null null

NARRATIVE
THEORY
EDITED BY JAMES PHELAN AND PETER J. RABINOWITZ
A Companion to Narrative Theory
Blackwell Companions to Literature and Culture
This series offers comprehensive, newly written surveys of key periods and movements
and certain major authors, in English literary culture and history. Extensive volumes
provide new perspectives and positions on contexts and on canonical and postcanonical texts, orientating the beginning student in new fields of study and providing the
experienced undergraduate and new graduate with current and new directions, as
pioneered and developed by leading scholars in the field.
1
2
3
4
5
6
7
8
9
10
11
A Companion to Romanticism
A Companion to Victorian Literature and Culture
A Companion to Shakespeare
A Companion to the Gothic
A Feminist Companion to Shakespeare
A Companion to Chaucer
A Companion to Literature from Milton to Blake
A Companion to English Renaissance Literature and Culture
A Companion to Milton
A Companion to Twentieth-Century Poetry
A Companion to Anglo-Saxon Literature and Culture
12
13
14
15
A
A
A
A
Companion
Companion
Companion
Companion
to
to
to
to
Restoration Drama
Early Modern Women’s Writing
Renaissance Drama
Victorian Poetry
16 A Companion to the Victorian Novel
17–20 A Companion to Shakespeare’s Works: Volumes I–IV
21 A Companion to
22 A Companion to
23 A Companion to
American South
24 A Companion to
25 A Companion to
26 A Companion to
27
28
29
30
A
A
A
A
Companion
Companion
Companion
Companion
to
to
to
to
31 A Companion to
32 A Companion to
33 A Companion to
the Regional Literatures of America
Rhetoric and Rhetorical Criticism
the Literature and Culture of the
Edited by Duncan Wu
Edited by Herbert F. Tucker
Edited by David Scott Kastan
Edited by David Punter
Edited by Dympna Callaghan
Edited by Peter Brown
Edited by David Womersley
Edited by Michael Hattaway
Edited by Thomas N. Corns
Edited by Neil Roberts
Edited by Phillip Pulsiano
and Elaine Treharne
Edited by Susan J. Owen
Edited by Anita Pacheco
Edited by Arthur F. Kinney
Edited by Richard Cronin, Alison
Chapman, and Antony H. Harrison
Edited by Patrick Brantlinger
and William B. Thesing
Edited by Richard Dutton
and Jean E. Howard
Edited by Charles L. Crow
Edited by Walter Jost and Wendy Olmsted
Edited by Richard Gray and Owen Robinson
Edited by Shirley Samuels
Edited by Robert Paul Lamb
and G. R. Thompson
Digital Humanities
Edited by Susan Schreibman, Ray
Siemens, and John Unsworth
Romance
Edited by Corinne Saunders
the British and Irish Novel 1945–2000
Edited by Brian W. Shaffer
Twentieth-Century American Drama
Edited by David Krasner
the Eighteenth-Century Novel and Culture
Edited by Paula R. Backscheider
and Catherine Ingrassia
Old Norse-Icelandic Literature and Culture
Edited by Rory McTurk
Tragedy
Edited by Rebecca Bushnell
Narrative Theory
Edited by James Phelan and Peter J. Rabinowitz
American Fiction 1780–1865
American Fiction 1865–1914
NARRATIVE
THEORY
EDITED BY JAMES PHELAN AND PETER J. RABINOWITZ
ß 2005 by Blackwell Publishing Ltd
except for editorial material and organization ß 2005 by James Phelan and Peter
J. Rabinowitz and chapter 33 ß 2005 Peggy Phelan
BLACKWELL PUBLISHING
350 Main Street, Malden, MA 02148-5020, USA
9600 Garsington Road, Oxford OX4 2DQ, UK
550 Swanston Street, Carlton, Victoria 3053, Australia
The right of James Phelan and Peter J. Rabinowitz to be identified as the Authors of the Editorial
Material in this Work has been asserted in accordance with the UK Copyright, Designs, and
Patents Act 1988.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or
otherwise, except as permitted by the UK Copyright, Designs, and Patents Act 1988, without the
prior permission of the publisher.
First published 2005 by Blackwell Publishing Ltd
1 2005
Library of Congress Cataloging-in-Publication Data
A companion to narrative theory/edited by James Phelan and Peter J. Rabinowitz.
p. cm.—(Blackwell companions to literature and culture; 33)
Includes bibliographical references and index.
ISBN-13: 978-1-4051-1476-9
ISBN-10: 1-4051-1476-2
1. Narration (Rhetoric) I. Phelan, James, 1951– II. Rabinowitz, Peter J., 1944– III. Series.
PN212.C64 2005
808—dc22
2004025184
A catalogue record for this title is available from the British Library.
Set in 11/13 pt Garamond 3
by SPI Publisher Services, Pondicherry, India
Printed and bound in India
by Replika Press Pvt. Ltd, Kundli
The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry
policy, and which has been manufactured from pulp processed using acid-free and elementary
chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used
have met acceptable environmental accreditation standards.
For further information on
Blackwell Publishing, visit our website:
www.blackwellpublishing.com
To our students
Contents
Notes on Contributors
Acknowledgments
Introduction: Tradition and Innovation in Contemporary Narrative Theory
James Phelan and Peter J. Rabinowitz
x
xvii
1
Prologue
1 Histories of Narrative Theory (I): A Genealogy of Early Developments
David Herman
19
2 Histories of Narrative Theory (II): From Structuralism to the Present
Monika Fludernik
36
3 Ghosts and Monsters: On the (Im)Possibility of Narrating
the History of Narrative Theory
Brian McHale
60
PART I New Light on Stubborn Problems
73
4
Resurrection of the Implied Author: Why Bother?
Wayne C. Booth
75
5
Reconceptualizing Unreliable Narration: Synthesizing
Cognitive and Rhetorical Approaches
Ansgar F. Nünning
89
Contents
6
Authorial Rhetoric, Narratorial (Un)Reliability,
Divergent Readings: Tolstoy’s Kreutzer Sonata
Tamar Yacobi
vii
108
7
Henry James and ‘‘Focalization,’’ or Why James Loves Gyp
J. Hillis Miller
124
8
What Narratology and Stylistics Can Do for Each Other
Dan Shen
136
9
The Pragmatics of Narrative Fictionality
Richard Walsh
150
PART II Revisions and Innovations
165
10 Beyond the Poetics of Plot: Alternative Forms of Narrative
Progression and the Multiple Trajectories of Ulysses
Brian Richardson
167
11 They Shoot Tigers, Don’t They?: Path and Counterpoint
in The Long Goodbye
Peter J. Rabinowitz
181
12 Spatial Poetics and Arundhati Roy’s The God of Small Things
Susan Stanford Friedman
192
13 The ‘‘I’’ of the Beholder: Equivocal Attachments and the
Limits of Structuralist Narratology
Susan S. Lanser
206
14 Neonarrative; or, How to Render the Unnarratable in
Realist Fiction and Contemporary Film
Robyn R. Warhol
220
15 Self-consciousness as a Narrative Feature and Force:
Tellers vs. Informants in Generic Design
Meir Sternberg
232
16 Effects of Sequence, Embedding, and Ekphrasis in Poe’s
‘‘The Oval Portrait’’
Emma Kafalenos
253
viii
Contents
17 Mrs. Dalloway’s Progeny: The Hours as Second-degree Narrative
Seymour Chatman
269
PART III Narrative Form and its Relationship to History, Politics, and Ethics
283
18 Genre, Repetition, Temporal Order: Some
Aspects of Biblical Narratology
David H. Richter
285
19 Why Won’t Our Terms Stay Put? The Narrative
Communication Diagram Scrutinized and Historicized
Harry E. Shaw
299
20 Gender and History in Narrative Theory: The Problem of Retrospective
Distance in David Copperfield and Bleak House
Alison Case
312
21 Narrative Judgments and the Rhetorical Theory of Narrative:
Ian McEwan’s Atonement
James Phelan
322
22 The Changing Faces of Mount Rushmore: Collective Portraiture and
Participatory National Heritage
Alison Booth
337
23 The Trouble with Autobiography: Cautionary Notes
for Narrative Theorists
Sidonie Smith and Julia Watson
356
24 On a Postcolonial Narratology
Gerald Prince
25 Modernist Soundscapes and the Intelligent Ear:
An Approach to Narrative Through Auditory Perception
Melba Cuddy-Keane
372
382
26 In Two Voices, or: Whose Life/Death/Story Is It, Anyway?
Shlomith Rimmon-Kenan
399
PART IV Beyond Literary Narrative
413
27 Narrative in and of the Law
Peter Brooks
415
Contents
28 Second Nature, Cinematic Narrative, the Historical
Subject, and Russian Ark
Alan Nadel
ix
427
29 Narrativizing the End: Death and Opera
Linda Hutcheon and Michael Hutcheon
441
30 Music and/as Cine-Narrative or: Ceci n’est pas un leitmotif
Royal S. Brown
451
31 Classical Instrumental Music and Narrative
Fred Everett Maus
466
32 ‘‘I’m Spartacus!’’
Catherine Gunther Kodat
484
33 Shards of a History of Performance Art: Pollock
and Namuth Through a Glass, Darkly
Peggy Phelan
499
Epilogue
34 Narrative and Digitality: Learning to Think With the Medium
Marie-Laure Ryan
515
35 The Future of All Narrative Futures
H. Porter Abbott
529
Glossary
542
Index
552
Notes on Contributors
H. Porter Abbott is Professor of English at the University of California, Santa
Barbara. He works primarily in the areas of narrative, modernism, autobiography,
literature and evolutionary theory, and the art of Samuel Beckett. His most recent
book is The Cambridge Introduction to Narrative (2002). He is currently at work on a
study of Darwin, modernism, and representations of the conversion experience.
Alison Booth, Professor of English, has taught at the University of Virginia since
1986. Among her publications are Greatness Engendered: George Eliot and Virginia Woolf
(1992), the edited collection Famous Last Words: Changes in Gender and Narrative
Closure (1993), and articles in Narrative, Victorian Studies, American Literary History,
Kenyon Review, and other journals and essay collections. Her interest in prosopography
sparks her current project on ‘‘homes and haunts’’ and the location of national canons
in authors’ houses.
Wayne C. Booth sees himself now as having spent a lifetime trying to improve
communication, both by teaching and by producing books and articles. That effort
has often led him to the thesis of his chapter here. In most situations, and especially in
reading literature, you cannot fully understand (and thus communicate well) unless
you distinguish the author implied by the text from both the flesh-and-blood author
and the diverse characters and narrators met in any text. Every author and speaker
attempts to create an implied version superior to the everyday Self. And no one who
tries to understand anyone can hope to succeed without listening closely enough to
spot the differences among two – and sometimes three – contrasting personae.
Peter Brooks is author of a number of books, including The Melodramatic Imagination
(1976), Reading for the Plot (1984), Body Work (1993), and Troubling Confessions (2000).
Realist Vision will be published in 2005. For many years professor of comparative
Notes on Contributors
xi
literature and French at Yale University, he is currently University Professor (English
and Law) at the University of Virginia.
Royal S. Brown is a Professor at Queens College in the City University of New York
– where he currently chairs the Department of European Languages and Literatures –
and in the PhD Programs in Music, French, and Film Studies at the City University of
New York Graduate Center. He is the author of Focus on Godard (1972) and Overtones
and Undertones: Reading Film Music (1994) as well as numerous articles and critiques
dealing mostly with film and film music.
Alison Case is a Professor of English at Williams College, Williamstown, MA. She is
the author of Plotting Women: Gender and Narration in the Eighteenth- and NineteenthCentury British Novel (1999) and of several articles on Victorian narratives and
narrative techniques. She is currently collaborating with Harry Shaw on a guide to
nineteenth-century British fiction.
