narrator’s life.9 If one assumes that the psychologist is right, that narratives can indeed go astray and be more or less complete, we have then to ask what an incomplete narrative looks like and how it can be completed. It will be helpful to introduce the concepts of trauma compensation patterns and trauma defence mechanisms at this point. These mechanisms are psycho6

7

8

9

Werner Wolf has suggested introducing Wolfgang Iser’s (1975) concept of the Leerstelle to help conceptualize the incompleteness of the trauma narrative in more precise narratological terms. In fact, it is possible that the absence of an end from a person’s account might increase in prominence in proportion to how speechless the person was when terrified. Such a narrative, then, is endless and incomplete at the same time. This reservation is all the more pressing if we want to avoid following literary critics who make judgements of taste (e.g. Marcel Reich-Ranicki). Steering clear of such an approach is not as easy as it might seem; for example, scholars often choose what they study on the basis of their own aesthetic preferences. This is why the well-know therapist Roy Schafer once defined psychotherapy as the process of telling the same story over and over again—but ultimately telling it in a different and more complete way.


243

logical and narrative devices that suppress or distort memories originating in the traumatic situation. They function by means of screen memories, which are emotionally endowed with screen affects, thus generating screen narratives. A screen memory does not represent the actual scene of traumatic interaction itself; it does not contain the core narrative of the experience. Instead, it focuses on a marginal aspect of the experience, separates this aspect from its context,10 and thereby avoids the memory of the traumatic context altogether and sidesteps the unbearable affects of panic and horror.11 It should be stressed that, although (or indeed because) screen memories and screen narratives suppress and cut off crucial parts of the traumatic experience, they enable the person concerned to keep going—at least for a time, for, in the long run, they prevent that person from recovering properly, from becoming less traumatized by arriving at a more complete trauma narrative. This process of reaching narrative closure is what therapists mean by the therapeutic process of coming to terms with something. The second major aspect of the role of narration in the context of trauma and therapy is now apparent: not only is trauma really a narrative, albeit an incomplete one, therapy is a narrative process too, the process of completing the narrative. Narration holds the key to dissolving the destructive impact of a traumatic experience. It is now more clear what the concept of a complete narrative might mean. It requires that a recounted story completely fulfil its author’s individual needs for the purposes of trauma therapy. In general, this means that the narrative should be factually complete (i.e. the roles of perpetrator and victim, their intentions, and the actual events should all be accounted for and psychologically processed). It also means that the narrative should be emotionally rich (it should include even the most unpleasant and horrifying affects attached to the traumatic experience). Once more, the voice of the literature-lover in me protests. Violently disagreeing with the above view, it responds that the literary text is a work of art, belongs to the realm of the spiritual, and thus by its very nature exists on a level above that of the utilitarian satisfaction of the needs of 10

11

The defensive mechanisms of screening and dissociating the screen memory’s content from its traumatic context resemble what is known in classical rhetoric and poetics as metonymic shift (contrasted with the metaphorical association of images). In this respect, integrating psychotherapy and the cultural sciences could result in metonymy and metaphor being seen not only as different rhetorical techniques but also as reflecting differences in the processes by means of which the human mind comes to terms with its experiences (see Weilnböck 2004b). Bearing in mind the debate concerning the pictorial or iconic turn in the humanities, it should be highlighted that screen memories often take the form of visually intensified images (see Rohde-Dachser 2000:88, 96). As early as 1973, Heinz Kohut spoke in terms of super-intense Polaroid memories.

244

Harald Weilnböck

an author undergoing trauma therapy in particular and the mundane or egotistic intentions of an author in general. Otherwise, it is not art but something more trivial. But the psychotraumatologist in me retorts that all storytelling, whether in oral speech or written literature, is in some way associated with and dependent on the personal needs and intentions of its author, and if an author or critic thinks otherwise, it can only be due to a refusal originating from psycho-cognitive screen-processes. No one would dispute that narration, like therapy, only functions in conjunction with another person who listens or reads with empathy.12 The recipient does this by receiving the narrator’s transferences of affects and associations and engaging with them personally.13 In order to prevent us from losing sight of this fact, I propose extending our conceptual apparatus with two new terms: narrative trauma association and empathetic narrative resonance. These concepts are intended to allow us to grasp better the fact that people who listen to trauma narratives can do so in a therapeutic manner only if they tap into their own trauma associations and trauma narratives. This means becoming aware of the personal experiences and associations that resonate with the account of violence and injury told to them. Although the associations of trauma and violence made by the listener may vary in kind and in psychotraumatic magnitude, it remains the case that therapeutic storytelling and listening cannot take place without listeners having their own trauma narratives completed and placed in a position where their emotional aspects can begin to be considered. It is by no means easy to achieve this, for it is hard to remember traumatic scenes in which one has been a victim, and even harder to remember scenes in which one has been a perpetrator. This latter condition is the second major prerequisite for the success of the process of transference. By definition, narrative trauma associations consist of scenes of interaction that contain not only the victim but also the perpetrator. And, being psy12

13

The literary theorist in me often forgets about the listening other; this reflects the general situation in literary theory, which has seen very little qualitative sociological research on the reception of the media (see Charlton/Schneider 1997, Weilnböck 2003a, and Weilnböck/Garbe/Schoett 1999). Transference means an unwitting exchange of affects and associations that takes place between two individuals interacting verbally and non-verbally. Feelings, even fantasies, associations, and thoughts can move from one person to another, without either party consciously intending or even being aware of it (this is also true of mediated interaction such as that between authors and readers, authors and texts, and texts and readers). There is no reason to treat processes of transference as miraculous or occult phenomena. They happen all the time when people interact, and they are historically and biographically developed, based on real-life events and historical and personal experiences shared by a community or society. See Racker 1978, Bohleber/Drews 2001:347, Mertens/Waldvogel 2000:226–232, 758–771, Kernberg/Dulz/Sachsse2000:87–97, and Tschuschke 2001: 160–163.


245

chological phenomena, trauma associations always introduce both aspects at once, no matter whether we are listening to the narrative of a victim or perpetrator. So, having a trauma narrative completed and ready to be confronted emotionally really means being aware of our experiences—be it as victims, or be it as perpetrators. Given that the stimulation of empathetic narrative resonance is such a complicated and vulnerable process, and given that it can fail or be distorted, we need to ask when empathy (and therapeutic transference) can fail, and what we are left with when this happens. At this point, I would like to introduce the concept of screen empathy, which I define as a special kind of listener response in which the screen affects of the traumatized person, and not the traumatizing affects themselves, are shared and reaffirmed. The listener who experiences screen empathy does not realize that the story provided is incomplete and that some traumatizing experiences and affects have not been expressed. Such a listener cannot and does not want to help complete the story. Screen empathy is, as it were, empathy with a person’s defence rather than with that person’s suffering. In psychoanalytic terms, experiencing screen empathy with a traumatized person means engaging in a process of parallel (counter-)transference (‘konkordante [Gegen]-Übertragung’; see Racker 1978:124–134). The affects of narrator and listener function in parallel, serving the common function of fending off unpleasant affects; consequently, the therapeutic process of coming to terms with what happened cannot take place properly. For the process to be successful, a different kind of transference, which we can call complementary (counter-)transference, is necessary. In it, the interacting pair of narrator and listener / reader can move beyond screen affects and touch on the suppressed affects and associations that the traumatized person, initially unable to feel and articulate them, transfers by means of non-verbal (and verbal, but non-explicit) interaction. In this manner, the interactive process of working through and complementing the trauma narrative unfolds. Screen actions and screen empathy, therefore, inhibit what is potentially the most productive, creative, and healing of all forms of human interaction: the construction and sharing of a common narrative about the experience of mental injury. A few words on trauma dependency and compulsive repetition should be said here. Screen affects never truly fulfil a person’s need to come to terms emotionally with experienced events; they substitute for this by providing the necessary degree of self-stimulation without actually removing the oppressive burden of the unarticulated traumatic experience. The need for the substitute can become so great that people strive continuously to repeat parts of the traumatic situation in order to feel the associated screen affects. For this reason, they often turn to forms of interaction characterized

246

Harald Weilnböck

by addiction. The object of addiction might be a chemical substance or a certain form of personal interaction. In the case of the media, an activity such as reading books or watching films might be performed because it had the effect of stimulating controlled retraumatization.14 2. Face-to-face and / versus Mediated Narration As far as its function is concerned, mediated narration is no different in any fundamental sense from therapeutic trauma narrative. Nonetheless, there are inevitable secondary differences with associated methodological and procedural consequences. In literary and / or mediated interaction, there is no direct contact with the listening other. The readers of literature or the audience of a film cannot react directly in an attempt to influence the author, and even hypertext is not significantly different in this respect.15 Even so, all authors write against the background of internalized mental representations of themselves and their readers, a fact inherited from direct face-to-face interaction, which is also guided by internalized representations and ultimately pre-structured to a far greater degree than is generally assumed in the context of ideological traditions based on individual liberty and autonomy. Another secondary difference lies in the fact that mediated interaction bridges a time gap, for the point in time of narration is temporally different from the point in time of listening, thus adding a second time gap to that contained in the link between an event or experience and its narration. This increases the probability that the trauma associations of historically distinct traumatic themes will overlap and intersect. But this is also true of face-to-face interaction, for, by definition, psychological associations are not constrained by restrictions of chronology and thematic coherence. It would appear, therefore, that the dichotomy between face-toface and mediated narration is less clear-cut than is generally assumed.16 The functional question posed by mediated narration is essentially the 14

15

16

In the case of interaction with literature and the media in the public sphere, adopting a pattern of screen empathy and trauma dependency might involve an institutionalized form of reader-author relationship in which author and audience interact primarily through screen actions, thereby strengthening their psychological defences. It is impossible to say conclusively whether we should conclude that defence mechanisms can be preserved more easily in mediated interaction or whether it is rather the case that mediated interaction, because it has more substantial content and greater imaginative potential, makes it easier to access and modify defence mechanisms. The same may be true of the distinction between factual and fictional narration. When approached from the point of view of sociology and therapy, it is easier to see the extent to which a factual story can be fictitious in the sense that it did not take place as an actual event, even though it is factual in the subjective perception of the narrator due to the effects of over-coding and various defence mechanisms.


247

same as that posed by its face-to-face counterpart: what kinds of associations and affects are activated by the author of a mediated narrative and transferred to its readers, and in what ways does this transference help or hinder the process of coming to terms with personal traumatic experiences and issues of communal trauma? Indeed, the presence of the (twofold) time gap helps to show that mediated narration, rather than being totally different from face-to-face narration, as is frequently assumed, actually adds another highly significant aspect to the way in which we consider the functioning of the interactive processes in which narratives enable us to come to terms with (traumatic) experiences. Because it is located in the public rather than the private sphere, media narration reveals collective processes of coming to terms with things. In these processes, an entire community or nation deals with publicly present traumatic issues regarding its past (this, of course, implies that legal and political questions have to be dealt with at the same time as the emotional issues). In view of the above, it would appear heuristically more sensible to treat face-to-face and mediated narration as dissimilar but not ontologically different. The difference between face-to-face and mediated narration on the one hand and narration in therapy on the other has been widely ignored. It must, however, be considered in order to develop further my belief that the study of therapy can provide considerable conceptual benefits for narratology. Because of the at times almost obsessive attention given to the distinction between face-to-face and mediated narration (often the result of efforts to defend against interdisciplinary tendencies), it has been largely ignored that narration in therapy differs from face-to-face and mediated narration in four ways. It takes place in a relational setting that is (1) asymmetric, (2) intended to achieve a specific goal, and (3) designed to ensure a continuous process of narration and narrative construction by means of interaction over an extended period of time; furthermore, (4) consensual validation is available as a means of evaluation. These properties are discussed in the following paragraphs. (1) The first feature captures the fact that, because one of the participants in therapy is there to receive therapeutic help from the other, both participants have fundamentally different positions, different functions, and different responsibilities. (2) For the same reason, the procedures and functions of this kind of interaction have a certain degree of constraint and focus, since its sole objective is the provision of therapeutic assistance. The participants in cultural / mediated and face-to-face interaction, on the other hand, have symmetric positions (and responsibilities), and there is no predefined focus for the functions of each of these forms and the ends that individuals can try to achieve with them. For example, the principal objective of therapy lies in dismantling and restructuring

248

Harald Weilnböck

the client’s psychological defences, for it is the defence mechanisms activated in response to traumatic experiences that cause the client to suffer. There is no place for reinforcing and reaffirming (pathological) defence mechanisms in a therapeutic context. In cultural / mediated interaction, on the other hand, reaffirming defence mechanisms can often be the main objective of individual authors / narrators and readers / listeners in writing and reading / listening. To put it another way, the therapist bears a certain amount of responsibility, but the participants in cultural interaction are not subject to any responsibility whatsoever, for they are protected by the rights of free expression and reception that form the indispensable foundation of modern western societies. (3) The third important difference between therapy and most cases of face-to-face and mediated narration is that the latter do not obtain or even seek the stable continuation of a particular process of narration; they permit the participants to withdraw from it at any time (this contrasts with the life-long process of mentally constructing biographical narratives, from which there is no escape). Client and therapist, however, agree that they will, for a certain period of time, regularly return to the task of continuing their shared process of narration whether they feel like it or not. This has productive consequences for the process. Readers and writers, on the other hand, are not required to go back to a particular reading or writing process (this does not, of course, mean that they will not sometimes feel drawn to return to it, even if against their will). (4) The fourth feature concerns the evaluation of the process of narration. In (psychoanalytically based) textual analysis, the literary critic is forced to reject the procedure of consensual validation with which therapist and client together can investigate and evaluate their understandings or interpretations of the process. The critic is faced with a concrete text rather than a process. When these differences between narration in therapy and face-to-face as well as mediated narration are considered, it becomes clear that the context will affect the way in which we approach the question of when narration is successful in the sense of being completed for the purposes of trauma therapy. The issue must be viewed differently depending on whether the situation is one of therapeutic or literary interaction. In addition, without even introducing methodological considerations, it should be clear that literature requires more than just the creation of a character or narrator who provides a complete (for the purposes of trauma therapy) narrative. Therapeutic progress has been made when patients undergoing therapy reach the stage of being able to provide a factually accurate, logically consistent, and emotionally rich account of traumatic experiences in their past. The literary presentation of such a patient as a character or narrator,


249

however, need not produce an aesthetically pleasing result. Furthermore, there is no guarantee that such a literary presentation will not be given in a form that triggers defensive affects because of the reader’s own trauma associations.17 In other words, when putting an individual experience of successful trauma therapy into writing in the sense of turning it into literature, it would not seem particularly desirable for the account to include verbatim reports of the interaction that took place in the therapy sessions. In conclusion, certain methodological and conceptual means of describing and evaluating a therapeutic process of coming to terms with something are applicable in the context of therapy, but they are no longer sufficient when applied to cases of mediated narration. It is important to note, however, that this does not, as might otherwise be assumed, point to the presence of fundamental theoretical differences. The same theoretical tools can be used to look at whether and how narratives ease trauma in both literary / mediated and therapeutic interaction alike. Nonetheless, the differences described above do have methodological consequences, and these we must consider. Once more, we find that the core of the problem lies in the fact that literary theory deals with static texts, concrete representations of the outcomes of psychological processes, rather than with the processes themselves. If we were to leave the text and its analysis in order to concentrate on reader response (and author input) and measure the effects of a process of literary reading or writing on an individual reader (or author), perhaps by means of narrative interviews, we would use more or less the same methodological procedures and evaluative criteria as are employed in qualitative research on psychotherapy. However, the literary theorist looks at the text rather than the way in which people interact with it. This has the important conceptual benefit of showing us that there is no such thing as a successful and complete trauma narrative in the substantialist sense of a complete printed text. A trauma narrative is not a text in the true sense of the word; it is a heuristic concept for a mental narrative process. Thus, we should speak only of complete and / or successful trauma-easing processes (not texts). When dealing with texts, this means that it is not appropriate to ask whether a text as a whole should be regarded as a complete and successful trauma narrative. Instead, we should ask whether the text contains features that can have the function of triggering trauma-easing processes, in particular complementary transference reactions, in people when they read the text (or in an author while writing it). 17

In my conference presentation, I developed this argument using the example of James Moll’s non-fiction Shoah film The Last Days, which was produced by the Steven Spielberg Foundation and won an Oscar for best documentary in 1998 (see Weilnböck 2004c).