Seymour Chatman is Professor Emeritus of Rhetoric and Film Studies at the
University of California, Berkeley. He is the author of Story and Discourse (1978),
Coming to Terms (1990), Antonioni, or the Surface of the World (1985), and Antonioni: The
Complete Films (2004). His recent articles include discussions of narratology, film
adaptation, parody, and the term ‘‘literary theory.’’
Melba Cuddy-Keane is Professor of English and a Northrop Frye Scholar at the
University of Toronto, and a former President of the International Virginia Woolf
Society. She has published widely on modernist narrative, media, and culture and is
the author of Virginia Woolf, the Intellectual, and the Public Sphere (2003).
Monika Fludernik is Professor of English Literature at the University of Freiburg,
Germany. She is the author of The Fictions of Language and the Languages of Fiction
(1993), Towards a ‘‘Natural’’ Narratology (1996), winner of the 1998 Perkins Prize
from the Society for the Study of Narrative Literature, and Echoes and Mirrorings:
Gabriel Josipovici’s Creative Oeuvre (2000). She has edited a special issue of Style on
second-person fiction, a special issue of EJES on ‘‘Language and Literature, and (coedited with Donald and Margaret Freeman) a special issue of Poetics Today on
‘‘Metaphor and Beyond: New Cognitive Developments.’’ Work in progress includes
a project on the development of narrative structure in English literature between
1250 and 1750.
Susan Stanford Friedman teaches at the University of Wisconsin-Madison. Mappings: Feminism and the Cultural Geographies of Encounter (1998) won the 2000 Perkins
Prize from SSNL. She has also published Psyche Reborn: The Emergence of H.D. (1991),
Penelope’s Web: Gender, Modernity, H.D.’s Fiction (1990), Analyzing Freud: Letters of H.D.,
xii
Notes on Contributors
Bryher, and Their Circle (2002), Joyce: The Return of the Repressed (1993), and numerous
essays on narrative poetics.
David Herman, who is editor of the Frontiers of Narrative book series published by
the University of Nebraska Press, teaches in the English Department at Ohio State
University. He has recently authored Story Logic: Problems and Possibilities of Narrative
(2002), winner of the 2004 Perkins Prize from SSNL, edited Narrative Theory and the
Cognitive Sciences (2003), and coedited (with Manfred Jahn and Marie-Laure Ryan) The
Routledge Encyclopedia of Narrative Theory (2005).
Linda Hutcheon is Distinguished University Professor of English and Comparative
Literature at the University of Toronto. Michael Hutcheon is Professor of Medicine
at the University of Toronto. Together they have written Opera: Desire, Disease, Death
(1996), Bodily Charm: Living Opera (2000) and Opera: The Art of Dying (2004).
Emma Kafalenos teaches comparative literature at Washington University in St
Louis. She has published extensively on narrative theory, often in relation to the visual
arts or music, in journals including Poetics Today, Comparative Literature, 19th-Century
Music, Visible Language, Studies in Twentieth Century Literature, and Narrative. She is the
guest editor of the May 2001 issue of Narrative, devoted to contemporary narratology.
Catherine Gunther Kodat is associate professor of English and American Studies at
Hamilton College, Clinton, NY. She writes on dance, music, cinema, and literature;
her essays have appeared in American Quarterly, Representations, and Mosaic. An essay on
Faulkner and Godard is forthcoming in the Blackwell Companion to William Faulkner.
Susan S. Lanser is Professor of English and Comparative Literature and Chair of the
Women’s Studies Program at Brandeis University. She works in the fields of narrative
theory, gender studies, and eighteenth-century European culture and literature. Her
publications on narrative include The Narrative Act (1981) and Fictions of Authority:
Women Writers and Narrative Voice (1992), as well as many articles and essays.
Fred Everett Maus is an Associate Professor of Music at the University of Virginia.
He has written widely on such subjects as theory and analysis, gender and sexuality,
popular music, aesthetics, dramatic and narrative aspects of instrumental music.
Recent publications include the entries ‘‘Criticism: General Introduction’’ and ‘‘Narratology, Narrativity’’ in the New Grove Dictionary of Music and Musicians.
Brian McHale is Distinguished Humanities Professor in English at The Ohio State
University. He has been associated for many years with the journal Poetics Today, most
recently as coeditor. He is the author of Postmodernist Fiction (1987), Constructing
Postmodernism (1992), and The Obligation Toward the Difficult Whole: Postmodernist
Notes on Contributors
xiii
Long Poems (2004), as well as many articles on modernist and postmodernist poetics,
narratology, and science fiction.
J. Hillis Miller taught for many years at the Johns Hopkins University and then at
Yale University, before going to the University of California at Irvine in 1986, where
he is now UCI Distinguished Research Professor. He is the author of many books and
essays on nineteenth- and twentieth-century English, European, and American literature, and on literary theory. His most recent books are Others (2001), Speech Acts in
Literature (2002), On Literature (2002), and Zero Plus One (2003). He is at work on a
book on speech acts in the novels and stories of Henry James. A J. Hillis Miller Reader
is forthcoming.
Alan Nadel, Professor of literature and film at Rensselaer Polytechnic Institute, Troy,
NY, is the author of numerous books and essays on American literature, film, and
culture. His books include Invisible Criticism (1988), Containment Culture (1995),
Flatlining on the Field of Dreams (1997), and the forthcoming White America in
Black-and-White: Cold War Television and the Legacy of Racial Profiling. His essays
have won prizes from Modern Fiction Studies and PMLA, and his poetry has appeared
in several journals including Georgia Review, New England Review, Paris Review,
Partisan Review, and Shenandoah. The chapter in this volume is excerpted from his
nearly complete work-in-progress: The Historical Performative: Essays on the Cogency of
Narrative Media.
Ansgar Nünning has been Professor and Chair of English and American Literary and
Cultural Studies at the Justus-Liebig-University of Giessen, Germany, since 1996,
following a 10-year period at the University of Cologne. He is the founding director
of the Giessen Graduate School for the Humanities (GGK), and the Project Coordinator of the International PhD study course ‘‘Literary and Cultural Studies.’’ He is
the author of monographs on the structure of narrative transmission and the functions
of the narrator in George Eliot’s novels, historiographic metafiction, the development of the historical novel in England since 1950, and the twentieth-century
English novel. He has edited the Metzler Encyclopedia of Literary and Cultural Theory
(1998), collections of articles on unreliable narration and multiperspectivism, the
Metzler Encyclopedia of English Authors (2002, with Eberhard Kreutzer), Konzepte der
Kulturwissenschaften – Theoretische Grundlagen – Ansätze – Perspektiven (2003) and
Kulturwissenschaftliche Literaturwissenschaft (2004, with Roy Sommer).
James Phelan is Distinguished Humanities Professor in English at Ohio State
University. He is the editor of Narrative and, with Peter J. Rabinowitz, the coeditor
of the Ohio State University Press series on the Theory and Interpretation of
Narrative. He is the author of five books, the most recent of which is Living to Tell
About It: A Rhetoric and Ethics of Character Narration (2005).
xiv
Notes on Contributors
Peggy Phelan is the Ann O’Day Maples Chair in the Arts and Professor of Drama,
Stanford University. She is the author of the Survey essays for Art and Feminism (ed.
Helena Reckitt, 2001) and Pipilotti Rist (2001). She is also the author of Mourning Sex:
Performing Public Memories (1997) and Unmarked: The Politics of Performance (1993). She
coedited The Ends of Performance (1997) with Jill Lane, and with the late Lynda Hart,
she coedited Acting Out: Feminist Performances (1993). Currently she is writing a book
entitled Twentieth Century Performance.
Gerald Prince is Professor of Romance Languages at the University of Pennsylvania.
The author of several books, including Narratology (1982), A Dictionary of Narratology
(1987), and Narrative as Theme (1992), he is editor of French Forum and he coedits the
‘‘Parallax’’ series for the Johns Hopkins University Press as well as the ‘‘Stages’’ series
for the University of Nebraska Press.
Peter J. Rabinowitz, Professor of Comparative Literature at Hamilton College,
divides his time between music criticism and narrative theory. He is the author of
Before Reading (1987) and coauthor, with Michael Smith, of Authorizing Readers
(1998); with James Phelan, he is coeditor of Understanding Narrative (1994) and series
coeditor of the Theory and Interpretation of Narrative series at Ohio State University
Press. He is currently a contributing editor to Fanfare.
Brian Richardson teaches in the English Department of the University of Maryland.
He is the author of Unlikely Stories: Causality and the Nature of Modern Narrative (1997)
and the editor of Narrative Dynamics: Essays on Time, Plot, Closure, and Frames (2002).
He has written articles on many aspects of narrative theory and reader response, and is
currently completing a book on extreme narration and unusual narrators in contemporary fiction.
David H. Richter is Professor of English at Queens College and the Graduate Center,
City University of New York. He is the author of The Progress of Romance (1996) and
Fable’s End (1974) and the editor of The Critical Tradition (1998) and Falling into
Theory (2000). Richter is currently working on two research projects: on indeterminacy in biblical narrative and on a case of identity theft in the late eighteenth century.
Shlomith Rimmon-Kenan is Professor of English and Comparative Literature at the
Hebrew University of Jerusalem, where she also holds the Renee Lang Chair for
Humanistic Studies and currently heads the School of Literatures. Her current
research concerns the concept of narrative in various disciplines (psychoanalysis,
historiography, law), an offshoot of which is her work on illness narratives.
Marie-Laure Ryan, a native of Geneva, Switzerland, is an independent scholar based
in Colorado. She is the author of Possible Worlds, Artificial Intelligence and Narrative
Theory (1991), and of Narrative as Virtual Reality: Immersion and Interactivity in
Notes on Contributors
xv
Literature and Electronic Media (2001), which received the Jeanne and Aldo Scaglione
Prize for Comparative Literature from the Modern Language Association. She is also
the editor of Cyberspace Textuality (1999) and Narrative Across Media (2004), as well as
a coeditor, with David Herman and Manfred Jahn, of the Routledge Encyclopedia of
Narrative (2004).
Harry E. Shaw is Professor of English at Cornell University, where he has served as
Chair of the Department of English and is now Senior Associate Dean of the College
of Arts and Sciences. He is the author of The Forms of Historical Fiction: Scott and his
Successors (1983) and Narrating Reality: Austen, Scott, Eliot (1999). Essays on Scott, J. L.
Austin, and narrative theory have appeared in such journals as JEGP, diacritics,
Narrative, and European Romantic Review. He is currently at work on a book on the
nineteenth-century British realist novel for a general audience, in collaboration with
Alison Case.
Dan Shen is Professor of English and Director of the Center for European and
American Literatures at Peking (Beijing) University. Apart from her numerous
books and essays published in China, she has published more than 20 essays in
North America and Europe on narrative theory, stylistics, literary theory, and translation studies.
Sidonie Smith is Martha Guernsey Colby Collegiate Professor of English and
Women’s Studies and Chair of the Department of English at the University of
Michigan. As well as the works coauthored and coedited with Julia Watson, she is
the author of five other books on autobiography, the most recent of which is Human
Rights and Narrated Lives: The Ethics of Recognition (with Kay Schaffer, 2004).
Meir Sternberg is Artzt Professor of Poetics and Comparative Literature at Tel-Aviv
University and Editor of Poetics Today. His publications include Expositional Modes and
Temporal Ordering in Fiction (1978), The Poetics of Biblical Narrative: Ideological Literature
and the Drama of Reading (1985), Hebrews Between Cultures: Group Portraits and National
Literature (1998), and the two-part interdisciplinary review ‘‘Universals of Narrative
and their Cognitivist Fortunes’’ for Poetics Today.
Richard Walsh is a lecturer in English and Related Literature at the University of
York, UK. His publications include Novel Arguments: Reading Innovative American
Fiction (1995), and articles on narrative in Poetics Today, Style, and Narrative.
Robyn R. Warhol, Professor of English at the University of Vermont, has published
Having a Good Cry: Effeminate Feelings and Popular Forms (2003), Feminisms (1997), and
Gendered Interventions: Narrative Discourse in the Victorian Novel (1989).
xvi
Notes on Contributors
Julia Watson is Associate Professor of Comparative Studies at Ohio State University.