250

Harald Weilnböck

3. Literary Working through of Traumatic Experience Having clarified to some extent what we should look for in order to understand what it means for a trauma narrative to be completed and ease trauma in a literary context, we can illustrate our findings with a concrete example. This is “Hunter-Tompson-Musik”, a short story by the contemporary German author Judith Hermann. Its main character, Hunter, seems to be entirely without family and home, and is caught in a dependency on the ramshackle hotel resembling a retirement home in which he lives, listening to music in his melancholy and radically withdrawing from most forms of social contact even though the desire to form friendships has not entirely died within him. Hunter’s behaviour, in particular his associations and fantasies, show that the narrator is depicting his personal situation as one characterized by dissociation (evident in his intense aggressiveness and rejection of social contact), double-binding, a deep-seated conflict between dependency and autonomy, and melancholy and depression. For example, when he speaks to the main female character for the first time, one of the narrative’s central events, Hunter suddenly associates with her the image of a cartoon duck strangled by another character (see Weilnböck 2002b). This does not make any sense, either narratologically or psychologically, unless we assume that the situation implicitly reflects a certain amount of prior traumatic experience. Combined with our knowledge of the way in which a character like Hunter interacts with others, we can conclude from this that his unsuccessful encounter with the young woman is a retraumatizing experience. The reasoning involved here takes place on the level of what I would call the implied biographical and psychological experiences of a literary character, which corresponds to Alan Palmer’s concept of the fictional mind in action (see Alan Palmer’s article in this book). The narrator does not tell us anything about the biographical history of Hunter prior to his present situation. However, readers see and sense Hunter’s state of mind via trauma associations of their own that resemble the experiences implicit in the presentation of Hunter as a character. Readers’ associations even include images and thoughts pertaining to biographical issues inherent in the character of Hunter. Most importantly, the text is structured in such a way that it permits and supports readers in coming to terms with trauma associations of their own that resonate with it. This is by no means self-evident, for Hunter and the narrator do not engage in or relate such a process of coming to terms. Nonetheless, although the protagonist of the narrative does not manage to alleviate his own suffering, the text gives readers the opportunity of reading it in a way that will help them come to terms with and increase the completion of their own personal trauma narratives.


251

How does the author of the text provide readers with this opportunity? The narrative is constructed in such a way that it activates two kinds of transference. One is defensive in nature, containing screen effects, primarily aggression and melancholy. The other serves to create a space in which the defended-against affects can appear, thereby furthering the process of coming to terms with what has happened. The author presents a seemingly independent third-person (authorial or extra- and heterodiegetic) narrator. In actual fact, however, the narrator’s perspective coincides with the subjective perspective of the main character (Hunter). This fusion of perspectives, although almost imperceptible to readers, helps them to develop empathetic resonance with the state of mind depicted in Hunter. The fusion of the third-person narrator with the most privileged narratorial voice maximizes both the impression made by Hunter on the reader and the transference that occurs. Reading about Hunter provides readers with the opportunity to activate effectively the trauma associations to which they have access because of their own biographical experiences. However, the fusion of the perspectives of character and narrator is by no means a necessary condition for this to take place. Something else is going on as well. The author, Judith Hermann, does not allow readers to know for sure at any point in the narrative that the perspectives of narrator and protagonist have merged; in this sense, she deceives the reader. By doing so, she brings about the transference of particular affects to the reader. Through her narrator, she makes the reader experience the affects of being in an illusory or unfaithful relationship. Furthermore, by deceiving the reader regarding crucial questions of self and other, she initiates a destructive double-binding relationship between reader and text / narrator of the very same kind as that which is an implicit part of Hunter’s previous history. As a result, the dependency-autonomy conflict depicted in Hunter is likely to be re-enacted between reader and text / narrator. In a way, then, the reader is offered a relationship with the same structure and, for the purposes of psychotraumatology, the same causes and effects as the relationships Hunter must have lived through (the implied biographical experience of his character) in order to reach the situation he is in when the narrator presents him to the reader. We cannot tell whether or to what extent the author does this consciously. It does not matter anyway, for, irrespective of the intentions behind it, the most likely outcome of the situation is that readers will be tempted to adopt Hunter’s frame of mind without even realizing that they are doing so. The reader, to put it in Alan Palmer’s terms, ‘constructs the fictional mind’ of the narrator or Hunter ‘as a continuous mind’ (see Palmer in this volume), and does so in a way that involves entering into a double-binding relationship with that mind. Thus, through the narrator, the author

252

Harald Weilnböck

is able to bring about (parallel) defence transference between herself and the reader. This means inducing in readers screen affects and screen associations of the same kind as those that characterize the emotional state depicted in Hunter. Consequently, readers are likely to fend off certain memories and feelings attached to their individual trauma associations, just as Hunter and the narrator avoid dealing with the associations of Hunter’s personal history. Several possible consequences for the reading process itself can be identified (this is done here for heuristic purposes only). First, readers will abandon critical distance in order to closely follow the evaluations and suggestions of the narrator, which ultimately means adopting Hunter’s perspective in each and every scene that involves him. For example, when the narrator shows Hunter getting angry with Leach, the owner of the hotel, for (in his view) not looking after his post properly and being a generally egotistic person (116f), readers believe Hunter and become angry like him. Another example occurs when Hunter states in the closing sentences of the story that the young woman is bound to leave the hotel early the next morning, taking with her the cassettes he gave to her as a present (119). Readers are likely to be convinced that this is what will happen. Finally, when the narrator shows Hunter enjoying his evening routine of listening to his music (137), readers will themselves slip into a melancholy mood and perhaps even think to themselves that this is the best possible way of listening to music. It follows that readers will not conceive of Hunter’s situation as a state of suffering at all. There is nothing to make them do so; Hunter clearly does not see himself as a suffering person in any way. The most importance consequence of this is that the reader does not even begin to reflect on what is perhaps the most intriguing aspect of the plot, the earlier biographical history and actual life experiences that brought about Hunter’s situation in the first place. When readers fail to reflect in this way, the defence strategy of the narrator (and perhaps even the author as well) has served its purpose successfully. Readers who forget to reflect in this manner do so due to the defensive transference brought about by the narrator, who seems determined not to tell us anything about Hunter’s previous history. In doing so, the narrator creates the conditions that allow readers to avoid realizing and sensing any trauma associations of their own that are touched on by the story. Hence, they will not come to terms with the traumatic experiences concerned. All empirical readers, of course, will perceive the text in their own individual subjective ways, but, as individual subjective perception is never completely isolated or different from that of other individuals, we can still assume that there is a limited number of probable reader responses. When faced with Hunter


253

as depicted in the narrative, readers have, one way or another, to decide in their hearts and minds whether to accept these transferences and whether and when to escape from and / or come to terms with them. Having ascertained that there is a narrative device able to cause defensive parallel transference, we can now focus on the second kind of transference at work in “Hunter-Tompson-Musik”. While the narrator collaborates, as it were, with the main character for reasons of psychological defence, it can be shown that the author has built a small number of inconsistencies, however subtle, into the narrator’s perspective. When these inconsistencies appear in the narrative, they give readers the chance to realize that the perspectives of narrator and protagonist have merged. If readers recognize these subtle clues in the narrative, they can transcend the (parallel) transference directed at them and enter into a process of complementary transference instead. They can loosen the grip of the screen affects and become capable of feeling with empathy the affects and associations that the narrator tries to keep at bay, that the narrative transfers all the more intensely to readers, and that readers defend themselves against when dealing with their own trauma associations. In summary, the narrator strives to implement (parallel) defensive transference, but the author of the text, in giving the reader a certain amount of opportunity to detect inconsistencies in the narrator’s perspective, creates a space in which complementary transference can take place, thereby supporting the therapeutic process of coming to terms with trauma associations, defence mechanisms, and defensive transferences. Determining the probability of such complementary transference taking place in practice (in other words, calculating the percentage of a group of Judith Hermann readers that we would expect to reach the stage of coming to terms with their screen affects) is a somewhat complex qualitative empirical problem. As far as Hermann’s text is concerned, it boils down to the question of how effective or recognizable the breaks in the narrator’s perspective are (or how easy it is for readers to overlook them). The breaks are extremely subtle, so subtle that we cannot be absolutely sure that they were consciously constructed by the author. The least subtle of the three breaks I have identified occurs when Hunter gets angry with Leach, the hotel owner, because he does not want to go and check whether Hunter has received any post (116f). Such behaviour is, indeed, rather impolite and annoying, but our view of the situation changes considerably when we note that it is implied in another scene that Hunter can see from the counter whether or not he has received any post. But this is not explicitly stated; it takes a highly analytic reader to notice the inconsistency. It is difficult to answer the qualitative empirical question of the likelihood of readers initiating psychological processes in which they come

254

Harald Weilnböck

to terms with their affects. Difficult—but not, the literary theorist in me says, impossible. We have simply to recognize that the best way of dealing with the question lies in exchanging textual analysis for a multidisciplinary research context of which qualitative reader-response studies are a part and in which methods from the social sciences are used. In the last few years, I have participated in the development of two research frameworks for reader-response studies, one involving individual interviews and the other group interviews. The former developed out of a Deutsche Forschungsgemeinschaft (German Research Foundation) project (1999, 2003a), the latter I derived from the context of a group-analytic training (Weilnböck 2002a, 2002b, 2003b). The reader responses that I introduced above as heuristic concepts were actually observed in a group analysis of Hermann’s text. Some of these responses turned out to be of central importance for understanding the text and the analysis session itself. It would be desirable, I would argue, to integrate such methods into our research, even though we can also make considerable progress without going beyond textual analysis—what I have said here, for instance, is based entirely on textual analysis.

4. Interactional Narratology Having taken Judith Hermann’s text as a concrete example of how the concepts of psychotraumatology can be applied to the study of a literary narrative, it is perhaps appropriate to take a step back and ask how this interaction-based perspective on aesthetic products and media output relates to literary narratology proper. In more general terms: what specific difficulties arise when we try to create an interdisciplinary narratology? Postulating that interaction-related functions and transferences connect a narrative text or film to its recipients clearly takes us beyond the world of descriptive systems that classify types of narrator and situations of narration on the basis of purely formal criteria. On the functional level, all narrators are different. One way of analysing Hermann’s narrative in terms of formal description would be to employ Wayne C. Booth’s concept of the unreliable narrator in order to describe its narrator systematically. Additional assistance can be found in the introduction to literary narratology by Martinez and Scheffel. Drawing on Martínez-Bonati’s distinction between mimetic and theoretical sentences in the words of the narrator, Martinez and Scheffel (1999:99–100) make the following comment in parentheses: ‘we shall also treat cases in which the narrator provides subjectively coloured evaluations as instances of theoretical unreliability [unreliability that occurs in the narrator’s more abstract and


255

philosophical statements].’18 This draws attention to the most important interaction-related function of texts (both for psychotraumatology and in a general sense). Just as in face-to-face narrative interaction, it is this kind of subjective colouring that allows a textual author / narrator to convey a narrative and / or interact with a certain motive, process experiences of conflict-ridden and / or psychotraumatic interaction, and perhaps try to re-enact such conflicts on the level of transference between text and reader. Martinez and Scheffel, though, touch on the subjective colouring of narratives in an aside, as if it had nearly been forgotten and was not considered of sufficient narratological importance to merit systematic treatment. Nonetheless, the introduction by Martinez and Scheffel stands out because it tries to point the way towards an interdisciplinary approach to narratology and explicitly attempts to go beyond purely descriptive systems for analysing narratological phenomena: ‘unlike Stanzel’s closed typological circle, which aims to subsume an imaginary totality of types, we present a descriptive model designed to operate with an inventory of freely combinable features (to which more can always be added) rather than a limited number of typical forms of narration’ (Martinez / Scheffel 1999:94). Putting this into practice, Martinez and Scheffel provide a chart of different narrative situations based on a set of three distinctions: hetero- / homodiegetic, extra- / intradiegetic, and (for focalization) author / character / neutral. Certain combinations of these features produce types that cannot be illustrated with concrete examples in the form of complete narrative texts (Martinez / Scheffel 1999:95). The system of Martinez and Scheffel is clearly an open one (in the sense that further features could be added to produce further types) but it is still based on forms rather than interaction-related patterns and functions. It covers criteria of time (order, duration, and frequency), mood (distance and focalization), and voice (time of narration, level of embedding, involvement of the narrator in events, and the subject and addressee of narration). Although the consideration of subject and addressee nominally reflects the nature of narration as a process of interaction, the system does not offer a way of describing or explaining the functions of narrative interaction. Any approach based on an inventory of features (even if an extendable one) and lacking functional criteria will, by its very nature, be limited to describing labels for formal types. If we are to adopt an interdisciplinary approach 18

An essential difference between text-based and interaction-based theories may lie behind the use of the word ‘sentences’ by Martinez and Scheffel, which implies written language printed on paper. I suggest using the term ‘utterances’ instead in order to reflect the mental and interaction-related nature of narration processes.

256

Harald Weilnböck

to narratology, we must assemble criteria for defining the affect-related psychological functions of narrative interaction. Although they fail to adopt such a functional approach, Martinez and Scheffel do identify different degrees of narratorial involvement in the narrated events (Martinez / Scheffel 1999:82). Following Lanser, their scale extends from the third-person (heterodiegetic) to the first-person (homodiegetic) narrator and includes a total of six types of narrator: (1) the uninvolved narrator, (2) the uninvolved eyewitness, (3) the witnessprotagonist, (4) the minor character, (5) the co-protagonist, and (6) the sole protagonist. The concept of narratorial involvement, indeed, would make an ideal functional / interaction-related criterion (as would the idea of subjective colouring). This would, of course, require the formulation of specific criteria with which to measure involvement, something that Martinez and Scheffel do not attempt. Also, the third-person / heterodiegetic narrator is not fully integrated into the scale. Identified with the uninvolved narrator, it constitutes one extreme of the scale; the other five types are all variants of the first-person / homodiegetic narrator. This is perfectly acceptable as far as the internal structure of the text is concerned, for a third-person narrator is not normally part of the narrated world and thus cannot normally have the same position relative to the characters as a first-person / homodiegetic narrator. When we study interaction outside the text, however, we concentrate not on the narrator’s formal position relative to the characters but on the narrator’s involvement with the reader implicitly addressed by the narrative. In this context, it is indisputable that a heterodiegetic / third-person narrator can be a highly involved narrator (i.e. the implied psychic situation of the narrator can be highly involved affectively with the events and interactionally with the addressee), clear that we must be prepared for different levels of involvement when dealing with first-person and third-person narrators alike, and clear above all that we should expect different affect-related psychological situations in each individual case. Interaction-related criteria for measuring narrative involvement are not defined in the models of Lanser and Martinez and Scheffel, nor is adequate attention given to the third-person narrator. Such models cover only the relation between the narrator and the narrated world and its characters. They do not account for the different types and functions of interaction between a narrator / text and addressees / readers (let alone the mediated interaction of author and reader), and if they do consider the relation between the two sides, it is implicitly treated as one that does not involve interaction. Their genuine desire to open the way to an interdisciplinary approach to narratology means that Martinez and Scheffel (1999:84–89) give consideration to the subject and addressee of narration. They observe tellingly


257

that ‘research in literary theory has previously focused on the role of the fictive narrator or the role of the reader and neglected the interaction of the two sides of the communication situation created in a fictional narrative’ (Martinez / Scheffel 1999:85; somewhat surprisingly, no mention is made of the author here). They go on to say that an attempt to rectify this imbalance can be found in recent work that approaches literary theory from the perspective of social history and considers the communicative function of texts. Martinez and Scheffel, however, do not refer to any specific examples (they proceed to discuss historical ways in which narrators have explicitly addressed their readers). This leaves us with the wide field of social history in general to consider. We find research on the macrosociological, ideological, and demographic features of historical epochs and audiences. But, interesting as it may be to learn about the macrosociological features of historical audiences, it does little to help us identify the affect-related psychological functions that influenced how readers read and interacted with a text; the question posed by Martinez and Scheffel (1999:84), ‘who narrates to whom?’, cannot be approached effectively in this way. The fact that Martinez and Scheffel (1999:85) assert that social history can provide a way to complement their primarily intratextual approach to narratology with a means of approaching the issue of text-reader interaction shows us just how little we know about the mental and affect-related psychological aspects of how people interact with texts. To their credit, Martinez and Scheffel do give extensive consideration to narratological models employed in disciplines other than literary theory, albeit in a chapter on “Further Horizons” at the end of the book (Martinez / Scheffel 1999:145–159). Their introduction marks the limit of the extent to which narratology is prepared to go in becoming interdisciplinary at this point in time—a truly interdisciplinary integration of the intratextual and interaction-based perspectives remains a task for the future.

5. The Difficulties of Becoming Interdisciplinary It was the objective of the Hamburg Conference on Narratology beyond Literary Criticism to rectify this shortcoming, and the contributions from researchers working on various narratological issues (narratology being treated as a broad, inclusive discipline) certainly provoked inspired, intense, and productive discussion. However, it seems to me that throughout the two days of the conference, we completely failed to talk and think about just how difficult the task confronting us is, about how challenging, sometimes even nerve-wrecking, it can be to go beyond the familiar and build interdisciplinary bridges. Each of us, trained in a particular academic

258

Harald Weilnböck

field, is profoundly moulded by her or his home discipline. As our home disciplines are fully-developed institutions with little inclination to encourage forays beyond their borders, it is up to their individual members to make the effort to venture into unfamiliar territory. The representatives from outside the Anglo-American countries must have felt this more than most, for they work in environments where the borders between academic disciplines generally seem more rigid than they are elsewhere. Becoming interdisciplinary (and international), then, is by its very nature a process of conflict in which we are torn between wanting to step into the unknown and, more or less consciously, wanting to shy away from leaving the secure ground of our academic home disciplines. Our position is, as it were, an ambivalently (anti-)interdisciplinary one. In his famous and much-quoted book Anxiety and Method, Georges Devereux was, as far as I am aware, the first person to highlight the extent to which hermeneutic disciplines are troubled by the fact that the researching subject is a subjective one. The researching subject is constantly plagued by subjectivism and thus longs to obtain a (self-)reflexive external view of its observations and conclusions. Devereux also emphasized that the necessity of (self-)reflection, which has long been known in philosophical hermeneutics, albeit mostly as a theoretical problem in the form of the hermeneutic circle, is not just an intellectual issue; it is also, and primarily, an issue of affects and concerns the affect-related psychological structures and patterns behind the habits of the individuals and institutions that engage in scientific and critical activity. The title of his book itself shows that Devereux sees methodological issues (the selection or rejection of certain methods and theories) not just as a matter of epistemological reasoning but primarily as the expression of the affect-related psychological anxieties, needs, and interests of the subjects involved in performing research.19 One source of anxiety lies in the fact that reconstructive / hermeneutic research always means trying to obtain knowledge about oneself as a subject in relation to the objects being studied. We have since learnt (and will no doubt continue to learn) that such methodological anxieties can hinder rather than encourage the process of acquiring knowledge in the humanities. Ultimately, they can even subvert what may have been intended as taking a step toward an interdisciplinary approach using multiple methodologies. We all know that if anxiety is not consciously accepted and dealt with as such, it tends to turn into aggression that generally triggers psychological defence mechanisms. In any case, one’s field of view narrows, which is not a particularly productive epistemological consequence. 19

The ideas of Habermas on the interdependence of insight and interests should be recalled here.


259

Moreover, as Stavros Mentzos has since pointed out, this kind of semior unconscious process of affect defence occurs in both individuals and institutions / structures. Institutions too have habits and behavioural patterns, that is, they have mechanisms of (inter)action and defence, and they foster mentalities.20 Thus, engaging in interdisciplinary work in the social sciences after Devereux (whose book was first published as early as 1973) and Mentzos means engaging with the anxieties that methods and changes of method can arouse in individuals and institutions. It also depends on our being prepared to focus self-reflexively on the process by which the subject carries out research on objects, with the result that the research process itself and its institutional administration become objects of metaresearch. In the humanities, the study of literature (the home of text-based theoretical approaches) in particular stands out for its inexperience in considering and implementing measures of (self-)reflexive scrutiny and regulation. Thus, when students of literature engage in interdisciplinary projects that also involve the social and psychological sciences, it is all the more important to identify (self-)reflexive procedures that fit their particular needs. From the perspective of the qualitative social sciences, which are more familiar with these procedures and have developed methods of narrative (group) interviewing,21 one step that could be taken is immediately apparent: undertaking a reconstructive close reading of critical interaction as it unfolds in real-life discussions between colleagues on interdisciplinary research in narratology. In addition, our theory would benefit because we would have the opportunity to consider whether narration in institutional interaction and narrative formation can be seen as a process that not only transcends the text or medium but also extends beyond individual subjects in some way, being generated by plural entities, groups, and discourses in group interaction and group narration.22 At the end of the two-day conference in Hamburg, it seemed to me that (anti-)interdisciplinary anxieties had played a role at various points in the debate. Thus, I felt encouraged in attempting a provisional close reading

20

21

22

See Mentzos (1988) on the concepts of institutional defence and the institutional unconscious. For methods of narrative interviewing see Fischer-Rosenthal/Rosenthal 1997; for methods of group interviewing see Bohnsack 2000. This introduces an area of research that has been ignored by narratologists and has not received much attention to date in the study of literature in general: group analysis and qualitative social research on the interaction and narration found in and performed by groups, networks, and institutions in which interaction occurs (see Tschuschke 2000, and Haubl/Lamott 1994 on group-analysis; on narrative group interviewing in the context of media studies see Weilnböck 2002a, 2002b, 2003b).

260

Harald Weilnböck

of one particular sequence of interaction.23 In general, I had two impressions. First, in spite of the desire to broaden its horizons, narratology, having developed out of the study of literature, is still somewhat inclined to confine the scope of its study to aesthetic and / or textual artefacts and treat social and psychological aspects of narration as belonging to an area of research distinct from narratology proper. Second, narratology appears to have a certain tendency to see itself as a discipline for others to draw on, and does not reciprocate in the sense that it would give systematic consideration to the possibility of looking for useful theoretical imports from other disciplines concerned with narration. I cannot say for certain whether these impressions are correct. In any case, I believe it would be worthwhile giving further attention to and systematically studying institutional (self-)reflection and the regulation of methodological processes. The highest level of institutional interaction, in Germany the Deutsche Forschungsgemeinschaft (German Research Foundation, or DFG), cannot be exempted from such an analysis. This is clear from the curious fact that the DFG, which until a few years ago attached considerable importance to encouraging and supporting interdisciplinary projects, has more recently made major funding decisions that seem to point in exactly the opposite direction. This top-level institutional phenomenon has already received a name: it is generally referred to as a Rephilologisierung (return to philological substance of the humanities). The DFG itself, however, does not appear to be paying much attention to the wider and long-term consequences of this (unarticulated) policy, let alone taking action to prevent interdisciplinary humanities from disappearing. What we do not know and need to understand better is how such a situation comes about. How do the individual and group (institutional) decision-making processes work? How do decisions relate to the objectives stated by those who make them? Are these processes subject to any procedures for performing (self-)regulation and preserving appropriate standards, and if so, how do such procedures function? In narratological terms, this amounts to asking what structures of interaction lie behind and what actions result from the institution’s narrative about interdisciplinarity. Knowing more about all this is desirable in the long run because there is one future development in the humanities that is not hard to predict: interdisciplinary research will grow. In part, this will take place in defiance of institutional resistance and our own individual uncertainties; it will take place, that is, by working with and making us reflect on those uncertainties, as my psychological self says to my literary self, and vice versa. It will take 23

The close reading could not be included in this article and will be published separately (see note 1).


261

place because interdisciplinary research is not only worthwhile and promising in itself but also provides us with the best possible tools for approaching crucial social questions that by their very nature involve interaction (e.g. the interaction of individuals and groups / societies in culture and the media) and interdisciplinarity. Moreover, these social questions are increasingly being given high priority by international research-funding organizations such as the European Union.24 So, it will happen! Interdisciplinary research will become an increasingly central aspect of the humanities of the future. The question is when and where this will happen. Being a literature studies person myself, I very much hope that it will not happen outside the field of literature. Psychology can probably manage without the narratology developed in the study of literature;25 there is always art itself if we want to find a good poem or drawing with which to illustrate a psychological concept such as dissociation.26 But the narratology developed in the study of literature, I believe, should not succumb to the temptation of isolating itself from sociology and psychology, for it will then be less well-equipped to help solve the social questions mentioned above. Instead, the study of literature should play an active role in revealing the enormous interdisciplinary potential of narratology, a potential evident when we consider that there can be few theoretical concepts that appear in as many fields of research as narration does. If this is to happen, it would appear that the study of literature must go beyond the realm of textual theory and develop an additional, interaction-based theory that focuses on how human minds interact in the writing and reading of texts. Returning to the basic definition of narration provided by Martinez and Scheffel (1999:9), we might say that such a narratology would deal not only with the ‘certain something’ of narration but also with its ‘someone’ and its ‘someone else’. 24

25

26

The European Union’s research-funding system is organized on the basis of general themes concerning the present technological and social state of the society (or societies) of Europe. There are no quotas for particular fields of research. The only category open to the humanities is entitled ‘Citizens and Governance in Knowledge-based Societies’. When investigating the FP6-2002-Mobility-5 programme (the 2003 Marie Curie IntraEuropean Fellowships with a budget of € 110,000,000 for 2003–04), I was unable to find a single literature-related project among those funded in previous years (see ). When psychologists have considered language-related questions (as I believe they should do more often, making greater use of the study of literature as a stimulus and source of ideas), they have shown themselves capable of performing their own language studies; for a recent example, see Searles 2000:427–444. In the discussion at the conference, I pointed out that the literary concept of metalepsis (discussed by Werner Wolf in his article in this book) corresponds to the psychological concept of dissociation. I have since realized that a not inconsiderable number of psychologists working on dissociation use metaleptic paintings and drawings for illustrative purposes. This is one way in which an interdisciplinary bridge could be built between narratology and psychology.

262

Harald Weilnböck

References Bohleber / Drews 2001 Bohleber, Werner / Drews, Sibylle (eds.): Die Gegenwart der Psychoanalyse – die Psychoanalyse der Gegenwart. Stuttgart 2001. Bohnsack 2000 Bohnsack, Ralf: “Gruppendiskussion.” In: Flick, Uwe / Kardorff, Ernst von / Steinke, Ines (eds.) Qualitative Forschung. Ein Handbuch. Reinbek 2000: 369–336. Charlton / Schneider 1997 Charlton, Michael / Schneider, Silvia (eds.): Rezeptionsforschung. Theorien und Untersuchungen zum Umgang mit Massenmedien. Opladen 1997. Devereux 1973 Devereux, Georges: Angst und Methode in den Verhaltenswissenschaften. Frankfurt a.M. 1973. Dornes 1993 Dornes, Martin: Der kompetente Säugling. Die präverbale Entwicklung des Menschen. Frankfurt a.M. 1993. Fischer / Riedesser 1998 Fischer, Gottfried / Riedesser, Peter: Lehrbuch der Psychotraumatologie. München 1998. Fischer-Rosenthal / Rosenthal 1997 Fischer-Rosenthal, Wolfram / Rosenthal, Gabriele: “Narrationsanalyse biographischer Selbstrepräsentation.” In: Hitzler, Ronald / Honer, Anne (eds.): Sozialwissenschaftliche Hermeneutik. Eine Einführung. Opladen 1997: 133–65. Geißler 2002 Geißler, Peter: “Psychoanalyse und Körper. Überlegungen zum gegenwärtigen Stand analytischer Körperpsychotherapie.” In: Psychoanalyse und Körper 1, 2002:37–85. Habermas 1968 Habermas, Jürgen: Erkenntnis und Interesse. Frankfurt a.M. 1968. Haubl / Lamott 1994 Haubl, Rolf / Lamott, Franziska (eds.): Handbuch Gruppenanalyse. Berlin 1994. Hermann 1998 Hermann, Judith: “Hunter-Tompson-Musik.” In: Sommerhaus, später. Frankfurt a.M. 1998:115–138. Hirsch 1998 Hirsch, Mathias (ed.): Der eigene Körper als Objekt. Zur Psychodynamik selbstdestruktiven Körperagierens. Gießen 1998. Iser 1975 Iser, Wolfgang: “Die Appellstruktur der Texte. Unbestimmtheit als Wirkungsbedingtheit literarischer Prosa.” In: Warning, Rainer (ed.): Rezeptionsästhetik. Theorie und Praxis. München 1975. Kernberg / Dulz / Sachsse 2000 Kernberg, Otto F. / Dulz, Birger / Sachsse, Ulrich (eds.): Handbuch der Borderline-Störungen. Stuttgart 2000. Kohut 1973 Kohut, Heinz: Narzissmus. Eine Theorie der psychoanalytischen Behandlung narzisstischer Persönlichkeitsstörungen. Frankfurt a.M. 1973.


263

Martinez / Scheffel 1999 Martinez, Matias / Scheffel Michael: Einführung in die Erzähltheorie. München 1999. Mauser / Pietzcker 2000 Mauser, Wolfram / Pietzcker, Carl (eds.): Trauma. Jahrbuch für Literatur und Psychoanalyse 19. Würzburg 2000. Mentzos 1988 Mentzos, Stavros: Interpersonale und institutionalisierte Abwehr. Frankfurt a.M. 1988. Mertens / Waldvogel 2000 Mertens, Wolfgang / Waldvogel, Bruno (eds.): Handbuch der psychoanalytischen Grundbegriffe. Stuttgart 2000. Mitchell 1988 Mitchell, Stephen A.: Relational Concepts in Psychoanalysis. Cambridge 1988. Pantoja 2001 Pantoja, Andréa P. F.: “A Narrative-Developmental Approach to Early Emotions.” In: Forum: Qualitative Social Research, 2001. Pietzcker 1992 Pietzcker, Carl: Lesend interpretieren. Zur psychoanalytischen Deutung literarischer Texte. Würzburg 1992. Racker 1978 Racker, Heinrich: Übertragung und Gegenübertragung. München 1978. Raguse 1994 Raguse, Hartmut: Der Raum des Textes. Elemente einer transdisziplinären theologischen Hermeneutik. Stuttgart 1994. Rohde-Dachser 2000 Rohde-Dachser, Christa: Das Borderline-Syndrom. Stuttgart 2000. Searles 2000 Searles, Harold F.: “Hinweise auf eine Borderline-Psychopathologie durch (a) Pausen und (b) Satzbaustörungen in der Sprache des Patienten.” In: Kernberg / Dulz / Sachsse 2000:427–444. Tschuschke 2001 Tschuschke, Volker (ed.): Praxis der Gruppenpsychotherapie. Stuttgart 2001. Weilnböck / Garbe / Schoett 1999 Weilnböck, Harald / Garbe, Christine / Schoett, Silja: “Narrative Interviews und rekonstruktive Fallanalyse in der medienbiografischen Forschung.” In: Groeben, Norbert (ed.): Lesesozialisation in der Mediengesellschaft. IASLSonderheft zum DFG-Forschungsschwerpunkt. Tübingen 1999:218–232. Weilnböck 2001 Weilnböck, Harald: “Psychotraumatologie. Über ein neues Paradigma für Psychotherapie und Kulturwissenschaften.” In: literaturkritik.de, 2001. Weilnböck 2002a Weilnböck, Harald: “Die Anwendung der Gruppenanalyse in der Kulturvermittlung. Trauer / -Abwehrarbeit in einer Sitzung des ‘Gruppenanalytischen Literaturseminars’ über Judith Hermanns ‘Hunter-Tompson-Musik’.” In: Gruppenanalytische Arbeitshefte 28, 2002:49–74.