With Sidonie Smith she has coauthored Reading Autobiography: A Guide for Interpreting
Life Narratives (2001) and coedited four collections of essays, most recently Interfaces:
Women, Autobiography, Image, Performance (2002).
Tamar Yacobi teaches at Tel-Aviv University. Her primary interests include narration, (un)reliability, reading, ekphrasis, the dramatic monologue, and Isak Dinesen.
Recent publications include ‘‘Ekphrasis and Perspectival Structure,’’ in Cultural
Functions of Intermedial Exploration (ed. Erik Hedling and Ulla-Britta Lagerroth,
2000) and articles in Poetics Today and Narrative.
Acknowledgments
The editors would like to thank, first, all of our contributors because they have made
working on this volume a consistent source of intellectual pleasure and energy. We also
would like to thank several people who have provided invaluable editorial advice: Jamie
Abaied, Elizabeth Jensen, Jessica Kent, Nancy Rabinowitz, Greta Rosenberger, and
James Weaver. Aaron McKain has earned our eternal gratitude through his patient and
thorough work on the index. We are grateful as well to Andrew McNeillie, who
recruited us to edit this Companion, and to Jennifer Hunt, who guided us in the first
stages of its preparation. We owe a special thanks to Emma Bennett and Karen Wilson,
who so ably guided the manuscript through production. Above all, we are indebted to
Jenny Roberts, copy-editor extraordinaire.
The editors and publishers are also grateful to the following for their kind
permission:
‘‘The Edge’’ by Sylvia Plath, reprinted by permission of Harper Collins Publishers
Inc. Copyright ß 1963 Ted Hughes.
‘‘A Time to Talk’’ by Robert Frost, reprinted by permission of Henry Holt and
Company. Copyright 1916, 1969.
Figure 22.1 Portrait Monument of Lucretia Mott, Elizabeth Cady Stanton, and
Susan B. Anthony, Architect of the Capitol, <http://www.aoc.gov/cc/art/rotunda/
suffrage.htm>.
Figure 22.2 Mount Rushmore, used by permission of the US Government
Printing Office, <http://bensguide.gpo.gov/3-5/symbols/mountrushmore.html>.
Figure 22.3 The English Romantic Poets, used with kind permission of James
Sullivan of Ard Ri Design, <http://www.ardridesign.com/gallery.html>.
Figure 22.4 Irish Writers Montage, used with kind permission of James Sullivan
of Ard Ri Design, <http://www.ardridesign.com/gallery.html>.
Introduction: Tradition and
Innovation in Contemporary
Narrative Theory
James Phelan and Peter J. Rabinowitz
Writing an introduction to a wide-ranging collection of essays is always a matter of
navigating between Scylla and Charybdis – but in this book, the metaphor has more
specificity than usual. Scylla, you’ll recall, was a monster who inhabited a cave in a
cliff on a spit of land jutting off from the coast; her many arms plucked sailors from
the boats that came too near. Charybdis was a whirlpool that threatened any boats that
tried to evade the cliff. Elements of those images impinge on this introduction – and
on narrative theory itself – in several ways.
How, for instance, should the introduction be structured? On the one hand, we can
aim for clarity, simply summarizing the contents. Such a Cliffs Notes approach may
provide a certain stony intelligibility, but it runs the risk of plucking the spirit from
both the essays and our readers. On the other hand, we’re faced with the everexpanding whirlpool of self-reflexivity. An introduction, after all, can itself easily
become a narrative – and a narrative about the current study of narrative (which in
turn includes a history of the study of narrative) risks spinning into an endless loop,
especially when written by authors who are, by disciplinary training, acutely selfconscious.
The task is made more difficult still because Scylla and Charybdis are not simply
potential dangers to be evaded. Like the Sirens (and it’s appropriate that the volume
includes three essays on music and one that deals tangentially with musical issues),
they offer seductions as well. Indeed, one can argue that the discipline of narrative
theory itself divides into two attractive and productive ways of doing things, ways
parallel to the distinction between the cliff and the whirlpool. On the one hand, we
have the search for a stable landing, a theoretical bedrock of the fundamental and
unchanging principles on which narratives are built. This approach is often associated
with what is called structuralist (or classical) narratology, and especially after the rise
of post-structuralism, it is often viewed as old-fashioned, even quaint – and it is often
2
James Phelan and Peter J. Rabinowitz
believed to yank the life out of the works it considers. But as we hope this collection
will make clear, it’s still an enormously vital area of study, and it’s still producing
illuminating work, though nowadays its claims tend to be more modest – about
‘‘most narratives’’ or ‘‘narratives of a certain historical period’’ rather than about ‘‘all
narratives.’’
On the other hand, we have a discipline experiencing a voracious spin. That
whirlpool is generated in part by the self-conscious and self-critical nature of much
narrative theory. What the two of us have in the past labeled ‘‘theorypractice,’’ for
instance, uses the interpretive consequences of particular theoretical hypotheses as a
way of testing and re-examining those very hypotheses. But the vortex of narrative
theory comes as well from what’s often called the ‘‘narrative turn,’’ the tendency of the
term ‘‘narrative’’ to cover a wider and wider territory, taking in (some would say
‘‘sucking in’’) an ever-broadening range of subjects for inquiry. Narrative theory, over
the years, has become increasingly concerned with historical, political, and ethical
questions. At the same time, it has moved from its initial home in literary studies to
take in examination of other media (including film, music, and painting) and other
nonliterary fields (for instance, law and medicine).
It should therefore be no surprise that our volume represents both Scylla and
Charybdis, both the search for enduring fundamentals of narrative theory and the
engagement with the many turnings of contemporary theory. True, the introduction
steers, on the whole, closer to the cliff, largely because so many of the essays
themselves offer sufficient spin to satisfy anyone in search of the pleasures of the
whirlpool. Nonetheless, we are sufficiently self-conscious to warn you that, if writing
this introduction involved navigating a familiar but tricky path, so too does reading
it. For reasons that will become increasingly clear, we urge you to recognize that our
navigational choices were not inevitable. You very well might have charted a different
route – though of course you can’t choose that route until you’ve read much more than
this introduction.
The book opens with a prologue that sets out narrative theory’s modern history. In the
first essay, ‘‘Histories of Narrative Theory (I): A Genealogy of Early Developments,’’
David Herman surveys the origins of the field. But since Herman is strongly
influenced by the notion of ‘‘genealogy’’ promoted by Nietzsche and Foucault, his
essay refuses to move in a simple linear fashion. He is particularly concerned with the
way early structuralist narratology – an attempt to study narrative by treating
‘‘particular stories as individual narrative messages supported by a shared semiotic
system’’ – grew out of ‘‘a complex interplay of intellectual traditions, criticotheoretical movements, and analytic paradigms distributed across decades, continents,
nations, schools of thought, and individual researchers.’’ Herman uses Wellek and
Warren’s influential Theory of Literature as a nodal point, moving in and out from there
to show the overlapping connections among a wide range of superficially competing
critics who nonetheless represent, if not ‘‘a singular continuous tradition of research,’’
then at least ‘‘a cluster of developments marked by family resemblances.’’ One of
Introduction
3
Herman’s key points is that what can be assimilated from any theorist’s work at any
given time is fundamentally dependent on the paradigms in force at the time of
reading. His history thus twists back on itself, as he demonstrates how ‘‘old’’
theoretical works in ‘‘new’’ contexts take on new resonances.
In ‘‘Histories of Narrative Theory (II): From Structuralism to the Present,’’ Monika
Fludernik likewise resists a linear chronology in which gradual developments contribute to an ever-more complete critical arsenal. Rather, Fludernik self-reflexively
applies narrative theory to the history of narrative theory itself, setting out two
competing ‘‘plots.’’ She then follows the second of these to show how narrative theory
has spread out in widening branches. Specifically, she moves from formalist study to
pragmatics (including issues of gender and politics) on to the study of media and of
the narrative turn in a variety of social sciences, coming finally to issues of linguistics
and cognition. Her essay dexterously ends with both a return to its opening and a
possible glimpse of the future.
The prologue ends with Brian McHale’s ‘‘Ghosts and Monsters: On the
(Im)Possibility of Narrating the History of Narrative Theory.’’ Here McHale questions
the work of Herman and Fludernik by using the insights of narrative theory to
question the task we assigned to them. Specifically, McHale defines two different
kinds of history – what he calls ‘‘history of ideas’’ and ‘‘institutional history’’ – and
argues that the friction between them makes a true history of narrative theory an
impossibility. In even more provocative terms, McHale suggests that there is
an irreconcilable opposition between narrative theory that privileges ‘‘structure’’ (the
Scylla of classical narratology) and narrative theory that privileges ‘‘history’’
(the Charybdis of narrative turnings). Though the necessity of sequence dictates that
McHale’s essay be last in the prologue, that does not mean that you should take his
word as final. Our own position is that the prologue is a provocation that opens up
more questions than it settles.
Part I, ‘‘New Light on Stubborn Problems,’’ looks at some of the disputes that have
continued as an undertow in the flow of narrative theory for the past 40 or 50 years –
disputes that therefore remain central to its continued movement. In the first essay,
‘‘Resurrection of the Implied Author: Why Bother?,’’ Wayne C. Booth – already
introduced in Herman’s and Fludernik’s histories – returns to a concept that he first
introduced in The Rhetoric of Fiction in 1961: the implied author, the author’s ‘‘second
self’’ as reflected in a text. Many subsequent theorists (including several contributors
to this volume) have argued that the concept is either useless or redundant; Booth, in
contrast, believes that it’s more important than ever. Starting from a reconsideration
of the concerns that led him to develop the idea in the first place, Booth goes on
to show how the concept improves our understanding, not only of fiction (for which
it was developed), but also of poetry and of the way we present ourselves in our dayto-day interactions.
The next two essays deal, in different ways, with another key narratological term
introduced by Booth in The Rhetoric of Fiction: the ‘‘unreliable narrator.’’ The first is by
4
James Phelan and Peter J. Rabinowitz
Ansgar Nünning, one of the critics dissatisfied with Booth’s concept of the implied
author. In ‘‘Reconceptualizing Unreliable Narration: Synthesizing Cognitive and
Rhetorical Approaches,’’ Nünning uses that dissatisfaction as an entry into a detailed
account of the battles that have erupted over unreliable narration, with particular
attention to the question of how readers in fact recognize unreliability when they
come across it in the text. He argues that unreliability cannot be defined simply in
terms of the text’s ‘‘structural or semantic’’ aspects; it also involves the ‘‘conceptual
frameworks’’ brought to the text by its readers. More generally, he contends that
an adequate model of unreliability needs to combine the latest insights offered by
the apparently divergent arguments of rhetorical and cognitive narrative theorists.
He ends his essay, provocatively, with a series of six questions that remain to be
answered.
If Nünning hopes to reduce the difficulties involved in accounting for unreliable
narration by developing a new synthetic theory, Tamar Yacobi (‘‘Authorial Rhetoric,
Narratorial (Un)Reliability, Divergent Readings: Tolstoy’s Kreutzer Sonata’’) offers a
somewhat different perspective on the same terrain by moving up to the next level
of generality. Yacobi argues that unreliability is best understood as a ‘‘readinghypothesis’’ that allows readers to resolve apparent textual contradictions. She
makes her case by looking at the wide disparities that separate different readings of
Tolstoy’s novella. After grouping these disparate readings into a few sets with
common features, she goes on to show how all these groups are in turn generated
by the operation of a small number of shared integrative mechanisms that readers use
to deal with incongruities that arise when processing a text.