264

Harald Weilnböck

Weilnböck 2002b Weilnböck, Harald: “‘Dann bricht sie in Tränen aus.’ Übertragungen von Trauer / -Abwehr im Text und im ‘Gruppenanalytischen Literaturseminar’ über Judith Hermanns ‘Hunter-Tompson-Musik’.” In: Trauer. Jahrbuch für Literatur und Psychoanalyse 22, 2002:241–261. Weilnböck 2003a Weilnböck, Harald: “Leila: Dissoziative (Medien-)Interaktion und Lebensweg einer jungen Erwachsenen. Eine (medien-)biografische und psychotraumatologische Fallstudie.” In: Forum: Qualitative Social Research, . Weilnböck 2003b Weilnböck, Harald: “Das ‘Gruppenanalytische Literaturseminar.’ Zur Anwendung der Gruppenanalyse in der Kulturvermittlung. Mit neuen Aspekten zur Interpretation von Heiner Müllers Prosatext ‘Vater’.” In: Gruppenanalyse 1, 2003:63–85. Weilnböck 2004a [forthcoming] Weilnböck, Harald: “‘[…] dazu passend: Rotwein mit Eierkognak zur Hälfte in einem bauchigen Glas.’ Borderline literarische Interaktion und Gewalt am Beispiel von Ernst Jüngers Kriegsschriften.” In: Hagestedt, Lutz (ed.): Ernst Jünger: Politik – Mythos – Kunst. Berlin / New York 2004. Weilnböck 2004b [forthcoming] Weilnböck, Harald: “Zur dissoziativen Intellektualität in der ‘Nachkriegszeit’. Historisch-psychotraumatologische Überlegungen zu Metapher / Metonymie und Assoziation / Dissoziation bei kritischen, neukonservativen und postmodernen Autoren.” In: Seidel, Günter / Eckart, Wolfgang (eds.): Historische Traumaforschung. Gießen 2004. Weilnböck 2004c Weilnböck, Harald: “Claude Lanzmanns ‘Shoah’ und James Molls ‘Die letzten Tage’. Psychotraumatologische Analysen von Bearbeitungen der Shoah im Film.” In: Schmitz, Walter (ed.): Erinnerte Shoah – Die Literatur der Überlebenden. Dresden 2004:444–494. Weilnböck 2005 [forthcoming] Weilnböck, Harald: Borderline literarische Interaktion am Beispiel Ernst Jüngers. Vols. 1 and 2 (Habilitation thesis, Berlin), Würzburg.

The Eye of the Beholder Narratology as Seen by Social Psychology Wolfgang Kraus (München) 1. Introduction At first sight, a fundamental difference between social science and narratology meets the eye: social science concerns people. It concerns the homo narrans, a very special animal that produces narratives. Narratology, on the other hand, concerns narratives, the stories produced by this story-telling animal. Social science takes individuals (positioned in a social context) as its starting point; narratology provides a way to deal with stories, something produced by (and for) the homo narrans through the practice of storytelling. Both fields have undergone fundamental changes, and many links have developed between them, but the differences between their underlying conceptions can still be felt today. Thus, there is good reason to examine the relationship between these two distinct fields of enquiry. Interest in narrative approaches has grown steadily in the increasingly diverse field of social science in recent decades. Hevern’s analysis (2003) of the PsycINFO citation index illustrates this: he found that the number of citations in which the terms ‘narrative’ or ‘narration’ are used grew by a surprising 2569 percent between 1960 and 2001. The last ten years, in particular, have seen a massive growth in interest in narrative. Successful as narrative may have been in securing a place in the social sciences, it is not immediately obvious what social psychologists are trying to achieve when they incorporate narrative concepts into their work. Indeed, there may even be some suspicion among narratologists that their concepts could be misused in such a way that much of their work on refining models and terms would be undone. Certainly, the number of citations alone says nothing about the state of narrative concepts when they are used in social psychology. Nor does it provide any indication of whether social psychologists follow the guidelines provided by narratology for the use of such concepts.

266

Wolfgang Kraus

From the perspective of social psychology, the importing discipline, the success of the narrative approach is quite remarkable and calls for explanation: ‘Why, at this particular juncture, is there so much interest in narrative?’ (Freeman 1998:45). Freeman seeks to identify more than causes; he is concerned with the historical place of this development, with understanding the current state of society. It is easy to see that use of narratological tools in the social sciences is increasing; in the following article, I hope to follow Freeman’s example by considering this development in a critical light and putting it in a historical perspective.

2. Positioning the Inquirer: Social Psychology’s View of Narratology In my search for an answer to Freeman’s question, I shall restrict my argument to the field of identity theory. This may seem to mean adopting a very narrow perspective, but it is justified by the fact that identity theory is a particularly dynamic area of social psychology. Furthermore, the rise of the concepts of narrative identity and self-narrative means that it is a field in which the narrative approach is well and undeniably established today. Thus, identity research is a prominent contributor to the rise of narrative reflected in the PsychINFO database. Classical approaches to identity in social psychology take as their starting point the idea that it is in adolescence that people develop an idea of who they are and who they want to become in their social worlds. So, identity is about the self as a continuous self in time (the relatively stable part) as well as about developing a personal project (the dynamic part). Ricœur (1990:140) has distinguished between these aspects of identity as idem as sameness and ipse as selfhood. Classical identity theory has been questioned recently because of findings which suggest fundamental changes in this identity system. The question is how to conceive of identity in a rapidly changing world and how to develop a model of the individual that does not presuppose some solid inner core but conceives of the development of the individual as a construct unfolding in the world in—and through—its symbolic representations. The current situation in identity theory provides a way to help us distinguish the various perspectives that can be adopted towards narratology and the various positions on which they are based. We find a general consensus among identity theorists that identity should be modelled as a process. This process has an agent and is realized in one way or another, which means that there is an outcome, a product of the process, fragile and provisional as it may be. And the process clearly takes place in a social world. If we juxtapose this with narratology, we can say that the latter

Social Psychology’s View

267

focuses on narrativity. Narrativity becomes recognizable—apparent—as a result of the act of narrating performed by a narrator, who thereby creates a fictional world in which his or her narrative is situated. Social psychology focuses on

Narratology focuses on

identity, which is a

narrativity, which is perceptible in

process associated with an agent taking place in a

Ù ?

telling performed by a narrator, creating a

social world, and analyses its

story world, and leading to

constructs

narratives

Table 1: Identity Theory and Narratology: A Superficial Affinity

In my arrangement of table 1, I have established an affinity, justifiable at first sight at least, between conceptual levels. It relates similar levels of investigation. On a general level, for instance, we may ask: what is the formal relationship between the two disciplines—social psychology and narratology—or between basic concepts—such as processes, agents, and so on—within them? When looking closer at the various levels, however, we are confronted with many questions. Are we suggesting that identity is replaced by narrativity, that common ground exists, that hierarchical relationships are involved? Are some relationships more important than others; is there more affinity on the level of narrativity than, for instance, on that of agency? There are many ways to approach such questions; I shall adopt a functional perspective. The functionality of any given narrative approach in social psychology, however, depends on the theoretical position in identity theory from which we analyse it. 3. The Quest of Identity Theory In order to understand what identity theory seeks in its approach to narratology, it is important to identify the underlying conceptual needs involved. Interest in narratology results from a search for ways to answer certain questions—identity theory employs narratology not for its own sake but

268

Wolfgang Kraus

as a source of assistance in dealing with fundamental changes in the construction of identity in modern societies. There may be general agreement on the basic facts, but the same can hardly be said of the way in which these changes are understood and conceptualized. Identity theory displays a general tendency, fuelled by historical sociology and contemporary social analysis, to criticize cognitivist models (cf. Bruner 1990). They are held to be ahistorical, ideological, and focused on coherence, and, it is argued, they fail to give proper consideration to historical changes and personal meaning-making. Historical development is no more than a content matter for such theories; it does not enter into their formal structure. Contrary to such models, current theories highlight the relevance and historical specificity of formal aspects and the importance of personal meaning-making processes (Keupp et al. 2002). Identity construction today, critics insist, is marked by fragmentation and diffusion, the lack—or at least decline—of coherence, the disappearance, as Lyotard has observed, of grand narratives. This is closely connected with individualization and changes in society that have been discussed under the rubrics of late modernity (Wagner 1994) and post-modernity (Gergen 1991). At the present time, identity theory is, generally speaking, concerned with the constructability of identities. The expanding scope of identity construction has characterized the transitions from one social configuration of modernity to another. ‘These transitions entail social processes of disembedding and provoke transformations of social identities, in the course of which not only other identities are acquired but the possibility of construction is also more widely perceived’ (Wagner 1994:157). Individualization processes are marked by the fact that the construction of coherence is no longer guaranteed by membership of large social groups (e.g. the Church, trade unions, social milieus or classes). The task of creating coherence is very much the responsibility of the individual person. The development of identity thus becomes a process, a never-ending story, always open for change. It is no longer a matter of constructing and realizing one’s personal project. Instead, it implies constantly rearranging and reframing oneself, testing and negotiating one’s limits. As illustrated by the main character in Martin Amis’s Rachel Papers, these acts of construction are performed in a situational context. What clothes would I wear? Blue madras suit, black boots, or the old black cord suit with those touching leather elbow-patches. What persona would I wear? On the two occasions I had seen her last August I underwent several complete identity-reorganisations, settling finally somewhere between the pained, laconic, inscrutable type and the knowing, garrulous, cynical, laugh a minute, yet something demonic about him, something nihilistic, muted deathwish type. Revamp those, or start again? (Amis 1973:45)


269

Although the importance of individual requirements for a model can be debated, table 2 is a reasonable general portrayal of some of the more important questions being explored in current research. Individual theoretical decisions notwithstanding, any identity theory will be required to provide answers to this set of questions. What is needed, then, is a model of identity construction dealing with identity as a process. It must also allow for the possibility of a self open to multivoicedness. Such a self presupposes many selves and is not to be understood as a single closed unity. Its multiplicity demands a special exploration of agency and the self-other relationship. Furthermore, the approaches of identity politics have made it clear that the self-other relationship should be considered in terms of the construction of difference and power. As Sampson (1993) has shown, such strategies consist of the devaluation of the other and the construction of serviceable others to sustain power relationships. Identity thus becomes a reflexive project, open to self-analysis and self-scrutiny, as Montaigne demonstrated long ago. Identity theory is concerned with the individual’s way of creating meaning, of giving form and reason to a life. Identity theory must answer questions regarding – – – – – – – –

process-orientation multivoicedness how individuals create meaning the self-other-relationship power/positioning community narratives cultural difference reflexivity

Table 2: Identity Theory in Late Modernity: Model Requirements

4. The Narratological Option The starting point for a deconstructive identity theory of this kind lies primarily not in narratives as closed units but in narrativity as a multifaceted resource for the understanding of self-construction. There is an obvious similarity with certain changes in narratology which have been described as manifestations of deconstructivism and post-structuralism and have led to an extraordinary diversification that has produced a multitude of narratologies (cf. Nünning / Nünning 2002). Social psychologists talk about a narrative turn when describing the reorientation they witnessed in the 1980s and 1990s, but narratologists use more dramatic metaphors to

270

Wolfgang Kraus

describe the resurgence of their field of study. Currie, for instance, writes of a narratological ‘phoenix coming out of the ashes’. He lists three features of this resurrection: a progression from discovery to invention, from coherence to complexity, and from poetics to politics. According to Currie, the first change (that from discovery to invention) reflects a general move away from the assumption that narratology is an objective science that uncovers inherent formal and structural properties in the narratives it studies. In short, poststructuralists moved away from the treatment of narratives (and the language system in general) as buildings, as solid objects in the world, towards the view that narratives were narratological inventions construable in an almost infinite number of ways. Structure came to be seen as a metaphor used by readers of a structuralist bent to give the impression of stability in the object-narrative meaning (Currie 1998:2).

This view was accompanied by a post-structuralist preference for terms which highlight the active role of the reader in the construction of meaning and terms ‘which challenge the idea that a narrative is a stable structure by borrowing their metaphors from the semantic field of movement, like process, becoming, play, difference, slippage and dissemination’ (Currie 1998:3). The similarity with the concerns of identity theory is obvious. Terms like ‘openness’ and ‘movement’ indicate the presence of common ground between the two projects. Currie’s second shift (that from coherence to complexity) intersects with identity theory’s interest in the question of coherence. Post-structuralist narratology acknowledges the contradictory aspects of narratives, preserves their complexity, and resists the temptation to reduce them to stable meanings or coherent projects. According to traditional critics, the most profound hidden design in a narrative was its unity. They believed that exposing that unity revealed the formal, thematic, or even polemical coherence of a work. This critical quest for unity can be read as a desire to present narrative as a coherent and stable project. In the view of the post-structuralist critic, however, this is no more than a way of reducing the complex heterogeneity of a narrative. Deconstructed narratology, then, aims at a less reductive way of reading, one that is not guided by notions such as the coherence of the authorial project. This deconstruction of narratology is closely connected with the diversification of narratology, since deconstruction ‘is no respecter of boundaries, least of all the boundary between literature and the real world’ (Currie 1998:3f.). Currie’s third change, finally, captures the transition from poetics to politics. It can also be seen as part of the deconstructive legacy, for deconstruction introduced new ways of exposing ideology. Concentration on


271

binary oppositions, for example, was a key part of the apolitical tradition of structuralism, but became more political in the hands of some poststructuralists. The post-structuralist critic often retains the old obsession with binary oppositions, but tends to view them as an unstable basis for meaning and as a place where the values and hidden ideologies of a text are inscribed. Even if this approach to aporetic oppositions has not always been presented in political terms, part of the legacy of deconstruction consists of providing new approaches to finding ideology in narrative. Today, we find overtly political narratologies being articulated using demonstrably deconstructive vocabulary. They bring distinctly deconstructive approaches to bear on issues in the politics and ideology of narrative. A typical deconstructive reading, for example, will treat a certain binary opposition as a hierarchy in which one term of the opposition enjoys superiority over the other, often proceeding to demonstrate that the text contains counter-suggestions which upturn the hierarchy. For Currie, then, diversification, deconstruction, and politicization are the three mutually implicative characteristics of the transitions taking place in contemporary narratology. This diversification in theoretical assumptions and methodological approaches opens much common ground with identity theory, but it raises the question of whether narratology can still be seen as a coherent field. Currie’s answer is—still—a positive one. Roman Jakobson’s structuralist dream of a global science of literature has yielded to an uncontrolled fracturing of narratological method. Yet paradoxically, the particularity of texts or readers only becomes recognisable through a shared descriptive vocabulary which in itself constantly threatens to homogenise the heterogeneity it advances. It is this paradoxical model of change, the simultaneity of standardisation and diversification, which makes it still possible … to talk of narratology, if only provisionally, as if it were a unified entity (Currie 1998:14).

From the point of view of social psychology, then, narratology can be provisionally seen as a unity held together by a common descriptive vocabulary. Given the diversity of the narratological family, it is all the more important to identify the specific members (theoretical positions) that social psychologists are courting. Only then can the relationship between social psychology and narratology be seen as something more than one of superficial references. Fortunately, the discussion has, in social psychology at least, reached what Freeman has called a ‘post-polemical phase’: ‘By this I mean that most of those who have been working in the area of narrative for some time, are generally less concerned to indict the status quo than was once the case. … Many have come to find in narrative inquiry a remarkably fertile arena for reflection and exploration’ (Freeman 2001:283).

272

Wolfgang Kraus

5. Narrative Identity: The Call for a Postclassical Perspective It turns out, therefore, that social psychology and narratology present us not so much with a division between them as one cutting through both of them alike. It can be identified as that between post-modern / poststructuralist and structuralist approaches, or, more generally between a plurality of postclassical narratologies and classical, structuralist narratology (Nünning / Nünning 2002:3). Post-structuralist narratology and postmodern identity theory both developed out of the same critical endeavour, one encompassing literary criticism and architecture, cultural studies and philosophy. Deconstruction has left its mark on all theories concerning the individual and the social world. Sometimes, of course, its programme is questioned or emphatically rejected, but this only demonstrates how successfully it has made itself felt. So, while it would be wrong to subsume all recent approaches to identity theory under the deconstructive paradigm, it is clear that they must all position themselves in relation to its theses in one way or another. Given this shared deconstructive basis, it does not take much effort to grant narrativity an important role in the construction of identity. The foundation for such a move, however, is the ‘homogeneity’ of the deconstructive project rather than an easily accepted consensus regarding narrativity. From this perspective, introducing narrativity into identity theory does not solve theoretical problems but presents us with another problem: that of narrativity itself. Thus, post-modern identity theory will benefit only if it engages in a critical evaluation of the concept. One of the central problems here is the coherence trap of stories: the inherent tendency of self-narratives to present themselves as superficially coherent and stable projects. Post-structuralist narratology criticizes homogenization and coherence in the analysis of stories; similarly, post-modern identity theory takes opposition to classical concepts of identity on the same grounds. The affinity between the two fields of theoretical enquiry results from a critical attitude towards unity and the use of power in narratives. It is here that post-structuralist narratology and identity theory find common ground. Thus, post-modern approaches take as a basic idea the concept of narrative identity as developed by psychologists like Sarbin (1986) or Gergen and Gergen (1988). These authors have proposed an understanding of identity as narrative identity. Although the term ‘narrative identity’ has since become established, few authors venture to suggest a concise definition of it. Widdershoven (1993:7) has taken up the challenge, defining it as ‘the unity of a person’s life as it is experienced and articulated in stories that express this experience’. As Gergen and Gergen (1988) have emphasized, these self-stories are highly variable, but only within a specific social and


273

cultural frame. Each society has its own set of stories and rules for their construction. The storytellers therefore are not free in the stories they tell. Their stories find acceptance only if they follow these patterns in their self-stories and obtain the consent of other individuals with them. From a post-modern perspective, the most interesting part of Widdershoven’s definition is the idea of experiencing unity. This can serve as a starting point for a discussion of Freeman’s enquiry into the reasons for interest in narrative at this particular juncture. Wagner’s analysis of late modernity (1994) suggests that the reason may lie in the unifying properties of self-narrative. The telling of a self-narrative means constructing a coherent life, a task no longer made possible by society. Self-narratives do what stories are good at doing: they create coherence and fill a void whose presence is felt all the more intensely in late modernity. According to Levi-Strauss (1977:11), individual self-experience is always fragmented. The difference from earlier periods and other cultures lies in the fact that our societies at this moment in time no longer offer models of coherence for our fragmented experiences. More than ever before, this task is left to individuals to handle. When there is no (historical) agent capable of providing the content of life with meaning, there seems to be no (coherent) story to be told and understood by others. White (1981:7) takes a similar line in his analysis of medieval annals, which offer only a minimum of information and no story at all. According to White, these annals locate us immediately ‘in a culture hovering on the brink of dissolution’. The level of dissociation in subjects, then, is intricately connected with society and its readability. Can we link this argument with narratology? Indeed we can: in narratological terms, making meaning needs emplotment. This holds true for individuals as well as for societies. A plot, as E.M. Forster (2000:87) has stated, answers the question ‘why?’. Consider the death of the queen. If it is a story we say: ‘And then?’ If it is a plot we ask: ‘Why?’ That is the fundamental difference between these two aspects of the novel. A plot cannot be told to a gaping audience or to a tyrannical sultan or to their modern descendant, the movie-public. They can only be kept awake by ‘And then—and then—’ they can only supply curiosity. But a plot demands intelligence and memory also (Forster 2000:87).

From my point of view, emplotment requires not only individual capabilities, but also social intelligence and social memory, the social intelligibility of a historical period and the lives within it. Experiencing personal unity, then, not only requires a subject capable of producing it, but also a society that its members can read. The question of unity is not only a personal one; it is intricately connected with society. For identity theory, the discussion turns not on the death of the subject but on the interplay between

274

Wolfgang Kraus

self and other, between the subject and society, between individual strategies and social resources for making meaning. Storytelling is suspiciously like an ideologically based cure for dissociated subjects and dissociative social tendencies. Narratives help to create ready-made coherence, which must, as de Certeau argues, be critically deconstructed: ‘From politics to advertising, the story makes believe and by that it makes do, it takes up this and neglects that, it classifies. On the other hand it produces oblivion, it institutes silence on the things it does not talk about. And because it is always ‘full’ and closed, it makes even forget that it is withholding something’ (cit. in Adam / Revaz 1996:91, my trans.). So, there are two positions regarding the function of narrative in identity theory. Both accept the importance of narratives for the development of identity. One position is content because it has discovered narrativity as flexible means of constructing identity. It accepts the structuralist project of reading narratives as solid objects with specific meanings. The other position criticizes the ideological note in the idea of closure and coherence behind this understanding, insisting instead on the multiplicity of meanings, on their interactive construction. It demands that social psychology take into account the critical steps taken by post-structuralist narratology: the heterogeneity of self-experience should be analysed and narrative closure deconstructed.

6. Who Speaks? Who Sees? Who Holds Power? The diminished readability of a society characterized by subjects no longer embedded in it is seldom employed as an argument in identity theory, which usually focuses more prominently, if not exclusively, on the individual. Discussion in this area has been stimulated by the deconstructive study of the death of the subject; the result is a long and varied controversy about abandoning the concepts of identity and the self altogether. Genette’s famous question ‘who speaks?’, although posed with narratives in mind, finds open ears among social psychologists, who have battled over the concept of the self for a long time. Fortunately, what Freeman (2001) would call a postpolemical stage has been reached in this area. Although social psychology has expended a considerable amount of energy in analysing the subject, it is generally accepted that there is an empirical self, not in any philosophical sense but in the sense of a social, corporeal being. In a sense, we have come to accept again (or at last) what the psychiatrist Morton Prince proposed a century ago. After discussing at length the therapy of a psychiatric patient called Sally Beauchamp, a famous case of multiple personality disorder at the time, he concluded: ‘Philosophise as


275

you will, there is an empirical self which may be designated the real self’ (Prince 1905:233). This does not mean falling back to a position in which the self is seen as a simple unity, let alone having an essentialist understanding of the self. On the contrary, the question concerns the situation behind the scenes, the interplay between the speaking subject and the multiplicity of selves constituting any single person. It makes more sense to talk about the coherence of the subject not as a harmonious state but as a ‘complex coherence’ (Camilleri 1991:89), which implies the constant (re-)establishment of equilibrium between the various pieces of feedback an individual obtains from his or her situational context. Thus, identity is not present in us but ‘structured by, or constituted by difference’ (Currie 1998:17). When we look at a person more closely, we must decompose him or her into a complex set of relations which we can describe as a unity in no more than a very superficial manner. The discussion of the self in psychology often has a somewhat monadic tone, for it tends to focus on the subject isolated from its dynamic relationship with the world and with others. Bruner is thus perfectly correct when he states that the self is a ‘surprisingly quirky idea—intuitively obvious to common sense, yet notoriously evasive to definition’ (2002:63). The self must therefore be understood as processed, socially embedded, and readable through the self-stories in which it manifests itself discursively. Narratology steps in at this point with analytic tools for examining the relationship with the other. ‘Telling oneself about oneself’ (Bruner 2002: 64), as well as telling oneself to others, is the basic principle of self-construction. To understand this process, we need to analyse the processes (the telling), the relationships (between teller and listener), and the form and content of such self-stories. If we accept that this relationship with the self and the other is realized in the process of communication, then we can conclude that that is where the self exposes itself and becomes readable. As an individual, the author of a self-story is seen—metaphorically speaking—as a split personality, a person with many selves, constantly trying to rearrange him- or herself into a provisional unity. These moves must be reflected in the process—and product—of telling. Benveniste proposes a tripartite division into three subjects in the structure of discourse: the speaking subject or material agent of the discourse; the subject of speech or purely linguistic subject of the discourse, designated by personal pronouns; and the spoken subject or subject produced through or by the discourse as a result of its effect on a listener / receiver (cf. Kerby 1997:132). These subjects can be understood as different communicative addressers addressing different addressees on their respective levels of communication (Jahn 2000:N2.1).

276

Wolfgang Kraus

This example alone is enough to indicate the potential uses of narratological terminology in analysing the discursive production of self-stories. Narratology offers a wide array of categories and analytic tools ‘to analyse stylistic choices that determine the form or shape of a narrative text …, and also the pragmatic features that put this text into context within the social and structural framework of a narrative act’ (Jahn 2000:N1.4). What social psychology does with the self, then, narratology does with the relationship between the participants in a discourse on the various levels of communication. Seen from a post-modern perspective, it is not surprising that the concepts of Bakhtin have gained particular prominence in work on identity and the self in social psychology. His analysis of polyphony in narrative (i.e. a diversity of voices on different levels of communication) and his introduction of the concept of heteroglossia (highlighting linguistic elements inherited or learned from others), are employed in current theories of the self—an example is the concept of a dialogical self (Hermans 2001). Genette’s second question, ‘who sees?’, involves the focalization of narratives. ‘Functionally, focalisation is a means of selecting and restricting narrative information, of seeing events and states of affairs from somebody’s point of view’ (Jahn 2000:N3.7). In the view of social psychology, the concept of focalization is highly relevant to our understanding of the structure of power in narrative discourses. Whose gaze is the privileged one, which focus is preferred, which is suppressed? Approaches based on identity politics are particularly interested in questions of power and privilege as they appear in self-stories. Recent developments in psychology such as positioning theory (Langenhove / Harré 1999) show a considerable affinity with narratology’s idea of focalization, which can be external (a narrator) or internal (a character). According to Langenhove and Harré, it is necessary to analyse power structures not only within a narrative but also in the performance of narrating (i.e. the communicative process). On the level of interaction, at least, this brings us closer to answering the question of whether individuals are the authors of their stories or simply their protagonists. This question is highly important for individual lives. When we analyse practices of social exclusion, we help to empower people in their individual lives. As the privileged medium of self-construction, narratives are a prominent battleground in the fight against exclusion and suppression. The practical question here is whether it is possible to create what we might call counter-stories in addition to simply analysing the mechanisms involved. According to Nelson, ‘because identities are narratively constituted and narratively damaged, they can be narratively repaired. The morally pernicious stories that construct the identity according to the requirements of an abusive power system can be at least partially dislodged and replaced by identity-constituting counterstories that


277

portray group members as fully developed moral agents’ (2001:xii). This focus on the suppression of and attack on identity accompanies Sampson’s critique of the rational modern self as ‘a most peculiar self’ in world history, characterized by implicit structures of dominance with regard to ‘serviceable others’ (1993:4).

7. Self-Narratives: Examining the Process What is the analytic unity we have in mind in models of the narrative construction of identity? Clearly we are looking at self-narratives, but we have still to determine their properties as stories. One central feature should be stressed: self-stories are rarely written down. More often than not, they are unavailable as texts, being orally constructed and communicated instead. Thus, unless they are recorded in some medium or other, which is certainly not the norm, they tend to elude analysts. There is a not inconsiderable distance between this situation and a narratology stemming from the analysis of literary production and presupposing a text consciously constructed by a reflective author / subject. One solution may be to focus on autobiographical self-(re)construction. Autobiographical research readily presents itself as a potential source of common ground between narratology and social psychology. It has inspired very interesting work on self-theory and self-construction. In the paradigm of ongoing, never-ending identity construction, however, its use is limited because of its focus on rather complete self-stories. The stories with which we are dealing in identity theory, however, are usually orally produced, and they lack the qualities of elaborate literary constructions, are volatile, and have a presentation that changes continually according to audience and circumstances. The narratological movement makes no secret of its belief that stories are everywhere; it is less widely recognized that such stories can be very different from those originating in literary production. Wellformed self-stories, as Gergen and Gergen, following Labov, conceive of them, do not seem to appear all that often. What about fragmentation, about people who have too many stories to tell, who get lost in their manifold social relationships and rarely, if ever, provide the well-formed stories Gergen and Gergen are looking for? After all, as Kerby remarks, ‘we are at most only part-time Montaignes. We tend to narrate ourselves only when the situation calls for it’, continuing, ‘at the level of everyday reality our narratives most certainly have little of the consistency or coherence of full-blown autobiographies or stories. In fact, I am not even sure, that the majority of people have too great a desire or need to know who they are’ (1997:129).

278

Wolfgang Kraus

This situation calls for thorough reflection. First, we turn to narratology for help in translating and adapting its concepts for use with oral storytelling. Second, as oral storytelling produces a different kind of story with different form and content (Bauman 1986), oral self-stories must be analysed in terms of these properties and as performances (Langellier / Peterson 2004). My third point is that, once more, it is of paramount importance to integrate situational context into the analysis of narrative: ‘Narrative performance thus refers to a site of struggle over personal and social identity rather than to the acts of a self with a fixed, unified, stable, or final essence which serves as the origin or accomplishment of experience’ (Langellier 2001:151). The concept of narrative most frequently referred to in social psychology is based on the work of Labov and Waletzky. It has the advantage of a high level of refinement and a long history in social science (Bamberg 1997). Even when a narrative does not meet all the standards of this concept, the deviations can be interpreted with respect to what they say about the construction of identity (Kraus 2002). When analysing everyday communication, however, it may be helpful to choose a concept which is closer to Prince’s idea of the minimal narrative (1989). Another proposal, also originating in Prince, is the idea of distinguishing different levels of narrativity (1999:43f.). Thus, a text can be more or less narrative, and certainly more or less elaborate. Boje (2001) has proposed yet another strategy. While retaining a high standard for classifying texts as narratives, he has introduced the concept of antenarrative to refer to storytelling of a ‘lower’ kind, in which the question of causality is not addressed. Essentially, then, his proposal is based on Forster’s distinction between story and plot. More often than not, stories in organizations, Boje’s principal area of interest, fall short of providing answers to the causal question. It is here that he sees the need for the notion of antenarrative. When printed, stories in literature have reached the end of their production process, but the status of stories in real life (whatever that may be) is unclear. They may lie at the beginning of a process of individual definition—or at the end of one. They may already have been integrated into the self or may still be undergoing evaluation as candidates for further integration. They may or may not still be being tested before various audiences. The evaluation of these narrative qualities can lead to plausible conclusions only if the subject as narrator is included in the analytic process. Thus, focusing on self-stories which have become fundamental for an individual’s understanding of him- or herself involves time and stories— communication in time through and about stories. Those stories can never be closed in any final sense.


279

8. Constructing the Social: Community Narratives Self-narratives are about constructing an identity in a social world. This world comes into being as a personally narrated one characterized by personal relationships with the other. The narrator positions the main character in his or her story in relation to other characters. On the level of discursive presentation the characters are listeners and joint presenters of the story. Telling thus becomes an act of negotiating meaning in which the narrator is not free in his or her telling. Instead, he or she needs the confirmation of others. The telling of self-stories becomes an act of negotiation and joint construction. Narratives, however, refer not only to relationships between individuals but to whole communities as institutionalized contexts for creating meaning. When social psychologists look at organizations, for instance, they are interested in how these communities create integration and differentiation, how they produce coherence, stability, and change. Narratives enter the frame on the level of individual narratives (in the sense that they jointly construct and relate to this communal picture) as well as on that of organizations (where community narratives represent a more or less official narrative). Here again, as with self-stories, we are dealing not only with one narrative for each community but with a set, a stream of community narratives. The individual (re)construction of this social world is a process which is constantly mediated by other agents and the process of experiencing community. As with agency, here too we must focus on the self-other relationship. This leads to an issue which is undeniably the concern of social psychology: why is it that stories in and about a community change, are created, told, retold, and live on in a web of stories? It is not so much the single story itself that is of interest here but rather the process of creating, sharing, and remaking stories. If we are to understand groups and organizations as story-telling structures, the question is how these stories develop, how they attain validity, and how we should analyse the processes of their change. Over the last fifteen years, social psychology has investigated the narrative construction of organizations. The question that arises is whether organizations are not what we see on organizational charts or in buildings but rather something constructed by people in the organizations in a communal—often controversial—effort to give meaning to their actions. The ‘micro-processes of talk thus become the macro-structure of organisation’ (Cerulo 1997:401). Analysing organizations thus means analysing storytelling in them. Deconstruction here involves ‘ways of reading that decanter or otherwise unmask narratives that posit authoritative centres’ (Boje 2001:19).