The essays by J. Hillis Miller and Dan Shen take up a different complex of
problems that have long puzzled narrative theorists, problems that often show up
in terms of the relation of form to content or style to meaning – or form to substance,
as Miller puts it in his close reading of Henry James’s The Awkward Age (‘‘Henry
James and ‘Focalization,’ or Why James Loves Gyp’’). Although many narratologists
have argued that the purpose of theory is not to produce new readings, Miller insists
that narratological distinctions are only valuable if they serve interpretation – indeed,
more radically in an age that often seems to resist judging interpretations, that they
‘‘are useful only if they lead to better readings or to better teachings of literary works.’’
A good reading for Miller, though, is not necessarily a simple or a stable one: it can
well end in a whirlpool of its own. His patient, theoretically sophisticated analysis
demonstrates how, in this formally anomalous narrative (anomalous, at least, within
the James canon), James succeeds in breaking down the distinction between substance
and form and how the ‘‘right reading’’ of the text turns out to be one that reaches
‘‘undecidability’’ as its conclusion.
Shen, in ‘‘What Narratology and Stylistics Can Do for One Another,’’ starts by
looking at a familiar assumption: there is a rough equivalence between the story/
discourse distinction central to narratological thinking (the distinction between
‘‘what’’ is told and ‘‘how’’ it is told) and the content/style distinction (what is
expressed and how it is expressed) central to much stylistic thinking. Using a brief
Introduction
5
interchapter from Ernest Hemingway’s In Our Time as her case in point, she goes on to
show that the similarities are only superficial – and that a full understanding of style
in narrative requires an interdisciplinary approach that synthesizes the insights of
both narratology and stylistics.
The final essay in Part I, Richard Walsh’s ‘‘The Pragmatics of Narrative Fiction,’’
turns to one of the most enduring, and most vexing, problems of narrative theory –
the nature of fiction. Walsh surveys the history of attempts to explain what fiction is
and how it operates, pointing out that, for all their rich variety, ‘‘modern accounts of
fictionality generally turn upon one or more of a small repertoire of theoretical
gambits.’’ In addition, as he demonstrates through patient exploration of central
modern accounts, especially those based in speech act theory and possible worlds
theory, all these maneuvers reduce to various kinds of displacement ‘‘by detaching the
fictive act from the domain of truth.’’ (Despite their differences, there are interesting
methodological similarities between Walsh’s and Yacobi’s essays.) As an alternative,
Walsh offers a pragmatic account of fictionality in which relevance, rather than truth,
becomes the key term, and he demonstrates its explanatory power by analyzing the
opening of Kafka’s The Trial.
Part II, ‘‘Revisions and Innovations,’’ groups together essays that offer significantly
new views of some basic concepts in narrative theory. The authors arrive at their fresh
takes on these concepts in different ways: some focus on narratives that cannot be
adequately addressed with our existing understandings, others focus on the logic of
theory itself or on the uncomfortable fit between theory and readerly experience, and
some use a combination of these methods. All of these essays engage in ‘‘theorypractice’’; in this respect, the section provides the self-conscious reader not only with some
provocative new theorizing but also an implicit primer in how to revise existing
theory.
Focusing primarily on James Joyce’s Ulysses and bringing in other avant-garde
narratives of the twentieth century, Brian Richardson deftly moves ‘‘Beyond the
Poetics of Plot’’ to consider ‘‘Alternative Forms of Narrative Progression.’’ Richardson
begins by pointing out that the dominant models, developed by R. S. Crane, Paul
Ricoeur, Peter Brooks, and others, conceive of plot as a sequence of logically connected events involving coherent characters that provides, in Brooks’s words, ‘‘design
and intention’’ for a narrative. But Richardson casts doubt on these models by
pointing to a tradition of novel writing that rejects this conception, moving its reader
from beginning to end through radically different logics. Richardson then delivers the
big payoff: an insightful survey into the range of these alternatives, starting with
the multiple principles of progression in Ulysses and ending with what he calls the
aleatory progressions of the Dadaists and their descendants such as William
Burroughs.
In ‘‘They Shoot Tigers, Don’t They? Path and Counterpoint in The Long Goodbye,’’
Peter J. Rabinowitz’s primary concern is with time – and our tools for accessing its
representation. Rabinowitz starts by puzzling over similarities and differences
6
James Phelan and Peter J. Rabinowitz
between treatments of time in music and narrative, and he ends by proposing that
narrative theory needs a new concept, one that refers to the order in which a character
experiences the events of a narrative. He calls this concept path, and he sees it as
supplementing the well-established distinction between ‘‘story’’ or ‘‘fabula’’ (referring
to the chronological order of events) and ‘‘discourse’’ or ‘‘sjuzhet’’ (referring to the order
of presentation of those events). We need the concept, he argues, because characters
may experience events neither in the story order nor in the discourse order, and that
difference can matter for readers. Rabinowitz demonstrates the interpretive value of
this concept through a revisionary reading of Raymond Chandler’s The Long Goodbye,
carefully tracing the story order, the discourse order, and the paths of Philip Marlowe
and the other characters involved in the ‘‘tiger trap’’ episode. Recognizing the
counterpoints among the different paths not only leads us to reinterpret the significance of that episode but also exposes Marlowe’s failure to understand it, a revelation
that overturns the standard readings of the novel.
In moving from Rabinowitz’s essay to Susan Stanford Friedman’s ‘‘Spatial Poetics
and Arundhati Roy’s The God of Small Things,’’ we shift attention from time to space.
Friedman observes that narrative theory has privileged time over space, and she wants
to redress the balance – not by demoting time but by restoring space ‘‘to its full
partnership with time as a generative force for narrative.’’ Drawing on other theorists
who have advocated more attention to space, such as Edward Soja, Mikhail Bakhtin,
Franco Moretti, and Lawrence Grossberg, she proposes that we stop regarding space as
a static background and begin recognizing it as an ‘‘active, mobile, and ‘full’ ’’
component of narrative. More specifically, Friedman proposes that we pay more
attention to the role of borders and border crossings in the generation, development,
and resolution of narratives. Exploring the consequences of her spatial poetics through
an energetic and insightful reading of Roy’s novel, Friedman shows how the spaces in
the novel contain ‘‘multiple borders of desirous and murderous connection and
separation, borders that are continually erected and transgressed in movement that
constitutes the kinetic drive of the plot.’’
Susan S. Lanser’s essay, ‘‘The ‘I’ of the Beholder: Equivocal Attachments and the
Limits of Structuralist Narratology,’’ makes an interesting companion piece to Walsh’s
as it takes a fresh look at the well-known assumption that the ‘‘I’’ of fiction is different
from the author, while the ‘‘I’’ of nonfiction is identical to the author. Arguing that
this assumption oversimplifies the situation, she proposes a more complex scheme
with three primary categories. Attached texts such as letters to the editor and
scholarly essays are those in which the primary ‘‘I’’ and the author are identical.
Detached texts are those in which the primary ‘‘I’’ and the author are not identical
(e.g., the Pledge of Allegiance; any fiction with an unreliable narrator) or in which the
relation between the two is not consequential for textual meaning (e.g., a joke, a
national anthem). Equivocal texts are those in which the primary ‘‘I’’ is both associated with and distinct from the author, moving between being attached and detached.
Novels and poems are typically equivocal texts. Lanser builds on this taxonomy by
arguing that genres such as lyric and narrative have default conditions – lyrics are
Introduction
7
attached, while novels are detached – but that under many conditions these default
lines are transgressed. Readers intuitively know how to navigate these fault lines, but
theorists have rarely paid attention to how they manage to do so. Lanser’s essay, with
its suggestive analyses of how we read such different works as Sharon Olds’s poem
‘‘Son,’’ Ann Beattie’s short story ‘‘Find and Replace,’’ and Philip Roth’s novel The
Human Stain, provides a theoretical scheme that accounts for these interpretive
practices.
In ‘‘Neonarrative; or, How to Render the Unnarratable in Realist Fiction and
Contemporary Film,’’ Robyn Warhol offers a different kind of taxonomy, one based
on the conventions governing what can and cannot be represented in narrative.
Warhol examines the phenomena of the ‘‘disnarrated’’ and ‘‘the unnarrated’’ as
instances of the larger category of the ‘‘unnarratable’’ in order to identify genre
conventions and changes in those conventions. The disnarrated, first identified by
Gerald Prince, is the narration of something that might have happened or was
imagined to have happened but did not actually happen. The unnarrated is the lack
of narration about something that did happen; it can be found in those passages in a
narrative ‘‘that explicitly do not tell what is supposed to have happened, foregrounding the narrator’s refusal to narrate.’’ Both the disnarrated and the unnarrated are, for
Warhol, strategies for representing the unnarratable, which she divides into four
types: the subnarratable (that which is taken for granted and so not worthy of
narration); the supranarratable (that which is ineffable); the antinarratable (that
which social convention labels as unacceptable for narration); and the paranarratable
(that which formal conventions render unnarratable). Building on these categories,
Warhol shows how, over time, Hollywood films have created what she calls ‘‘neonarratives’’ by transgressing the formal or social conventions governing the unnarratable.
In ‘‘Self-consciousness as a Narrative Feature and Force: Tellers vs. Informants in
Generic Design,’’ Meir Sternberg turns to a neglected feature of narrative representation, self-consciousness – not just of narrators but also of characters. Sternberg makes
three crucial claims: (1) self-consciousness should be mapped along a continuum
from total to totally absent, from what he calls tellers who are fully aware of audiences
other than themselves to informants whose only audience is themselves; (2) selfconsciousness is always mediated – the narrator’s mediated by the author, the
characters’ by the narrator and the mimetic situation; and (3) the significance of
self-consciousness for narrative’s form and functioning has not yet been appreciated.
Sternberg offers a careful elaboration and exemplification of these points, including an
analysis of why the phenomenon has not yet been given its due. The result is a
persuasive case that the phenomenon of self-consciousness – in particular, the unselfconscious end of the continuum – deserves much more attention.
In ‘‘Effects of Sequence, Embedding, and Ekphrasis in Poe’s ‘The Oval Portrait,’ ’’
Emma Kafalenos offers another approach to understanding the structure of narrative,
as she employs function analysis to explore the interrelation of the three terms of her
title in Poe’s story. Revising models proposed in the work of Vladimir Propp and
Tzvetan Todorov, Kafalenos’s function analysis focuses on how events in narrative
8
James Phelan and Peter J. Rabinowitz
proceed through a series of five stages: equilibrium, disruption, efforts by characters
(or actants) at alleviating the disruption, the success or failure of those efforts, and
finally the establishment of a new equilibrium. In examining the Poe story, which
involves both ekphrasis (the representation of a visual composition in a literary work)
and an embedded narrative, Kafalenos makes good on her contention that ‘‘function
analysis demonstrates the magnitude of the effect that the way events are told can
have on interpretations of the causes and effects of those events.’’ More specifically,
Kafalenos shows how the final paragraph in Poe’s story, which narrates the death of a
young girl, has one effect in a function analysis of the embedded narrative and
another, different, effect in a function analysis of the story as a whole. Furthermore,
she identifies the factors that contribute to this difference, paying special attention to
the role of ekphrasis. This insightful reading of ‘‘The Oval Portrait’’ thus serves as a
model for the way we understand (1) the link between the events of an embedded
narrative and those in its frame, and (2) the role of ekphrasis in narrative.
Kafalenos focuses on one kind of ‘‘intratextuality,’’ as she examines the interaction
of several texts represented or embedded within a single story. Seymour Chatman’s
essay, ‘‘Mrs. Dalloway’s Progeny: The Hours as Second-degree Narrative,’’ shifts attention to intertextuality. His exploration builds on Gérard Genette’s work in Palimpsests
on second-degree texts (texts that are explicitly intertextual with a prior, source text).