280

Wolfgang Kraus

This, however, is no easy task in an increasingly globalized world in which the very notion of community is clearly open to debate. Understanding the construction of community narratives is not only a theoretical question but also a political one, as communitarian discourse has amply shown. It aims at making clear how ideologies are upheld and how they can be changed. It also asks, following research on late modernity, whether a lack of social cohesion, an acceleration of social change, a heightened mobility will lead to the loss of stories and thereby to the loss of community. ‘Man is in his actions and practice essentially a story-telling animal. … By neglecting the question ‘of what story or stories do I find myself a part?’ leaves us as individuals unscripted and hence powerless to detect the disorders of moral thought and practice’ (McIntyre cit. in Abbott 1997:281). How, for instance, can a sense of community be developed when there are no shared stories because of the immense cultural and biographical diversity in areas inhabited by immigrants? Where is cohesion going to come from in an individualized society if the reserve of shared stories is constantly growing smaller? Many narratologists, no doubt, would agree that culturally shared stories are being progressively undermined and overthrown by empirical science and critical rationalism. Some of these thinkers, moreover, fear that the decline of narrative will make it more difficult to maintain any kind of social cohesion at all given that the latter has relied so much on common stories in the past. In times of rapid social change, the ‘guiding myths developed by previous generations fail to fit the territory’ (Hinchman / Hinchman 1997:236). Thus, not only is loss registered; a need to fight against it, to become involved in the construction of community, is also expressed: ‘What matters at this stage of the late twentieth century is the construction of local forms of community within which civility and intellectual and moral life can be sustained through the dark ages which are already upon us’ (McIntyre cit. in Fisher 1997:307). Even if we are sceptical of communitarian analysis, we cannot deny that the construction of a sense of belonging is changing. Narratives thus have a central role to play because they are a means of constructing social bonds. 9. Social Psychology and Narratology: The Old and New Question of Virtuality We should not fail to see that narratology is also about people and that social psychology has many stories to tell. The question of the epistemological status of narrativity draws together many different aspects of the individual. To date, this discussion has been extensively (but not exclusively) shaped by philosophical contributions. Impressive as these findings


281

are, it is not easy to work out their consequences for social psychology and its organization. What is meant by ‘narrative psychology’? Is it a subfield of psychology in the ways developmental or cognitive or biological psychology are? Probably not. Rather, many theorists would claim that ‘narrative psychology’ refers to a viewpoint or a stance within psychology which is interested in the ‘storied nature of human conduct’ (Sarbin 1986)—how human beings deal with experience by constructing stories and listening to the stories of others. Psychologists studying narrative are challenged by the notion that human activity and experience are filled with ‘meaning’ and that stories, rather than logical arguments or lawful formulations, are the vehicle by which that meaning is communicated. This dichotomy is expressed by Jerome Bruner as the distinction between ‘paradigmatic’ and ‘narrative’ forms of thought … Sarbin (1986) proposes that narrative becomes a root metaphor for psychology (Hevern 2003).

The relative importance of a narratological approach to psychology is heavily dependent on the epistemological place we are prepared to give narrativity in our concept of human being. Hinchman and Hinchman (1997:xix f.) distinguish strong narrativists from weak ones. The groups disagree over whether there is such a thing as pre-reflective, pre-narrative, unemplotted experience (the weak school), or whether all human experience possesses at least an implicit narrative structure (the strong school). While this dichotomous picture may highlight the difference between the two positions, in the long run, it may be more helpful to investigate the ground between them. This is what Eakin, for example, does on the level of self-theory when he proposes distinguishing different registers of the self. He develops a model of ecological and interpersonal selves that precede other selves in child development. This concept suggests ‘that it is time to discard restrictive notions of the self and the subject that make of them little more than metaphysical or narrative puppets, opening the way for a much broader, experientalist approach to the nature and origin of subjectivity’ (Eakin 1999:25). Eakin is interested not in identifying himself as a strong or weak narratologist but in covering findings from, for example, child psychology and thus helping to establish narratology as a basic metaphor by tying it together with other basic findings of the social sciences. Others, like Sartwell, criticize a restrictive focus on language in current theoretical thinking. According to Sartwell, ‘alongside the hermeneutic tradition that affirmed the absolute hegemony of language as if it were a liberation, there developed a narratology that developed everything into stories and that made stories definable in terms of telos’ (2000:3). Sartwell provides not only criticism but also a detailed effort ‘to sketch the limits of narrative as a category, and some of its political sources and implications’ (2003: 10). He thereby makes an undeniable contribution to the development of the narratological project.

282

Wolfgang Kraus

Narratology proper has witnessed other attempts to explore narrativity. They adopt a wider view than social psychology, positioning narrativity in the long epistemological history of mimesis. At the same time, however, they seek to liberate it from its confinement to literary criticism and view it in the light of a new debate on virtuality in our culture. Thus, Schaeffer (1999) traces the development of fiction in humankind, basing it on anthropological functionality. His reflection on mimesis and analysis of anti-mimetic positions brings the discussion to a point where it is in a good position to consider new developments in fiction that are based on the computer revolution and have led to new presentations of virtuality such as video games. Ryan (2001) ventures deep into the labyrinth of principles behind the construction of computer games. Her journey is an exploration of the construction of virtuality—it is human beings that do the programming and they that fall for the supposed reality of the virtual world (or don’t they?). Thus, again we find ourselves facing fundamental reflection on mimesis and perception beyond the confinements of disciplinary boundaries. And, because the users of such games are intended to become characters in their narratives, narrative is being analysed as a form of participation and performance as well. This presents identity theory with a most useful discussion of coherence, performance, and immersion in narrative. Bal’s project, finally, investigates the role of narrativity in the context of a larger cultural theory. She advocates ‘a narrative theory that enables the differentiation of the place of narrative in any cultural expression without privileging any medium, mode, or use; that differentiates its relative importance and the effect of the narrative (segments) on the remainder of the object as well as on the reader, listener, viewer’ (1999:21). For Bal, narratology should be understood as part of a cultural theory. When this perspective is adopted, cultural artefacts, events, or domains can be analysed in detail. Cultural analysis should not be taken literally—or analytically—as meaning the ‘taking apart’ of culture. Rather, cultural analysts interpret the way in which cultures take things, people, and themselves apart. Narratology is of great relevance for such a reorientation of, simultaneously, close reading and cultural studies into cultural analysis (Bal 1999:21).


283

10. Conclusion: Narratology as a Resource for Social Psychology It is wrong to see social psychology as something beyond, foreign, external to narratology. The narrative turn in social science is about twenty years old now, and there has been an extensive discussion of the concept of narrativity in social science. Narratology itself has developed into such a plurality of distinct narratologies that viewing the two as binary opposites is far too simple to do justice to the situation. Refined as the discussion has become, the shared search for resources is shaped primarily by the theoretical focus adopted in each particular instance. From the point of view of a theory of narrative identity open to post-modern concerns, interest in narratologies is primarily directed at contributions to be made by the post-structuralist debate. This, however, is only part of the story. The more the narrative approaches in psychology become concerned with empirical research, the greater their interest in the methodological possibilities provided by narratology. Here, I suspect, mutual acceptance is only beginning. Another productive point of exchange lies in the discussion of mimesis and virtuality. Some narratological contributions move convincingly beyond the limits of literary criticism and thereby provide potential benefits for the theory of the subject that are yet to be recognized by those working in other fields. To conclude, therefore, when we consider the benefits of looking beyond social psychology into the multifaceted world of narratologies, we find a mixture of close affinity and surprising strangeness. – Common grounds. Even if the concepts of a narrative identity theory do not commit themselves unreservedly to a deconstructive position, the questions arising from the latter’s critique have initiated an intensive, sometimes even impassioned debate. Post-structuralist and deconstructivist narratologies make welcome contributions to this discussion. They offer a view of narrativity that makes it compatible with the post-modernist discussion of identity. – Analytical resources. The critical project of deconstructivism was easily established as a source of common ground on the level of theory, but the ‘shared descriptive vocabulary’ (Currie) of current narratologies has not yet been exploited by narrative identity theory. Although some narratologists seem hesitant to accept the very idea that their concepts might migrate to other fields of enquiry, recent discussions show that many narratological concepts (e.g. the concept of metalepsis) may be of use in the analysis of non-fictional autodiegetic narratives. – Narrative as performance and immersion. In the technological world of computer games, performance becomes important in two ways, first through the developer, who designs the active role of the player / par-

284

Wolfgang Kraus

ticipant, and second by the player, who takes on that role. Ryan’s investigation of the design and use of computer games has shown that there is much more to be discovered from such analysis. It is also an exploration of the relationship between reality, virtuality, and authenticity, and as such is a useful contribution to the theory of the subject. – Art as an illustration of theory. Literary production, the home of narratology, provides fruitful material for social psychology too. Many authors are working and thinking about narrativity as a theoretical concept. When reading the works of Salman Rushdi, Julian Barnes, and many others, for instance, we find ourselves participating in a theoretical discourse expressed in literary work. Thus, literary productions such as these are also theoretical reflections on humanity and reality and the place of narrativity in the latter. – ‘Literature is a wilderness, psychology is a garden’ (Albright 1994:19). What about literature as the wilderness, films and paintings as the other of theoretical production? Even without what we might call theoretically informed literary production, literary narratives offer valuable—and irreplaceable—insights into human subjectivity and imagination, because, as E. M. Forster has said, art, through the homo fictus, tells us something about ourselves: We cannot understand each other, except in a rough and ready way; we cannot reveal ourselves, even when we want to; what we call intimacy is only a makeshift; perfect knowledge is an illusion. But in the novel we can know people perfectly, and, apart from the general pleasure of reading, we can find here a compensation for their dimness in life. In this direction fiction is truer than history, because it goes beyond evidence, and each of us knows from his own experience that there is something beyond the evidence (Forster 2000:70).

References Abbott 1997 Abbott, Philip: “Storytelling and Political Theory.” In: Hinchman, Lewis P. / Hinchman, Sandra K. (eds.): Memory, Identity, Community. The Idea of Narrative in the Human Sciences. Albany, NY 1997:281–306. Adam / Revaz 1996 Adam, Jean-Michel / Revaz, Françoise: L’analyse des récits. Paris 1996. Albright 1994 Albright, Daniel: “Literary and Psychological Models of the Self.” In: Neisser, Ulric / Fivush, Robyn (eds.): The Remembering Self. Construction and Accuracy in the Self-Narrative. New York 1994:19–40. Amis 1973 Amis, Martin: The Rachel Papers. London 1973.


285

Bal 1999 Bal, Mieke: “Close Reading Today: From Narratology to Cultural Analysis.” In: Grünzweig, Walter / Solbach, Andreas (eds.): Grenzüberschreitungen. Narratologie im Kontext. Transcending Boundaries: Narratology in Context. Tübingen 1999:19–40. Bamberg 1997 Bamberg, Michael (ed.): Oral Versions of Personal Experience. Three Decades of Narrative Analysis. Journal of Narrative and Life History, Special Issue. Vol. 7 (1–4), London 1997. Bauman 1986 Bauman, Richard: Story, Performance, and Event. Cambridge, UK 1986. Boje 2001 Boje, David: Narrative Methods for Organizational and Communication Research. London 2001. Bruner 1990 Bruner, Jerome: Acts of Meaning. Cambridge, MA 1990. Bruner 2002 Bruner, Jerome: Making Stories. Law, Literature, Life. Cambridge, MA 2002. Camilleri 1991 Camilleri, Carmel: “La construction identitaire, essai d’une vision d’ensemble.» In: Les Cahiers Internationaux de Psychologie Sociale 1 / 2, 1991: 77–90. Cerulo 1997 Cerulo, Karen A.: “Identity Construction. New issues, New Directions.” In: American Review of Sociology 23, 1997:385–409. Currie 1998 Currie, Mark: Postmodern Narrative Theory. New York 1998. Eakin 1999 Eakin, Paul John: How Our Lives Become Stories. Ithaca 1999. Fisher 1997 Fisher, Walter R.: “Narration, Reason and Community.” In: Hinchman, Lewis P. / Hinchman, Sandra K. (eds.): Memory, Identity, Community. The Idea of Narrative in the Human Sciences. Albany, NY 1997:307–327. Forster 2000 Forster, Edward M.: Aspects of the Novel. London 2000. Freeman 1998 Freeman, Mark: “Mythical Time, Historical Time and the Narrative Fabric of the Self.” In: Narrative Inquiry 8, 1998:27–50. Freeman 2001 Freeman, Mark: “From Substance to Story. Narrative, Identity, and the Reconstruction of the Self.” In: Brockmeier, Jens / Carbaugh, D. (eds.): Narrative and Identity. Studies in Autobiography, Self and Culture. Amsterdam 2001:283–298. Gergen 1991 Gergen, Kenneth J.: The Saturated Self. New York 1991. Gergen / Gergen 1988 Gergen, Kenneth J. / Gergen, Mary M.: “Narrative and the Self as Relationship.” In: Berkowitz, L. (ed.): Advances in Experimental Social Psychology. New York 1988:17–56.

286

Wolfgang Kraus

Hermans 2001 Hermans, Hubert J. M.: “The Dialogical Self: Toward a Theory of Personal and Cultural Positioning.” In: Culture & Psychology 7 / 3, 2001: 243–281. Hevern 2003 Hevern, Vincent W.: Narrative Psychology: Basics. Narrative Psychology: Internet and Resource Guide. 2003. Hinchman / Hinchman 1997 Hinchman, Lewis P. / Hinchman, Sandra K. (eds.): Memory, Identity, Community. The Idea of Narrative in the Human Sciences. Albany, NY 1997. Jahn 2000 Jahn, Manfred: Narratology. A Guide to the Theory of Narrative. Version 1.3. 2000. Kerby 1997 Kerby, Anthony P.: “The Language of the Self.” In: Hinchman, Lewis P. / Hinchman, Sandra K. (eds.): Memory, Identity, Community. The Idea of Narrative in the Human sciences. Albany, NY 1997:125–142. Keupp et al. 2002 Keupp, Heiner / Ahbe, Thomas / Gmür, Wolfgang / Kraus, Wolfgang / Mitzscherlich, Beate / Straus, Florian: Identitätskonstruktionen. Das Patchwork der Identitäten in der Spätmoderne. Reinbek 2002. Kraus 2002 Kraus, Wolfgang: Falsche Freunde. Radikale Pluralisierung und der Ansatz einer narrativen Identität. Frankfurt / M. 2002:159–186. Langellier 2001 Langellier, Kristin M.: „‘You’re Marked’, Breast Cancer, Tattoo, and the Narrative Performance of Identity.” In: Brockmeier, Jens / Carbaugh, Donald (eds.): Narrative and Identity. Studies in Autobiography, Self and Culture. Amsterdam 2001:145–184. Langellier / Peterson 2004 Langellier, Kristin M. / Peterson, Eric E.: Storytelling in Daily Life. Performing Narrative. Philadelphia 2004. van Langenhove / Harré 1999 van Langenhove, Luk / Harré, Rom: “Introducing Positioning Theory.” In: Harré, Rom / van Langenhove, Luk (eds.): Positioning Theory. London 1999:14–32. Levi-Strauss 1977 Levi-Strauss, Claude: «Avant-propos.» In: Levi-Strauss, Claude (ed.): L’identité. Paris 1977:9–12. Nelson 2001 Nelson, Hilde Lindemann. Damaged Identities, Narrative Repair. Ithaca, NY 2001. Nünning / Nünning 2002 Nünning, Ansgar / Nünning, Vera: „Von der strukturalistischen zur ‘postklassischen’ Erzähltheorie: Ein Überblick über neuere Ansätze und Entwicklungstendenzen.“ In: Nünning, Ansgar / Nünning, Vera (eds.): Neue Ansätze in der Erzähltheorie. Trier 2002:1–34. Prince 1989 Prince, Gerald: A Dictionary of Narratology. Lincoln 1989.