After sorting out the different kinds of relations a second-degree narrative can have to
its source, Chatman turns to Michael Cunningham’s The Hours, his homage to
Virginia Woolf’s Mrs. Dalloway. Chatman illuminates Cunningham’s novel by considering it as what Genette calls a ‘‘transposition,’’ a work that invents new characters
but uses the source text as a basis for either revisions or patterns. More specifically,
Chatman treats The Hours as a ‘‘complement’’ to Mrs. Dalloway, a narrative that does
not seek to displace the source even as it transforms it. Within this framework,
Chatman focuses on the ‘‘how’’ of the transposition, tracing many relations between
elements of the two novels. Because Chatman attends both to similarities and to
differences, he is able to show Cunningham’s indebtedness to Woolf as well as
Cunningham’s creative use of his source. More generally, Chatman’s analysis works
to illuminate the specific craft of each novel – and the value of his approach to seconddegree narratives.
Part III brings together essays that consider ‘‘Narrative Form and Its Relationship to
History, Politics, and Ethics.’’ This section moves from biblical narrative to contemporary medical narrative, that last essay providing a bridge to Part IV’s move ‘‘Beyond
Literary Narrative.’’
David Richter’s ‘‘Genre, Repetition, Temporal Order: Some Aspects of Biblical
Narratology’’ begins with an acknowledgment of the challenges biblical narrative
presents to narratological models: it is not written by clearly identifiable authors, its
coherence cannot be assumed, and, indeed, its generic identity is often unclear. What,
then, can narratology offer to biblical studies? Richter answers that a flexible,
historically informed application of the concepts of genre, repetition, and temporal
Introduction
9
order can yield significant interpretive dividends. He starts by showing that different
readings of the book of Jonah stem from different assumptions about its genre.
Richter defends the view that it is satirical fable, but he emphasizes that such a
generic classification is likely to meet resistance because it entails accepting the idea
that some parts of the Bible are fiction. With his second example, the multiple uses of
repetition in the first book of Samuel, Richter works through their many possible
interpretations to arrive at the significant role of the redactor. He then proposes that
the repetitions end up making Samuel 1 a precursor of Rashomon or Faulkner’s
Absalom, Absalom, in which the motives of the tellers take on greater interest than
the story they tell. Richter’s third example is the temporal order of the second book of
Samuel; here late revelations lead to a radical reconfiguration of our understanding of
David’s character. These narratologically grounded interpretations help explain,
Richter argues, why religions have preferred to focus on ‘‘pericopes’’ (short segments
detached from context) rather than the Bible as a continuous narrative whole.
Questions about the historical stability of narratological categories continue in
Harry E. Shaw’s ‘‘Why Won’t Our Terms Stay Put?: The Narrative Communication
Diagram Scrutinized and Historicized.’’ Shaw argues for the necessary relation among
our understanding of basic tools of narrative theory, our larger conceptions of narrative, and the history of narrative. Focusing on the famous diagram that charts the
movement of narrative communication from real author through implied author,
narrator, text, narratee, implied reader, and, finally to real reader, Shaw begins by
showing how our different interests give different meanings to elements of the
diagram. Those committed to a rhetorical view of narrative find the concept of
the implied author a necessary element, while those committed to a view of narrative
as information find the concept unnecessary. (In this part of his argument, Shaw is
offering his own intervention in the debate about implied authors raised at
the beginning of the volume in the essays by Booth and Nünning.) Developing the
distinction between rhetoric and information further, Shaw notes that Genette’s Law –
the dictum that as we move across the diagram from tellers to audiences, the entities
become less substantive – is especially relevant for the rhetorician: while informationoriented theorists can give solidity to the narratee by restricting it to its observable
presence in the text, rhetoricians are better served by folding their insights about
the narratee back into their understanding of the narrator. After demonstrating the
consequences of this view for his reading of Thackeray’s narrator in Vanity Fair, Shaw
makes a self-conscious turn, showing the historical inflection of his own argument. He
notes that his theoretical predilections are part and parcel of his interest in accounting
for a specific period in the history of narrative, the nineteenth-century British novel.
Alison Case’s ‘‘Gender and History in Narrative Theory: The Problem of Retrospective Distance in David Copperfield and Bleak House’’ continues Shaw’s focus on the
necessary connection between form and history in the nineteenth-century British
novel, though her view of the terrain is different because she looks through the lens of
feminist narratology. Her essay, in effect, adds a missing historical dimension to James
Phelan’s account of ‘‘paradoxical paralipsis’’ in Narrative as Rhetoric. Phelan explains
10
James Phelan and Peter J. Rabinowitz
that the technique is in play when the narration of a naı̈ve character narrator is not
informed by the changes that his or her narrative eventually reveals. Phelan also
contends that the technique, though it violates strict mimesis, is often effective
because it allows the author to render the character’s change with the most emotional
force. Case argues that Phelan’s account works for twentieth-century cases but that in
the nineteenth century the technique is often part of ‘‘a gendered literary code.’’
Specifically, paradoxical paralipsis marks a narrator as ‘‘feminine’’ – which does not
necessarily mean ‘‘female’’ – by revealing a lack of authoritative control over the
narrative. Case develops her argument by contrasting Dickens’s different approaches
to the same problem in Bleak House and in David Copperfield: how to use a retrospective character narrator to represent sympathetically that figure’s prior naı̈ve consciousness. In Esther Summerson’s narration, Dickens uses paradoxical paralipsis and other
devices that signal Esther’s lack of narrative mastery, whereas in David’s narration, he
uses commentary from the narrator’s perspective that calls attention to David’s
mastery. Furthermore, Case argues, in establishing these links among gender, technique, and narrative mastery, Dickens is typical of his age.
James Phelan’s ‘‘Narrative Judgments and the Rhetorical Theory of Narrative: Ian
McEwan’s Atonement’’ shifts from a focus on form and history to a focus on form and
ethics. Phelan’s major theoretical claim is that the concept of narrative judgment is
central to a rhetorical understanding of narrative form, narrative ethics, and narrative
aesthetics, because judgment functions as the hinge that allows each domain to open
into the other two. Phelan develops the claim through his identification of three kinds
of judgments (interpretive, ethical, and aesthetic), his articulation of six theses about
their interrelation, and his analysis of Atonement, a complex novel that intertwines the
problems of ethical judgments by its characters with the problems of readers’ ethical
judgments of its storytelling. Atonement is overtly concerned with the relation between transgression and atonement as it shows, first, 13-year-old Briony Tallis’s wellmeaning misidentification of her sister’s lover as a sexual assailant and, second,
Briony’s realization of her mistake and her efforts to make amends. But after showing
Briony on the verge of atonement, McEwan reveals that he has encouraged a misidentification on his audience’s part: this novel we have been reading is not only his
but Briony’s. Furthermore, within the world of both their novels, Briony’s error has
been real but her atonement is pure fiction: her sister and her lover were never
reunited and, in fact, died before such a reunion could even be possible. Thus we
need to come to terms both with the ethics of Briony’s decision to fictionalize her past
and with the ethics of McEwan’s misidentification of the narrative we have been
reading. Phelan contends that McEwan guides us to see Briony’s justification as
ethically and aesthetically deficient even as he succeeds in making his own misidentification increase the aesthetic and ethical power of the novel.
The next two essays deal with lifewriting. Alison Booth’s ‘‘The Changing Faces of
Mount Rushmore: Collective Portraiture and Participatory National Heritage’’ investigates the connection between portraits and biographies and between collective
biography and politics. Her method is to focus on decisions about whose portrait/
Introduction
11
biography gets included in galleries of literary figures and at sites designed to
memorialize significant parts of American history such as Mount Rushmore and the
Hall of Fame of Great Americans. In each case, Booth traces a complex interaction
among the choice of an individual portrait (and its implied biography), the development of a collective portrait (for which she uses the term ‘‘prosopography’’), and the
significance of that collective portrait for the larger national community. Prosopography inevitably raises the political question of representativeness, of who gets
included or excluded and why. Consequently, as Booth notes in her summary,
prosopography involves its ‘‘presenters as well as the audience in collective memorial
representation, claiming a certain kinship in cultural heritage, forming a conjunction
of biography and history, and leaving a palpable afterimage of what is missing.’’
Sidonie Smith and Julia Watson continue the focus on lifewriting in ‘‘The Trouble
with Autobiography: Cautionary Notes for Narrative Theorists’’ – but they do so by
raising significantly different kinds of questions. They explore four particular trouble
spots in autobiographical practice that deserve more attention from narrative theory.
First, Smith and Watson analyze different motives for and effects of hoaxes (and
claims about hoaxes), though they conclude this section by underlining how, whatever
the motives, an autobiographical hoax induces a feeling of betrayal in its reader. Their
discussion of the second trouble spot, the postcolonial play with the borders of fiction
and nonfiction, includes both claims for autobiography in telling the story of another
and claims for fiction in telling the story of oneself. Such experimentation, Smith and
Watson contend, interrogates ‘‘the complicity of the autobiography canon and its
critics with dominant modes of self-representation and truth-telling.’’ Next they turn
to autobiographical narratives that bear witness to human rights abuses, which often
place the autobiographical subject in the position of speaking for the collective and of
appealing to the audience for recognition. These appeals, in turn, place a particular
ethical burden on the audience since recognizing the writer’s experience also entails
doing something with that recognition. Finally, in a discussion of materiality, Smith
and Watson note that contemporary mixed media autobiography raises questions
about the links among the body, the medium of representation, and the autobiographical subject. Like so many essays in the collection, this one ends in a kind of selfconscious spin: all four trouble spots, Smith and Watson contend, simultaneously
receive illumination from and offer challenges to contemporary narrative theory.
Gerald Prince picks up one thread of Smith and Watson’s discussion of postcolonial
lifewriting and weaves it into a larger essay ‘‘On a Postcolonial Narratology.’’ Prince
explains that his version of this hybrid ‘‘would . . . adopt and rely on the results of
(post)classical narratology but would inflect it and perhaps enrich it by wearing a set
of postcolonial lenses to look at narrative.’’ Prince elaborates on this conception by
moving, with impressive swiftness and clarity, through a remarkable number of
narratological categories and suggesting how they would look from this new perspective. To take just a few examples, with voice, one would focus on its ‘‘linguistic power
or communal representativeness’’; with narrator, one would look at postcolonial status
and diegetic situation. Most notably from the perspective of this volume, one would
12
James Phelan and Peter J. Rabinowitz
do with space what Susan Friedman does in her essay – pay attention to border
crossings and heterotopia. Prince concludes the essay by placing his sketch of a
postcolonial narratology within a larger vision of three important projects for contemporary and future narrative theory: (1) the continued re-examination and revision
of existing categories through the consideration of new examples of narrative – as in
this essay; (2) studying the role of various elements of narrative as part of an effort to
establish an empirical basis for narratology; and (3) reviving the effort to develop a
model of narrative competence. In this portion of his essay, Prince looks back to the
kinds of concerns motivating the essays in Parts I and II, while also looking ahead to
those raised in the Epilogue.
This look to the future continues in Melba Cuddy-Keane’s ‘‘Modernist Soundscapes
and the Intelligent Ear: An Approach to Narrative Through Auditory Perception.’’
This essay could serve as another model for the first of Prince’s projects for the future,
since it seeks to develop a vocabulary and methodology for dealing with representations of sound in narrative. At the same time, the essay resonates with the work of
Shaw and Case when it makes a historical claim that ‘‘modernity occasioned new
experiences for the ‘human sensorium,’ stimulating both a new perceptual knowledge
and a new apprehension of perception,’’ especially auditory perception. Cuddy-Keane
brings these two main concerns of the essay together in her analysis of Virginia
Woolf’s representations of sound perception in her fiction. Drawing on such terms as
auscultation, auscultize, and auscultizer (patterned on focalization, focalize, and focalizer,
narratology’s terms for describing vision) as well as soundmark and soundscape (patterned on landmark and landscape), Cuddy-Keane demonstrates the range, innovation, and importance of Woolf’s use of soundscapes from ‘‘Kew Gardens’’ to The
Years. Cuddy-Keane’s supple analyses show that ‘‘by reading for sonics rather than
semantics, . . . we discover new forms of making narrative sense.’’
Shlomith Rimmon-Kenan’s essay, ‘‘In Two Voices, or: Whose Life/Death/Story
Is It, Anyway?,’’ is a meditation on the ethics of the double illness narrative by
Ilana Hammerman and her husband Jürgen Nieraad, Under the Sign of Cancer:
A Journey of No Return. The first half of the book is Nieraad’s account of his terminal
illness, acute myeloid leukemia; the second half is Hammerman’s account of that same
illness told from her perspective after Nieraad’s death. Rimmon-Kenan structures her
treatment as ‘‘a series of concentric circles: the relations between the dying husband
and his wife, the twofold act of narration, the appropriation of both husband and wife
by the medical ‘system,’ published responses to the narrative by doctors and other
readers, and my own appropriation as evidenced in this essay.’’ The result is an analysis
that, like Phelan’s, links matters of narrative form and ethics, though RimmonKenan’s focus on this nonfictional illness narrative and published responses to it –
as well as her own reflections on her personal investment in the issues raised by the
narrative – leads her to a rather different set of ethical questions. Is Hammerman’s
narrative inevitably an appropriation of Nieraad’s experience? Is it a just indictment
of the medical system? Is Rimmon-Kenan’s own analysis yet another appropriation?
By taking on these questions with intelligence and rigor and refusing to
Introduction
13
answer definitively the question of her title, she offers a compelling example of the
intellectual whirlpools that can result from a serious ethical engagement with
narrative.
As we’ve said, narrative theory has an expansionist quality; and in Part IV, ‘‘Beyond
Literary Narrative,’’ we gather up seven essays that exemplify its ability to contribute
to fields well beyond the traditionally literary. Contribute to – or swallow up? Even
among narrative theorists, there are doubts about the impact of this expansionist (one
might even say imperialist) potential, as revealed by the questions raised in several of
these essays.
The section moves, generally, from verbal to nonverbal fields. We begin with two
essays discussing the power of unacknowledged narrative mechanisms in contemporary culture. First, in ‘‘Narrative in and of the Law,’’ Peter Brooks reflects on the ‘‘role
of storytelling’’ in the legal realm. Stories are absolutely central to legal practice, but,
according to Brooks, the law rarely admits their centrality openly. Instead, it recognizes the importance of stories only implicitly and only negatively, through ‘‘its
efforts at policing narrative,’’ a policing that manages ‘‘the conditions of telling’’ so
that ‘‘narratives reach those charged with judging them in controlled, rule-governed
forms.’’ Narrative content, in other words, is repressed – and kept ‘‘under erasure.’’
Brooks ends his essay by calling for a ‘‘legal narratology,’’ especially one that deals
with the reception and construction of stories by the listeners: judges and juries.
Alan Nadel, in ‘‘Second Nature, Cinematic Narrative, the Historical Subject, and
Russian Ark,’’ studies the impact of unacknowledged conventions in narrative cinema.
Although most viewers have come to accept their position while watching narrative
films as ‘‘natural,’’ in fact such films ‘‘naturalize a counterintuitive experience by
creating the illusion that the viewer has acquired a privileged window on reality.’’
This naturalization requires viewers to engage in a kind of learning as forgetting
which has significant social and psychological consequences. Nadel focuses on the
conventions of the ‘‘Classical Hollywood Style’’ – exemplified here in particular by the
use of the close-up – and concludes with an analysis of Alexander Sukurov’s 2002 film
Russian Ark that shows how it ‘‘calls into question that style and the history of
cinematic narrative in which it participates.’’
The next three essays deal with music, an area that would seem, at first, even
further removed from narrative concerns than law and film. In ‘‘Narrativizing the
End: Death and Opera,’’ Linda and Michael Hutcheon argue that, in order to account
for opera’s synthesis of music and story, standard narratological models need to be
significantly altered. They demonstrate the power of such an alteration by turning to
a question about operas that center on suffering and death: how do such works of art
produce pleasure? Expanding the work of traditional narrative theorists (especially
that of Frank Kermode), they propose that the shared public experience of opera,
much like the Early Modern practice of contemplatio mortis, allows its audience to
‘‘rehearse death’’ and to confront its ‘‘mortal anxieties.’’
14
James Phelan and Peter J. Rabinowitz
In ‘‘Music and/as Cine-Narrative; Or, Ceci n’est pas un leitmotif,’’ Royal S. Brown
looks at a different kind of music attached to an independent story: film music.
Traditional accounts of film music have privileged the narrative aspects of this
independent story, reducing the music to little more than a catalog of themes and
motifs that double the narrative that we see on the screen. Brown looks instead at the
‘‘quasi-narrative properties’’ found in the music itself in the form of the codes on
which Western music depends. He zeroes in on the ways in which film music interacts
with our notions of time in the title sequences composed by Hugo Friedhofer for The
Best Years of Our Lives and by Bernard Herrmann for North by Northwest. His analysis
demonstrates how film music, ‘‘through violations of strictly musical codes’’ familiar
from the concert hall, ‘‘often rises above’’ the literal-minded duplication of the screen
action, providing a commentary that, at its best, provides ‘‘a kind of meta-text whose
‘story’ is the very substance of narrative.’’
In the last of the musical essays, ‘‘Classical Instrumental Music and Narrative,’’ Fred
Everett Maus deals with music that seems, on the surface, even further removed from
narrative concerns: nonprogrammatic instrumental music. Maus charts out the major
arguments in the debates that have arisen since musicologists first started to take
narrative theory seriously in the 1970s, with particular attention to the work of Marion
A. Guck, Susan McClary, and Anthony Newcomb. He then offers his own intervention,
arguing that the relationship between music and narrative is a loose analogy and that we
should think more carefully about the ‘‘poetics’’ of texts about music (rather than
considering them simply as literal representations of the music they discuss). From
this base, he launches his most important argument: we should think not about musical
works as ideal objects that are stable and consistent but rather about the ‘‘diversity of the
dramatic successions that different performers may create, even when starting from the
same score.’’ Maus demonstrates this point with a careful analysis of several different
performances of a passage from Beethoven’s Fifth Symphony.
Most of the essays in this collection ground their arguments in the details of a
particular text. In ‘‘‘I’m Spartacus!’ ’’ Catherine Gunther Kodat does something substantially different – she studies a particular figure, Spartacus, who shows up in many
different narratives, posing the question of why we continue to be fascinated by a figure
(a term she picks with care) so inextricably tied to defeat. To answer the question, Kodat
chooses to treat the repetition of Spartacus figures as an instance of what she calls, after
J. M. Bernstein, eidetic variation. At first glance, Kodat argues, Spartacus appears to
reveal ‘‘the strength of a single figure to hold together a crumbling narrative (the
fragmentary, incomplete, and contradictory early histories of a slave rebellion in which
Spartacus is the common thread)’’; but her exploration of the various versions of
‘‘Spartacus’’ in different media (especially Howard Fast’s novel, Stanley Kubrick’s
film, and Aram Khachaturian’s ballet) reveals, rather, an underlying tension between
narrative (as ground) and Spartacus as a figure that narrative cannot, in fact, contain. In
the end, Spartacus emerges as a queer figure whose story is inevitably ‘‘our’’ story.
The most radical essay in this section is arguably the one by Peggy Phelan, ‘‘Shards
of a History of Performance Art: Pollock and Namuth Through a Glass, Darkly.’’
Introduction
15
Focusing on the history of performance art, Phelan self-consciously explores both the
need for and (like Kodat) the limitations of narrative. Taking as her primary example
the action painting of Jackson Pollock, she confronts a glaring paradox. On the one
hand, narrative is necessary for anyone who wants to produce a history of performance
art. On the other hand, as she shows in her account of the way Pollock’s work has been
treated by critics (in particular critics who discuss it through the mediation of
Namuth’s photographs of Pollock’s work), narrative itself serves to counteract the
very spirit of performance art, whose purpose is ‘‘to unsettle the distinction between
subject and object, between doing and telling.’’ The essay ends with one of her
experiments in ‘‘performative writing,’’ as she reflects on Pollock’s action painting
via prose ‘‘that remains alive to action, whose telling force resides in each breath of the
renewing present tense.’’
The Epilogue contains two essays that, like the end of Gerald Prince’s, consider recent
developments in narrative and implicitly or explicitly point the way toward the future
of both narrative and narrative theory. Marie-Laure Ryan’s ‘‘Narrative and Digitality:
Learning to Think With the Medium’’ explores the development of digital narrative
over the last 25 years by focusing on the relation between the potential of software
systems and the realization of that potential in actual narratives. Ryan offers a skillful
survey of what she calls the ‘‘affordances’’ of three major kinds of digital narrative:
interactive fiction based on Infocom software; hypertext narrative based on Storyspace
software; and mixed-media narrative based on Flash software. She then builds on the
survey to develop her larger theoretical point about ‘‘thinking with the medium,’’ her
phrase for taking advantage of the potentiality of software. Thinking with the
medium ‘‘is not the overzealous exploitation of all the features offered by the authoring system, but an art of compromise between the affordances of the system and the
demands of [narrative] meaning.’’ From this perspective, one should not judge digital
narrative by the criteria of print narrative because the technologies underlying each
kind of narrative provide very different affordances. In other words, digital narrative
should not be found wanting because it is not as good as Shakespeare or Proust but
rather should be judged by how well it takes advantage of its potential to offer ‘‘freely
explorable narrative archives; dynamic interplay between words and image; and active
participation in fantasy worlds,’’ such as we find in multiplayer online computer
games.
The final essay, H. Porter Abbott’s ‘‘The Future of All Narrative Futures,’’ is a
meditation on the power of narrative form and its consequences for the future of
narrative, based on Abbott’s analysis of current ‘‘technologically assisted’’ narrative
entertainments. Although he begins by noting that such narratives (MOOs and
MUDs as well as the kinds Ryan analyzes) have expanded the domain of narrative
and focused more attention on interactivity, Abbott ultimately sees more similarity
than difference between the fundamental structures of these narratives and print
narrative than Ryan does. But Abbott takes the investigation further by asking
whether the shifts in emphasis of our digital age signal a developing shift in our
16
James Phelan and Peter J. Rabinowitz
behavior and cognition, one moving us away from coherent master-narratives toward
more fragmentary, open-ended, local narratives. Turning to a reading of the contemporary political scene, including a consideration of the conflicting master-narratives of
the September 11, 2001 terrorists and of George W. Bush and his followers, Abbott
again suggests that the old structures remain in place. This answer, however, suggests
the next question: what is the relation between current developments in narrative
e-entertainments and the old structures? Abbott finds in these developments a
resistance to ‘‘the givenness of narrative,’’ and a correspondingly greater interest in
the ‘‘prenarratable,’’ that is, experience that is not yet shaped into narrative. More
generally, Abbott suggests that this oscillation between the prenarratable and narrative corresponds with a similar oscillation between living and narrating our lives, an
oscillation that he sees as necessary for our mental health. Consequently, ‘‘the often
wonderful developments in the technology of entertainment will continue largely to
take place within constraints, narrative or otherwise, that give us the illusion that
time itself has a shape and that somehow we are equipped to read it.’’
As we look back on this introduction’s navigation – and that of our contributors –
between the stony clarity of apparently well-grounded knowledge and the whirling
waters of theoretical and interpretive innovation, we see several larger conclusions.
(1) Contemporary narrative theory is a flourishing enterprise precisely because it
remains strongly aware of its history and tradition even as it pursues its commitment
to innovation. If one navigational mistake is to steer too close to either Scylla or
Charybdis, an equally grave mistake is to sail on oblivious to both. (2) There is no one
best way to navigate between tradition and innovation, and the field is flourishing
because its scholars have developed multiple paths even as they continue to invent
new ones. (3) The ongoing work of our contributors will play a substantial role in the
further flourishing of the field.
Prologue
1
Histories of Narrative Theory (I):
A Genealogy of Early
Developments
David Herman
[A fact becomes] historical posthumously, as it were, through events that may be
separated from it by thousands of years.