287

Prince 1999 Prince, Gerald: “Revisiting Narrativity.” In: Grünzweig, Walter / Solbach, Andreas (eds.): Grenzüberschreitungen. Narratologie im Kontext. Transcending Boundaries: Narratology in Context. Tübingen 1999:43–51. Prince 1905 Prince, Morton: The Dissociation of Personality. New York 1905. Ricœur 1990 Ricœur, Paul: Soi-même comme un autre. Paris 1990. Ryan 2001 Ryan, Marie-Laure: Narrative as Virtual Reality. Immersion and Interactivity in Literature and Electronic Media. Baltimore 2001. Sampson 1993 Sampson, Edward E.: Celebrating the Other. A Dialogic Account of Human Nature. Boulder 1993. Sarbin 1986 Sarbin, Theodore R.: “The Narrative as a Root Metaphor for Psychology.” In: Sarbin, T. R. (ed.): Narrative Psychology. The Storied Nature of Human Conduct. New York 1986:3–21. Sartwell 2000 Sartwell, Crispin: End of Story. Toward an Annihilation of Language and History. Albany, NY 2000. Schaeffer 1999 Schaeffer, Jean-Marie: Pourquoi la fiction? Paris 1999. Wagner 1994 Wagner, Peter: A Sociology of Modernity. Liberty and Discipline. London 1994. White 1981 White, Hayden: “The Value of Narrativity in the Representation of Reality.” In: Mitchell, W. J. T. (ed.): On Narrative. Chicago 1981:1–23. Widdershoven 1993 Widdershoven, Guy A. M.: “The Story of Life. Hermeneutic Perspectives on the Relationship between Narrative and Life History.” In: Josselson, R. / Lieblich, A. (eds.): The Narrative Study of Lives. Newbury Park 1993: 1–20.

Author Index Aarseth, Espen J. 47, 62 Abbate, Carolyn 65, 66, 66, 79, 83, 105 Abbott, H. Porter 2, 4, 21 Abbott, Philip 280, 284 Abbott, Scott 192 Abelson, Robert 153, 172, 219, 236 Abrams, M.H. 147 d’Accone, Frank A. 79 Adam, Jean-Michel 274, 284 Ahbe, Thomas 286 Alber, Jan 105 Albrecht, Theodore 76, 79 Albright, Daniel 284 Allen, Woody 97, 102, 103, 105 Alm, Irene 79 Alphen, Ernst van 31, 42 Amis, Martin 268, 284 Anderson, Timothy A. 52, 63 Andersson, Sven-Gunnar 114, 123 Angehrn, Emil 178, 190 Ankersmit, Franklin R. 174, 190, 190, 193 Antaki, Charles 222, 233 Apollonius of Tyana 211 Appiah, Anthony Kwame 233 Aristides 203, 211 Aristotle 21 Aspden, Suzanne 106, 107 Aynsley, J. 43 Bach, Alice 210 Bach, Johann Sebastian 84, 106 Badura-Skoda, Paul 73, 79 Bakhtin, Mikhail M. 276 Bal, Mieke 25, 26, 27, 28, 31, 32, 32, 33, 42, 156, 170, 190, 209, 209, 282, 285 Bamberg, Michael 125, 146, 213, 214, 221, 222, 224, 225, 231, 233, 235, 237, 278, 285 Barnes, Julian 284

Barthes, Roland ix, 21, 31, 42, 45, 62, 125, 146 Bartholomäus 207, 210 Bartov, Omer 183, 183, 185, 190 Bateson, Gregory 56, 62 Bauman, Richard 278, 285 Baumgartner, Hans Michael 175, 176, 176, 181, 190 Beauchamp, Sally 274 Beethoven, Kasper Karl van 76 Beethoven, Ludwig van 65, 66, 66, 67, 69, 71, 74, 75, 75, 76, 76, 77, 77, 78, 79, 80, 81, 88, 100 Bely, Andrei 118 Benjamin, Walter 30 Bennington, Geoff 22 Benveniste, Émile 275 Berg, Harry van de 233 Berkhofer, Robert F., Jr. 181, 190 Berkowitz, L. 285 Bernhard, Thomas 29 Bernhart, Walter 106,107 Bernheim, Ernst 185, 190 Beyer, Marcel 26, 42 Bibace, Roger 237 Biber, Douglas 131, 132, 133, 134, 135, 146 Blank, Marc S. 52, 63 Boehm, Gottfried 25, 29, 42 Boer, Inge E. 42, 190 Bohleber, Werner 241, 244, 262 Bohnsack, Ralf 259, 262 Boje, David 278, 279, 285 Bokhour, Barbara 237 Böll, Heinrich 117 Bondarko, Alexander V. 112, 123, 124 Booth, Wayne C. 61, 254 Bordwell, David 2, 10, 21, 46, 46, 47, 50, 62 Borges, Jorge Luis 5, 30, 40 Bornemann, Christiane 101, 105

290

Author Index

Bortolussi, Marisa 125, 126, 147 Bourquin, Yvan 209, 210 Bowden, John 210 Bremond, Claude 1, 21 Brennan, T. 43 Breward, C. 43 Brockmeier, Jens 285, 286 Brogan, T.V.F. 106 Brooks, Peter x, 178, 190 Brown, Gillian 129, 137, 147 Browne, Sir Thomas 30, 33, 39, 42 Bruder, Gail A. 149 Bruner, Jerome 3, 5, 21, 125, 147, 268, 275, 281, 285 Brunner, Otto 192 Bryson, Norman 25, 26, 28, 42 Bultmann, Rudolf 207, 208, 209 Burgin, Victor 34, 43 Burridge, Richard A. 195, 209 Butler, Judith 224, 233 Butor, Michel 94 Caesar, Gaius Julius 206, 211 Caillois, Roger 46, 62 Camilleri, Carmel 275, 285 Campbell-Kibler, Kathryn 234 Canary, Robert 193 Capps, Lisa 125, 148, 220, 235 Caravaggio, Michelangelo Merisi di 42 Carbaugh, Donald 285, 286 Carroll, Lewis 92, 96, 103, 105, 106, 107 Case, Sue-Ellen 233 Casement, Roger 34, 35, 36, 37 Catan, Liza 225, 233 Cavell, Stanley 38, 43 Certeau, Michel 56, 62, 274 Cerulo, Karen A. 279, 285 Charlton, Michael 244, 262 Chartier, Roger 175, 190 Chatman, Seymour x, xiii, 1, 2, 21, 47, 62, 125, 127, 147, 149 Childs, Becky 129, 148 Chomsky, Noam 126, 147 Chung, Sandra 113, 123 Cicero, Marcus Tullius 185, 185 Cochrane, Lydia G. 190 Cohn, Leopold 211 Comrie, Bernard 113, 123

Cone, Edward T. 75, 79 Conrad, Joseph 30, 35 Conrad, Susan 131, 132, 133, 134, 135, 146 Conrady, Karl Otto 100, 105 Conze, Werner 192 Cooper, Barry 71, 79 Cornelius Nepos 208 Cornils, Anja 201, 209 Cortázar, Julio 88 Culler, Jonathan x, 28, 43 Culpeper, Jonathan 133, 149 Curie, Marie 261 Currie, Mark 270, 271, 275, 283, 285 Czerny, Carl 73, 74, 74, 79 Dällenbach, Lucien 90, 105 Dahlhaus, Carl 73, 79 Damasio, Antonio 152, 170 Daiute, Colette 233 Danto, Arthur C. 111, 123, 174, 191 Dawkins, Richard 13, 21 De Jong, Irene J.F. 209, 209, 210 Dennett, Daniel 153, 170 Derrida x Devereux, Georges 258, 262 Devlin, Ken 13 Disney, Walt 95, 95, 96, 102, 103, 105 Dixon, Peter 125, 126, 147 Doležel, Lubomír 151, 154, 171 Don, Abbe 3, 22 Donaldson, E.T. 129, 147 Dormeyer, Detlev 195, 195, 210 Dornes, Martin 241, 262 Dregger, Alfred 187 Drews, Sibylle 241, 244, 262 Droysen, Johann Gustav 174, 191 Duchan, Judith F. 149 Dukas, Paul 13 Dulz, Birger 244, 262, 263 Duve, Freimut 187 Duvenage, Pieter 194 Eakin, Paul John 281, 285 Echterhoff, Gerald 180 Eckart, Wolfgang 264 Eckert, Penelope 226, 233, 234

Author Index

Eder, Jens 179, 180, 191 Ehrich, Veronika 112, 123 Eisen, Ute E. 91, 195, 197, 198, 210 Eisenberg, Ann R. 148 Eliot, George 139, 151, 171 Eliot, T.S. 100, 105 Elster, Jon 152, 171 Emmott, Catherine 128, 147, 156, 171 Ende, Michael 93, 105 Ermarth, Elizabeth 181, 191 Ernst, Bruno 105 Escher, M.C. 84, 98, 99, 101, 103, 105, 106 Fauconnier, Gilles 153, 171 Fillmore, Charles 135, 137, 147, 219, 234 Fischer, Gottfried 240, 262 Fischer-Rosenthal, Wolfram 222, 223, 234, 236, 259, 262 Fisher, Walter R. 280, 285 Fiske, Susan T. 216, 237 Fivush, Robyn 284 Flämig, Walter 113, 123 Fletcher, Angus 88, 106 Flick, Uwe 236, 262 Fludernik, Monika xiii, 4, 22, 83, 92, 105, 125, 128, 147, 153, 156, 171, 179, 181, 182, 191, 217, 217, 232, 234 Fontane, Theodor 117 Forster, Edward M. 273, 278, 284, 285 Fowles, John 102 Francis, W. Nelson 134, 135, 135, 137, 138, 147 Frank, Joseph 31, 43 Freeman, Mark 217, 218, 234, 266, 271, 273, 274, 285 Frickenschmidt, Dirk 196, 201, 203, 204, 206, 208, 210 Fuchs, Esther 210 Fuchs-Heinritz, Werner 222, 234 Fulda, Daniel 176, 177, 191 Garbe, Christine 244, 263 García Landa, José Angel 109, 124 Gardner, Martin 105

291

Gates, Henry Louis 233 Gaudreault, André 2, 22, 47, 48, 49, 62 Geißler, Peter 241, 262 Genette, Gérard x, xiii, 2, 11, 22, 27, 32, 43, 47, 50, 56, 57, 62, 84, 84, 88, 88, 90, 105, 111, 111, 123, 198, 201, 274, 276 Gergen, Kenneth J. 268, 272, 277, 285 Gergen, Mary M. 272, 277, 285 Gerrig, Richard J. 125, 147 Glasheen, Anne-Marie 107 Gmür, Wolfgang 286 Goddard, Cliff 147 Gödel, Kurt 84, 106 Goethe, Johann Wolfgang von 30, 177, 239 Goffman, Erving 221, 234 Goodman, Nelson 85, 105 Goody, Jack 216, 234 Grabowicz, George C. 171 Grass, Günter 118 Greene, Graham 116, 123 Greenwood, John 216, 234 Groeben, Norbert 263 Gross, Larry 23 Griffith, D.W. 63 Grünzweig, Walter 285, 287 Gumperz, John 232, 234 Gunning, Tom 47, 48, 49, 61, 63 Habermas, Jürgen 63, 258, 262 Hagestedt, Lutz 264 Hahn, Ulla 26, 43 Hall, Rogers 230, 234 Hanks, William F. 22, 234 Harré, Rom 276, 286 Harth, Dietrich 173, 191 Harvey, John H. 237 Hatten, Robert 73, 79 Haubl, Rolf 259, 262 Hauser, Caspar 241 Haydn, Joseph 67, 80 Heath, Malcolm 21 Heath, Shirley Brice 217, 235 Heath, Stephen 21, 146 Hedling, Erik 80 Heer, Hannes 183, 184, 184, 186, 187, 187, 189, 191

292

Author Index

Heinemann, Michael 75, 77, 79 Heise, Ursula K. 16, 22 Helbig, Jörg 94, 105 Henisch, Peter 26, 43 Herman, David xiii, 6, 22, 84, 88, 106, 125, 125, 127, 128, 129, 129, 130, 131, 138, 139, 147, 148, 153, 155, 171, 179, 179, 180, 181, 186, 192, 215, 218, 219, 234, 235 Hermann, Judith xiii, 250, 251, 253, 254, 262, 263, 264 Hermans, Hubert 276, 286 Herndl, Diane Price 148 Heubeck, A. 210 Hevern, Vincent W. 265, 281, 286 Hewitt, Lynne E. 125, 128, 129, 137,149 Hickethier, Knut 83, 106 Hight, Jeremy 18, 22 Highter, Melissa 148 Hill, Deborah 129, 137, 149 Himmler, Heinrich 184 Hinchman, Lewis P. 280, 281, 284, 285, 286 Hinchman, Sandra K. 280, 281, 284, 285, 286 Hirsch, Mathias 240, 262 Hitler, Adolf 188 Hitzler, Ronald 234, 262 Hockey, Susan 148 Hölderlin, Friedrich 30 Hoffmann, E.T.A. 80 Hofstadter, Douglas 84, 106 Hogan, Patrick C. 216, 235 Hoiting, Nini 131, 141, 144, 149 Hollander, John 88, 106 Homer 201, 210 Honer, Anne 234, 262 Hornback, Bert G. 171 Horstkotte, Silke 25, 43 Houtkoup-Steenstra, Hanneke 233 Howard, Richard 44, 146 Hughes, E. 105 Huizinga, Johan 45, 46, 63 Hulse, M. 44 Ickes, William 237 Ingarden, Roman 154, 171 Iser, Wolfgang 154, 171, 218, 235, 242, 262

Jachnow, H. 124 Jackendoff, Ray 129, 148 Jahn, Manfred 11, 22, 125, 148, 179, 182, 192, 275, 276, 286 Jakobson, Roman 111, 112, 123, 271 Jander, Owen 75, 79 Jay, Martin 25, 43 Jefferson, Gail 226, 236 Jensen, Olaf 194 Johnson, Mark 137, 148 Johnstone, Barbara 127, 148 Jolly, Constance S. 81 Jones, Timothy 71, 79 Josselson, Ruthellen 217, 235, 287 Jost, François 2, 22, 47, 63 Joyce, James 117, 120, 123 Jünger, Ernst 264 Jureit, Ulrike 184, 188, 192 Kako, Edward T. 148 Kant, Immanuel 146, 190 Kaplan, Bernard 226, 237 Kardorff, Ernst von 236, 262 Keil, Frank C. 171 Kellner, Hans 193 Kellogg, Robert 201, 211 Kerby, Anthony B. 275, 277, 286 Kermode, Frank 118, 123 Kernberg, Otto F. 244, 262, 263 Keßler, Eckhard 173, 192 Keupp, Heiner 268, 286 Kidd, Robert F. 237 Kierkegaard, Sören 10 Kimball, John P. 149 Kinderman, William 66, 80 Kindt, Tom xi, 124, 191 King, Stephen 21 Kirsner, Kim 128, 149 Kivy, Peter 65, 66, 80 Klauck, Hans-Josef 207, 210 Knape, Joachim 174, 185, 192 Knowles, Gerry 131, 148 Knowlton, Jeff 18, 22 Koch, Torsten 194 Kocka, Jürgen 193 Kohut, Heinz 243, 262 Konova, Joana 173 Korobov, Neill 222, 235 Koselleck, Reinhart 176, 176, 180, 192, 194

Author Index

Kozicki, Henry 193 Kramer, Lawrence 73, 78, 80 Kraus, Wolfgang 265, 278, 286 Krieger, Murray 29, 43 Kristeva x Kucera, Henry 134, 135, 135, 137, 138, 147 Kuhn, Thomas 170, 171 Kwint, Marius 31, 43 Labov, William xi, 111, 112, 123, 137, 148, 232, 235, 277, 278 Lämmert, Eberhard 176, 192, 203 Lagerroth, Ulla-Britta 80 Lakoff, George 137, 148, 219, 235 Lamott, Franziska 259, 262 Landau, Barbara 129, 148 Langellier, Kristin 278, 286 Langenhove, Luk van 276, 286 Lanser, Susan Sniader x, xiii, 139, 148, 256 Lanzmann, Claude 264 Laurel, Brenda 22 Lebling, David 52, 63 Leh, Almut 183, 187, 193 Lehmann, Volkmar 113, 114, 120, 123, 124 Leichtentritt, Hugo 72, 80 Lemon, Lee T. 63 Lessing, Gotthold Ephraim 12, 14, 22, 32 Lethen, Helmut 189, 193 Lévi-Strauss, Claude 273, 286 Lewin, Jane 22, 43 Leyh, Peter 191 Lightfoot, Cynthia 233 Linde, Charlotte 137, 148 Lindemann Nelson, Hilde 286 Lloyd, B.B. 236 Lodato, Suzanne M. 106, 107 Lodge, David 84, 88, 106, 107 Lotman, Jurij M. xiii, 20, 111, 116, 124, 197 Luhmann, Niklas 46, 53, 63 Lund, Hans 80 Lutz, Heinrich 192, 194 Lyotard, Jean-François 3, 13, 22, 268 MacArdle, Donald W.