(Benjamin 1969: 265)
Introduction
When Tzvetan Todorov coined the French term narratologie (‘‘narratology’’) in his
1969 book Grammaire du ‘‘Décaméron,’’ he used the word in parallel with biology,
sociology, and so forth to denote ‘‘the science of narrative,’’ describing his work as a
fledgling effort within a field not yet fully born. Participating in a broader structuralist revolution that sought to use Saussurean linguistics as a ‘‘pilot-science’’ for
studying cultural phenomena of all sorts (Dosse 1997: 59–66), Todorov built on
precedents set by theorists such as the early Roland Barthes ([1957] 1972), who had
characterized diverse forms of cultural expression (advertisements, photographs, museum exhibits, wrestling matches, etc.) as rule-governed signifying practices or
‘‘languages’’ in their own right (Culler 1975). Founding narratology as a subdomain
of structuralist inquiry, researchers like Barthes, Claude Bremond, Gérard Genette,
A. J. Greimas, and Todorov followed Saussure’s distinction between la langue (¼
language viewed as system) and la parole (¼ individual utterances produced and
interpreted on that basis); they construed particular stories as individual narrative
messages supported by a shared semiotic system. And just as Saussurean linguistics
privileged la langue over la parole, focusing on the structural constituents and
combinatory principles of the semiotic framework of language, the narratologists
privileged narrative in general over individual narratives. Their chief concern was with
transtextual semiotic principles according to which basic structural units (characters,
20
David Herman
states, events, etc.) are combined, permuted, and transformed to yield specific
narrative texts.
I revisit the methods, aims, and legacies of structuralist narratology in the section
on ‘‘The Structuralist Synecdoche’’ below (see also Prince 1995). But it is worth
emphasizing at the outset that, despite the narratologists’ own claims for novelty, and
notwithstanding the profound impact their work has had on the study of narrative
since the 1960s, structuralist theories were not in fact as revolutionary as the
heady language in some of their more programmatic statements might suggest
(see, e.g., Barthes [1966] 1977). Rather, French structuralist narratology emerged
from a complex interplay of intellectual traditions, criticotheoretical movements,
and analytic paradigms distributed across decades, continents, nations, schools of
thought, and individual researchers. Marked by regions of nonsynchronous evolution
(Bloch 1988) and criss-crossing vectors of change, this complex of related research
initiatives makes it difficult to fix the exact geotemporal coordinates of any particular
innovation in the field. Accordingly, as I will argue in this chapter, a quite complicated story needs to be told about the rise – or diffusion – of methods for studying
stories.
Adopting a case-study approach, the next section examines René Wellek and
Austin Warren’s chapter on narrative fiction in their 1949 book Theory of Literature.
I focus on this influential text, published in mid twentieth-century North America
but bearing the traces of earlier German, Czech, and Russian scholarship, to underscore how modern-day research on narrative was shaped by the discontinuities and
dislocations associated with World War II, among other disruptive events. Caught up
in these events, Wellek and Warren’s chapter in Theory of Literature provides a point of
entry into the network of historical tendencies that conspired to make narrative theory
what it was and is. Subsequent sections of my chapter then zoom out to reveal the
broader contexts in which approaches to narrative theory emerged, interacted, and reemerged in different form.1 As my account suggests, early Anglo-American initiatives
in the field, though they have contributed importantly to recent research on narrative
(see section ‘‘Morphology II’’ below), must also be situated within a complex of
discourses and traditions that were rooted in Europe, and that later manifested
themselves in the structuralist narratology developed in France beginning in the
1960s. The final section of the chapter returns to Wellek and Warren’s account,
reassessing their discussion in light of the larger historical developments of which
I here attempt to give an overview.
A preliminary note about my title: I use the term genealogy in the sense pioneered
by Friedrich Nietzsche ([1887] 1968) and reinvigorated by Michel Foucault ([1971]
1984); in this usage, genealogy is a mode of investigation that seeks to uncover
forgotten interconnections; reestablish obscured or unacknowledged lines of descent;
expose relationships between institutions, belief-systems, discourses, or modes of
analysis that might otherwise be taken to be wholly distinct and unrelated. Genealogy
in this way seeks to denaturalize ‘‘the contingent [social, institutional, discursive, or
other] structures we mistakenly consider given, solid, and extending without change
A Genealogy of Early Developments
21
into the future as well as into the past’’ (Nehamas 1986: 110). Hence, in referring to
the chapter as a genealogy, I mean to clarify its aims: to situate recent theories of
narrative in a complex lineage, a network of historical and conceptual affiliations, and
thereby underscore how those theories constitute less a singular continuous tradition
of research than a cluster of developments marked by family resemblances. Such
resemblances are readily recognized yet notoriously difficult to define (Wittgenstein
[1953] 1958; see the final section below).
Theory of Literature and Narrative Theory: A Case Study
Published during the heyday of the New Criticism, Wellek and Warren’s Theory of
Literature helped shape Anglo-American formalist literary theory in general, as well as
research on narrative fiction in particular. The chapter on fiction reflects the authors’
overarching aim of developing an ‘‘intrinsic’’ as opposed to an ‘‘extrinsic’’ approach to
literary study; they eschewed ‘‘positivist’’ methods that subordinated literature to the
objects studied in other disciplines (e.g., history, social groups, or philosophical
arguments), advocating instead a theory of literature based on the nature of literary
objects in particular. In turn, in specifying the nature of narrative fiction as a
particular object of study, the chapter sets out many of the research foci that have
persisted as central concerns for later theorists of narrative:
. the notion that narrative fiction, including novels and short stories, constitutes
only a particular subtype of narratively organized discourse, not the canonical
form of narrative as such (Wellek and Warren 1949: 225);
. the jointly causal and temporal logic of stories, or narrative’s fusion of sequence
and consequence (p. 222);
. the distinction between the basic situations and events that form the story-stuff or
fabula of a narrative and the composition of those basic elements into the plot or
sjuzet (pp. 224–6); that is, the distinction between ‘‘story’’ and ‘‘discourse’’ (Chatman 1978) that was exploited so productively by Francophone structuralists such
as Todorov (1966) and Genette ([1972] 1980);
. mechanisms of characterization, including the social and anthropological bases for
character-types in fiction (pp. 226–8);
. point of view (pp. 230–3);
. framed tales, and the modes of narrative embedding that result from storieswithin-stories (p. 230) (cf. Genette [1972] 1980);
. the noncoincidence of author and narrator in fictional discourse (p. 230) (cf. Booth
1961);
. strategies for the representation of characters’ consciousness as well as speech
(pp. 233–4); and
. the notion that the ‘‘truth’’ of narrative fiction arises from the way its components
hang together to form a Kosmos sufficient unto itself, whereas the truth of a
22
David Herman
historical account depends on the extent to which it matches, in some sense, the
way the world is (pp. 219–22).
But is it the case that by targeting these issues the authors in fact anticipated or
prefigured subsequent approaches to narrative theory – for example, approaches
enabled by the birth of narratology in France? The story, as I have suggested, is not
that simple.
Thanks in part to Wellek’s own background as a specialist in English language and
comparative literature who was once affiliated with the Prague Linguistic Circle, and
who thus helped transmit ideas deriving from earlier European and Slavic scholarship
(including Saussurean linguistics), Theory of Literature constitutes a node within a
network of discourses and intersecting historical trends.2 Besides interacting synergistically with the guiding principles of the Anglo-American New Critics, whose
insistence on the autonomy of verbal art mirrored the Russian Formalists’ emphasis
on ‘‘literariness,’’3 the same Formalist ideas that helped shape Wellek and Warren’s
approach also affected the development of structuralist narratology. Influenced in
their own right by German scholars working on questions of narrative in the early
years of the twentieth century (see my next section), Viktor Shklovskii, Boris
Tomashevskii, Vladimir Propp, and other Formalists helped establish not only the
universe of discourse on which Wellek and Warren drew in their 1949 text, but also
the basis for some of the most recognizable topoi of narratology: for example, the
decomposition of the story level or content level into obligatory and optional
components, and the rethinking of characters as actants (see the next section and
the section on ‘‘The Structuralist Synecdoche’’ below for further discussion).
By the same token, rather than unfolding along a single timeline, the history of
narrative theory has acquired its structure from the distribution of research concerns
across parallel developmental trajectories; genealogy allows these trajectories to be
grouped together into a larger historical constellation. The student of narrative theory
would thus do well to look for family resemblances, but not necessarily causally
determinative relations, between Wellek and Warren’s characterization of narrative as
a transgeneric phenomenon and the later analogous account outlined by Barthes
([1966] 1977). The same goes for Theory of Literature’s use of the story–discourse
distinction and Genette’s use of it in Narrative Discourse, as well as for Wellek and
Warren’s probe into the anthropological bases of character types and Greimas’s
anthropologically inflected theory of actants, or general behavioral roles encoded in
narrative plots.
But what accounts, exactly, for the special synergy of the discourses and traditions
from which Wellek and Warren’s chapter on narrative fiction took rise, and to which
their study in turn contributed? In other words, when it comes to early developments
in narrative theory, what is the nature (and scope) of the historical constellation in
which Czech, Russian, French, and Anglo-American ideas and approaches can be
situated? The remainder of my chapter explores these questions.
A Genealogy of Early Developments
23
Early Twentieth-Century Narrative Poetics: ‘‘Morphological’’
Models in Germany and Russia
In his own history of occidental poetics during the period stretching from Aristotle
(1971) to the Prague School semioticians, Lubomı́r Doležel (1990) has identified a
constellation in terms of which a number of ideas about narrative can be grasped as
related developments. Doležel suggests that in the aftermath of Romanticism, ostensibly diverse approaches to (narrative) poetics were united by their common descent
from a broadly ‘‘morphological’’ research tradition. Tracing the morphological paradigm back to Goethe, and building on previous work by Victor Erlich (1965) and
Peter Steiner (1984: 68–98; cf. Steiner and Davydov 1977), Doležel characterizes the
emergence of morphological poetics in the Romantic era as a manifestation of a
broader epistemological shift.4 At issue is the replacement of a mechanistic by an
organic model for understanding the structure of the world (Doležel 1990: 55). As
Doležel puts it, ‘‘Anatomy and morphology share the assumption that an organism is
a set of parts. . . . [But whereas] anatomy is satisfied with separating and identifying
the parts, morphology informs us that the diverse parts make up a higher-order,
structured whole. Morphology is a theory of the formation of complex structures from
individual parts’’ (p. 56).
In the domain of narrative poetics, the morphological method was pioneered in
Germany in the early years of the twentieth century (Doležel 1990: 126–41; see also
Doležel 1989). It was then further developed by Russian Formalist theorists and
subsequently by Prague structuralists who built on the Formalists’ work.5 Foundational works of narrative morphology include a programmatic 1912 manifesto by the
German philologist Otmar Schissel von Fleschenberg, who analyzed higher-order
narrative structures into ‘‘rhetorical artistic devices,’’ and the Germanist Berard
Seuffert’s studies of what he identified as schematic building blocks of literary
narratives.6
As Doležel notes, German scholars such as Schissel, Seuffert, and also Wilhelm
Dibelius distinguished between the disposition (logical arrangement) and composition
(artistic arrangement) of the structural elements contained within narratives. They
focused on two aspects of compositional patterning in particular: patterns of action
and acting personae. Thus, in his 1910 study of the English novel, Dibelius anticipated Propp’s morphological study of Russian folktales by defining the notion of
‘‘role’’ as a ‘‘character in a certain function of the whole’’ (quoted by Doležel 1990:
133–4). Like Dibelius, Propp subordinated character to plot, focusing not on particularized actors but on recurrent, plot-based ‘‘functions’’ instantiated by various
individuals across specific tales (see below). In turn, anticipating the idea of actants
that Greimas developed on the basis of Propp’s work, Seuffert studied the system of
characters as an ingredient of narrative composition. Schissel, for his part, focused on
the composition of action rather than characters per se, parsing narrative structures
24
David Herman
into the patterned distribution of scenes and episodes, and identifying the ring
(Kranz) and frame (Rahmen) as basic compositional principles.