81

293

MacLaury, Robert E. 216, 237 Malina, Debra 84, 92, 103, 103, 106 Mann, Golo 174, 193 Mann, Thomas 117 Marcellus 203 Margolin, Uri 161, 171 Marguerat, Daniel 209, 210 Maron, Monika 26, 43 Martinet, Hans 211 Martinez, Matias 176, 193, 241, 254, 255, 255, 256, 257, 261, 263 Martínez-Bonati, Félix 254 Márton, László 26, 43 Marx, Adolf Bernhard 72, 73, 80 Maslov, Jurij S. 114, 124 Massumi, Brian 22 Mateas, Michael 125, 148 Maus, Fred Everett 65, 69, 80 Mauser, Wolfram 240, 263 McClary, Susan 69, 70, 80, 106 McCormick, Edward Allen 22 McEnery, Tony 131 McGraw, Nancy 148 McHale, Brian 84, 88, 96, 104, 106 McIntyre, Karen 280 McLamore, Alyson 79 McLaughlin, Kathleen 193 McLuhan, Eric 22 McLuhan, Marshall 17, 22 Megill, Alan 177, 193 Mehlig, H.R. 123 Meister, Jan Christoph 111, 124, 146 Mentzos, Stavros 259, 259, 263 Mercer, Sarah 83 Merenlahti, Petri 197, 209, 210 Mertens, Wolfgang 244, 263 Metz, Christian 2, 22 Mink, Louis O. 177, 193 Misch, Ludwig 72, 80 Mitchell, Stephen 263 Mitchell, W.J.T. 23, 25, 29, 32, 34, 43, 194, 287 Mitzscherlich, Beate 286 Moll, James 249, 264 Moller, Sabine 189, 194 Mommsen, Theodor 173, 174, 193 Montaigne, Michel de 269, 277 Morgenstern, Christian 100

294

Author Index

Mosthaf, Franziska 83, 106 Müller, Hans-Harald xi, 124, 191 Müller, Heiner 264 Naigles, Letitia 131, 141, 144, 145, 148 Nash, Christopher 84, 90, 106 Nattiez, Jean-Jacques 65, 66, 80 Naumann, Klaus 183, 184, 186, 187, 191, 193 Nebel, Bernhard 153, 171 Neef, Sonja 36, 43 Neisser, Ulric 284 Neitzel, Britta 45, 46, 63 Nerbonne, John 112, 124 Neubauer, John 69, 80 Nickels, Christa 187 Nipperdey, Thomas 193 Nitsche, Stefan Ark 209, 211 Nünning, Ansgar 23, 44, 81, 107, 191, 194, 198, 210, 218, 235, 237, 269, 272, 286 Nünning, Vera 23, 44, 81, 107, 191, 194, 218, 235, 237, 269, 272, 286 O’Brien, Flann 102 Ochs, Elinor 125, 148, 220, 235 Olson, David R. 216, 217, 235, 236 Onega, Susanna 109, 124 Ong, Walter 15, 22, 216, 236 Ostaijens, Paul van 43 Packard, Edward 93, 94, 102, 106 Palmer, Alan 151, 152, 161, 167, 172, 250, 251 Parsons, Mikeal C. 206, 211 Pavel, Thomas 154, 172 Pazhtnov, Alexey 64 Peer, Willi van 125, 149 Pellauer, David 193 Pelopidas 203, 211 Pembaur, Josef 75, 80 Perrin, Bernadotte 211 Perry, Menakhem 156, 163, 172 Peterson, Eric E. 278, 286 Petrie, Graham 107 Pfandl-Buchegger, Ingrid 83 Pfister, Manfred 83, 106 Pfotenhauer, H. 42

Philo of Alexandria 206, 211 Philostratus 211 Pier, John 105, 107, 209 Pietzcker, Carl 240, 263 Pinker, Steven 152, 172 Pirandello, Luigi 94, 95, 102, 103 Plato 11, 48, 63 Plutarch 201, 203, 206, 211 Podesva, Robert J. 234 Polanyi, Livia 232, 236 Pollak, Alexander 188, 193 Porath, Hann-Jörg 176 Potter, Jonathan 221, 222, 234 Prantl, Heribert 183, 193 Pratt, Mary Louise 5, 22 Preminger, Alex 106 Prieto, Eric 85, 86, 106 Prince, Gerald x, xiii, 2, 6, 22, 109, 110, 111, 111, 112, 116, 118, 120, 124, 278, 286, 287 Prince, Morton 274, 275, 287 Ps.-Kallisthenes 201, 211 Racker, Heinrich 244, 245, 263 Raguse, Hartmut 240, 263 Rauchenecker, Eva 120, 124 Reardon, Coleen 79 Reemtsma, Jan Philipp 183, 186 Reichenbach, Hans 112, 122 Reich-Ranicki, Marcel 242 Reis, Marion J. 63 Remarque, Erich Maria 117 Rembrandt (d.i. Harmensz van Rijn) 30 Reppen, Randi 131, 132, 133, 134, 135, 146 Revaz, Françoise 274, 284 Richardson, Brian 10, 22 Ricœur, Paul xiii, 174, 175, 193, 217, 266, 287 Riedesser, Peter 240, 262 Riegel, Klaus F. 225, 236 Rigney, Ann R. 174, 193 Rimmon-Kenan, Shlomith 9, 23 Roberts, Geoffrey 194 Roberts, Sarah 234 Robinson, J. Bradford 79 Rohde-Dachser, Christa 243, 263 Ronen, Ruth 154, 156, 172 Rosch, Eleanor 219, 236

Author Index

Rosen, Charles 67, 80 Rosenthal, Gabriele 222, 223, 234, 236, 259, 262 Ross, Garry 97 Rüsen, Jörn 174, 174, 189, 190, 192, 194 Rushdi, Salman 284 Ruysdael, Salomon van 30 Ryan, Marie-Laure xiii, 1, 7, 8, 9, 10, 11, 14, 18, 20, 23, 84, 86, 88, 89, 89, 90, 92, 100, 101, 102, 103, 106, 107, 125, 129, 147, 148, 154, 156, 158, 172, 180, 185, 242, 282, 284, 287 Ryle, Gilbert 158, 162, 172 Sachsse, Ulrich 244, 262, 263 Sacks, Harvey 226, 236 Sadie, Stanley 80 Salinger, Jerome David 114, 116, 124 Sampson, Edward 269, 277, 287 Sanders, Robert 222, 236 Sarbin, Theodore R. 272, 281, 287 Sartwell, Crispin 281, 287 Schaberg, Jane 210 Schaeffer, Jean-Marie 209, 282, 287 Schafer, Roy 242 Schank, Roger 153, 172, 219, 236 Scheffel, Michael 176, 193, 241, 254, 255, 255, 256, 257, 261, 263 Schegloff, Emanuel 226, 236 Scher, Steven Paul 106, 107 Schering, Arnold 75, 80 Scheuerl, Hans 46, 63 Schiller, Friedrich 239 Schily, Otto 187 Schindler, Anton Felix 74, 74, 75, 76, 77, 78, 81 Schmalfeldt, Janet 71, 73, 81 Schmid, Wolf 114, 116, 124 Schmitz, Thomas A. 209, 211 Schmitz, Walter 264 Schneemelcher, Wilhelm 210 Schneider, Silvia 244, 262 Schnelle, Udo 198, 211 Schoenfeld, Howard 92 Schoett, Silja 244, 263 Scholes, Robert 201, 211 Schüssler Fiorenza, Elisabeth 210

295

Searle, John 153, 159, 172 Searles, Harold F. 261, 263 Seaton, Douglass 65 Sebald, W.G. 25, 26, 27, 29, 30, 31, 31, 32, 33, 41, 43, 44 Seidel, Günter 264 Seiffert, Rachel 26, 44 Semino, Elena 131, 133, 149 Sengers, Phoebe 125, 148 Shakespeare, William 74, 74, 75, 75, 76, 77, 79 Sher, Benjamin 149 Shklovskii, Viktor 127, 149 Shopen, Timothy 123, 149 Short, Mick 126, 131, 133, 148, 149 Showalter, Elain x Sigman, Stuart J. 222, 236, 237 Simpson, Rita C. 148 Slobin, Dan 131, 141, 144, 149 Sobchack, Vivian 180, 194 Solbach, Andreas 285, 287 Solshenitsyn, Alexander 118 Speelman, Craig 128, 149 Spellman, Naomi 18, 22 Spielberg, Steven 249 Stafford, Barbara 25, 44 Stanzel, Franz K. xiii, 255 Stein, Malte 95 Steinke, Ines 236, 262 Stempel, Wolf-Dieter 194 Sternberg, Meir xiii Sterne, Laurence 88, 89, 91, 107 Stierle, Karlheinz 189, 194 Stoichita, Victor I. 83, 87, 107 Stone, Addison 226, 237 Stoppard, Tom 94, 95, 102, 103 Strasen, Sven 180, 194 Straus, Florian 286 Sturgess, Philipp J.M. 110, 118, 124 Suetonius 201, 208, 211 Sullivan, J.P. 210 Sutton-Smith, Brian 5, 23 Swales, John 148 Swift, Graham 83, 107 Talbot, Jean 224, 237 Talmy, Leonard 131, 132, 141, 142, 149 Tannen, Deborah 236

296

Author Index

Taylor, John A. 216, 237 Taylor, Michael 22 Taylor, Shelley E. 216, 237 Telemann, Georg Philipp 13 Terrazas, Paula 131, 141, 144, 145, 148 Thamer, Hans-Ulrich 183 Theophilus 198, 199, 200 Thiel, Helmut van 211 Thiele, Hans-Günther 182, 183, 184, 187, 194 Thomas, Jenny 148 Thompson, Kristin 2, 10, 17, 21, 23 Timberlake, Alan 113, 123 Todorov, Tzvetan 38, 40, 44, 50, 63, 109, 124 Tomashevsky, Boris 49, 63 Toolan, Michael 125, 149 Tovey, Donald Francis 75, 75, 81 Töteberg, Michael 105 Tracey, Martin 173 Tribe, Keith 192 Tschuggnall, Karoline 189, 194 Tschuschke, Volker 244, 259, 263 Turner, Mark 153, 172 Ueding, Gert 191, 192 Unamuno, Miguel de 102 Utzschneider, Helmut 209, 211 Vendler, Zeno 139, 149 Voltaire (d.i. François Marie Arouet) 178 Wachowski, Andy 97 Wachowski, Larry 97 Wackwitz, Stefan 26, 44 Wagner, Peter 29, 44, 268, 273, 287 Wagner, Frank 84, 88, 90, 92, 107 Wahlin, Claes 124 Waigel, Theo 187 Waldvogel, Bruno 244, 263 Waletzky, Joshua 278 Warhol, Robyn 148

Warning, Rainer 262 Watt, Ian 216, 234 Waugh, Evelyn 160, 172 Weiher, Anton 210 Weilnböck, Harald 83, 103, 239, 240, 243, 244, 249, 250, 254, 259, 263, 264 Welzer, Harald 189, 194 Wendland, Paul 211 Wenger, Etienne 226, 237 Werner, Heinz 225, 226, 237 Wertsch, James V. 152, 160, 162, 172, 226, 237 Wetherell, Margaret 221, 222, 233, 236 White, Hayden 7, 23, 174, 174, 175, 176, 176, 178, 178, 179, 194, 273, 287 Whiteley, J. 105 Whittall, Mary 79 Wierzbicka, Anna 216, 237 Wiesner, David 95, 107 Wilkins, David 129, 137, 149 Wilson, Andrew 131 Wilson, Robert A. 171 Wingender, M. 124 Wirth, Ilse 105 Wittgenstein, Ludwig xiv Wolf, Werner 3, 4, 9, 14, 23, 31, 32, 44, 66, 70, 81, 83, 84, 86, 88, 90, 92, 92, 96, 103, 107, 217, 237, 242, 261 Wong, Andrew 234 Worth, Sol 9, 23 Wynne, Martin 131, 133, 149 Yevseyev, Vyacheslav 118, 121, 124

6, 109, 117,

Zingrobe, Frank 22 Zubin, David 125, 128, 129, 137, 149 Zunshine, Lisa 161, 172

Contact Addresses Marie-Laure Ryan

[email protected]

Silke Horstkotte

[email protected]

Britta Neitzel

[email protected]

Douglass Seaton

[email protected]

Werner Wolf

[email protected]

Vyacheslav Yevseyev

[email protected]

David Herman

[email protected]

Alan Palmer

[email protected]

Daniel Fulda

[email protected]

Ute Eisen

[email protected]

Michael Bamberg

[email protected]

Harald Weilnböck

[email protected]

Wolfgang Kraus

[email protected]

Narratology Beyond Literary Criticism: Mediality and Disciplinarity (Narratologia - Volume 6)

Narratology Beyond Literary Criticism: Mediality and Disciplinarity (Narratologia)

Literary Criticism

Marxism and Literary Criticism

Textual and Literary Criticism

Marxism and Literary Criticism

Blanchot and Literary Criticism

Literary and Linguistic Approaches to Feminist Narratology

a Literary Criticism

What Is Narratology?: Questions and Answers Regarding the Status of a Theory (Narratologia - Volume 1)

Narratology in the Age of Cross-Disciplinary Narrative Research (Narratologia)

Elementary Guide to Literary Criticism

Classical Literary Criticism (Penguin Classics)

Literary Criticism: A New History

Some Principles of Literary Criticism

Mediating Criticism: Literary Education Humanized

The Cambridge History of Literary Criticism, Vol. 1: Classical Criticism

The Cambridge History of Literary Criticism, Vol. 1: Classical Criticism

Practical Criticism: A Study Of Literary Judgment

American Literary Criticism Since the 1930s

What Is Narratology?: Questions and Answers Regarding the Status of a Theory (Narratologia, 1)

A History of Feminist Literary Criticism

Philosophy, rhetoric, literary criticism: (inter)views

Elemente der Narratologie (Narratologia - Volume 8)

Psychoanalytic Literary Criticism (Longman Critical Readers)

The Bowdler Shakespeare (Cambridge Library Collection - Literary Studies) (Volume 6)

The Cambridge Shakespeare (Cambridge Library Collection - Literary Studies) (Volume 6)

Premediation: Affect and Mediality After 9 11

Literary subversions: new American fiction and the practice of criticism

The Aesthetics of Chaos: Nonlinear Thinking and Contemporary Literary Criticism

Goethe's Werther and the Critics (Literary Criticism in Perspective)

Narratology Beyond Literary Criticism: Mediality and Disciplinarity (Narratologia - Volume 6)