Grounded in a broadly morphological approach, this early focus on narrative
composition shaped later German narrative theory. Relevant studies include André
Jolles’s Einfache Formen, a Goethe-inspired morphology of ‘‘simple forms’’ (anecdotes,
sayings, legends, etc.) (Doležel 1990: 205 n.31), and two works on which Gérard
Genette drew in presenting the influential narratological model outlined in Narrative
Discourse ([1972] 1980): Günther Müller’s ‘‘Erzählzeit und erzählte Zeit’’ (Narrative
Time and Narrated Time), originally published in 1948 and reprinted in Müller’s
1968 book Morphologische Poetik (Morphological Poetics), and Eberhard Lämmert’s
(1955) Bauformen des Erzählens (Structural Components of Narrative). These later German
studies, extended by Genette, exploited the semantic potential of the idea of composition itself; they investigated how incidents may acquire idiosyncratic time-scales
and event structures when composed into narratives, that is, integrated into a complex
of other events to form a narrative whole. In parallel with this line of research, the
Russian Formalists had mapped the idea of ‘‘composition’’ onto that of ‘‘plot’’ (¼
sjuzet); for Shklovsky ([1929] 1990) plot encompassed not only the arrangement of
incidents ‘‘but also all the ‘devices’ used to interrupt and delay the narrative’’ (Erlich
1965: 242; cf. Wellek and Warren 1949: 226).
Meanwhile, German work on narrative composition was well-known in Russia by
the 1920s (Doležel 1990: 134–6). There, too, the morphological model flourished
(Erlich 1965: 171–2, 239, 249, Steiner 1984: 68–98), producing studies that would
mold later research on narrative theory, particularly the structuralist narratology
developed in France some four decades later. The Russian Formalists were committed
to developing stylistic models for the widest possible range of prose forms, including
the full gamut of narrative genres (Erlich 1965: 230–50, Shklovsky 1990 [1929]).
They studied, for example, narrative techniques such as the skaz, whereby a personified narrator figure in a written narrative prominently mediates between the author
and the audience to produce the illusion of spontaneous vernacular storytelling (Erlich
1965: 238, Prince 1987: 87–8). Concerned with such higher levels of narrative
structure, the Formalists studied prose narratives of all sorts, from Tolstoy’s historically panoramic novels to tightly plotted detective novels to (Russian) fairy tales. This
widened investigative focus would prove to be a decisive development in the history
of modern-day narrative theory. The new focus helped uncouple theories of narrative
from theories of the novel, shifting scholarly attention from a particular genre of
literary writing to all discourse or, in an even broader interpretation, all semiotic
activities that can be construed as narratively organized. The Formalists thus set a
precedent for the transgeneric and indeed transmedial aspirations of later French
structuralist theorists such as Bremond (1964) and Barthes ([1966] 1977) (Herman
2004). Further, the Formalists’ attempt to create a stylistics suitable for prose texts
constituted a starting-point for Bakhtin’s ([1929] 1984) (anti-Formalist) research on
the polyphonic or multivoiced profile of Dostoevsky’s fictions. A half-century later,
Bakhtin’s research helped shape the contextualist approaches to narrative that took
A Genealogy of Early Developments
25
issue with structuralist narratology, which in turn traced its roots back to the Russian
Formalists (see section on ‘‘The Structuralist Synecdoche’’ below).
Adopting the German theorists’ morphological approach to narrative composition,
and extending their framework of inquiry to include narratives of all types, the
Russian Formalists authored a number of pathbreaking studies. For example, in
distinguishing between ‘‘bound’’ (or plot-relevant) and ‘‘free’’ (or nonplot-relevant)
motifs, Boris Tomashevsky ([1925] 1965) provided the basis for Barthes’s distinction
between ‘‘nuclei’’ and ‘‘catalyzers’’ in his ‘‘Introduction to the Structural Analysis of
Narratives’’ (Barthes [1966] 1977). Renamed kernels and satellites by Seymour Chatman (1978), these terms refer to core and peripheral elements of story-content,
respectively. Delete or add to the kernel events of a story and you no longer have
the same story; delete or add to the satellites and you have the same story told in a
different way. Related to Tomashevsky’s work on free versus bound motifs, Viktor
Shklovsky’s ([1929] 1990) early work on plot as a structuring device provided what I
have already alluded to as one of the grounding assumptions of narratology: namely,
the fabula–sjuzet or story–discourse distinction, that is, the distinction between the
what and the how, or what is being told about versus the manner in which it is told.
What is arguably the most important Formalist precedent for modern narrative
theory, however, was furnished by Vladimir Propp’s Morphology of the Folktale ([1928]
1968), whose first English translation appeared in 1958. Explicitly associating his
approach with Goethean morphology (Doležel 1990: 141), Propp distinguished
between variable and invariant components of higher-order narrative structures –
more specifically, between changing dramatis personae and the unvarying plot functions performed by them (e.g., act of villainy, punishment of the villain, etc.). In all,
Propp abstracted 31 functions, or character actions defined in terms of their significance for the plot, from the corpus of Russian folktales that he used as his data-set; he
also specified rules for their distribution in a given tale. Lévi-Strauss ([1955] 1986),
in a protonarratological study that influenced French structuralists such as Bremond
(1973) and Barthes ([1966] 1977), would later build on Propp’s concept of functions
to formulate a deep-structural analysis of the ‘‘mythemes,’’ or ‘‘gross constituent
units,’’ structuring the Oedipus myth. Further, Propp’s approach constituted the
basis for structuralist theories of characters as ‘‘actants.’’ Extrapolating from what
Propp had termed ‘‘spheres of action,’’ Greimas ([1966] 1983) sought to create a
typology of general roles to which the (indefinitely many) particularized actors in
narratives could be reduced.
But before detailing the ways in which structuralist narratologists of the 1960s and
early 1970s built on the work of previous researchers, I need to explore another branch
within the genealogy of modern-day narrative theory. At issue is the Anglo-American
tradition of narrative poetics in which Wellek and Warren’s Theory of Literature
intervened in the mid twentieth century, but which extends back at least to the
tradition of novel theory that originated with Henry James (see Miller 1972) and was
codified by Percy Lubbock ([1921] 1957). This tradition began with an organicist
emphasis that can be traced back to Aristotle’s concern with the way well-made plots
26
David Herman
(mythoi) comprise elements contributing to a larger whole. Jamesian organicism
was congruent, on the one hand, with the emerging New Critical conception of
literary artifacts as ‘‘organic wholes’’ and, on the other hand, with the German, Czech,
and Russian theories of narrative morphology that surface in Wellek and Warren’s
chapter on fiction. But what is more, beginning as long as two decades before
the Francophone structuralists exposed the limits of earlier, linguistically oriented
European and Slavic models, Chicago School neo-Aristotelians such as R. S. Crane
(1953) and his student Wayne Booth (1961) leveled powerful critiques at analogous
trends in Anglo-American theories of fiction. Anglo-American and French structuralist approaches to narrative thus underwent a kind of staggered development,
following parallel evolutionary trajectories at nonsynchronous rates of change.
Morphology II: Organic Form, Anglo-American
Formalism, and Beyond
Henry James’s influential 1884 essay on ‘‘The Art of Fiction’’ (Miller 1972: 27–44)
provided important foundations for Anglo-American novel theory. Three interconnected aspects of the model outlined by James can be singled out for discussion here:
its characterization of fictions as organic wholes, its adoption of a descriptive rather
than prescriptive approach to narrative analysis, and its skeptical attitude toward
exact recipes or protocols for the production of fictional texts. The subsequent course of
early to mid twentieth-century Anglo-American narrative theory can be mapped out
as a series of (more or less dialectically related) responses to these aspects of James’s
approach.
In parallel with Aristotelian poetics as well as the later morphological models
discussed in my previous section, James characterizes narrative fictions as higher-order
verbal structures greater than the sum of their parts. For James, too, the components
of fiction derive their functional properties from their relation to the gestalt in which
they participate. But James’s metaphorical equation of literary works with living
organisms is even more explicit than that adopted by the Russian Formalists.
Questioning such standard oppositions as incident vs. description and description
vs. dialogue, James remarks: ‘‘A novel is a living thing, all one and continuous,
like any other organism, and in proportion as it lives will it be found, I think, that in
each of the parts there is something of each of the other parts’’ (Miller 1972: 36). It is
in this same context that James poses his famous two-part rhetorical question,
suggesting that characters and events cannot be understood in isolation from the
plot by which they are linked together and from which they acquire their significance
in the first place: ‘‘What is character but the determination of incident? What is
incident but the illustration of character?’’ (p. 37). Similarly, refusing to distinguish
between the form and content of fiction, James writes: ‘‘This sense of the story being
the idea, the starting-point of the novel, is the only one that I see in which it can
be spoken of as something different from its organic whole’’ (p. 40). Precisely because
A Genealogy of Early Developments
27
for James the content of fiction must be allowed to determine its form, and vice versa,
his account remains resolutely antiprescriptivist. He argues that it is impossible to
dictate, a priori, what makes for successful narrative fiction. Rather, ‘‘the good health
of an art which undertakes so immediately to reproduce life must demand that it
be perfectly free. It lives upon exercise, and the very meaning of exercise is freedom’’
(p. 33).
Percy Lubbock ([1921] 1957) took his inspiration from James’s novelistic practice
as well as his theory of fiction. Working in tandem with his New Critical contemporaries, who characterized verbal artifacts as organic wholes such that removing or
altering one part would destroy the entire work (Brooks [1947] 1992), Lubbock
espoused the doctrine of organic unity in fiction; he thus privileged James’s own wellwrought fictions over the more sprawling creations produced by writers such as
Tolstoy. However, Lubbock made the issue of ‘‘point of view’’ the cornerstone of his
account – to an extent not necessarily warranted by James’s own approach (Miller
1972: 1, Booth 1961: 24–5). In doing so, Lubbock appropriated James’s ideas to
produce a markedly prescriptive framework. His chief concern was to draw an
invidious distinction between showing (‘‘dramatizing’’ events) and telling (‘‘describing’’ or ‘‘picturing’’ events), suggesting that description is inferior to dramatization,
picturing to scene-making: ‘‘other things being equal, the more dramatic way is
better than the less. It is indirect, as a method; but it places the thing itself in view,
instead of recalling and reflecting and picturing it’’ (Lubbock 1957: 149–50). But if
in systematizing James’s ideas Lubbock transformed them from more or less tentative
descriptions into prescriptions, by the same token he construed fictional techniques as
part of a craft rather than mysterious emanations of artistic genius. More specifically,
in advocating for and not just describing the techniques whereby events are dramatized by being filtered through a reflector figure or fictional center of consciousness,
Lubbock suggested that specific methods or procedures are at the heart of the craft of
fiction.
In response, maintaining a focus on issues of narrative technique, but seeking to
restore the qualifications and complexities evident in James’s original statement of his
theory (as well as in his novelistic practice), Wayne C. Booth (1961) inverted the
terms of Lubbock’s argument. Instead of privileging showing over telling, Booth
accorded telling pride of place – making it the general narratorial condition of which
‘‘showing’’ is a localized effect, or rather an epiphenomenon. Thus, unlike Lubbock,
Booth resisted making point of view the paramount concern of the theory of fiction.
Indeed, Booth’s brilliant account revealed difficulties with the very premise of the
telling-versus-showing debate. He characterized showing as an effect promoted by
certain deliberately structured kinds of tellings, organized in such a way that narratorial mediation (though inescapably present) remains more or less covert. Booth also
suggested that an emphasis on showing over telling has costs as well as benefits,
cataloging important rhetorical effects that explicit narratorial commentary can be
used to accomplish – for example, relating particulars to norms established elsewhere
in the text, heightening the significance of events, or manipulating mood.
28
David Herman
More generally, Booth’s rigorous exposé of the incoherence of calls for artistic
purity – his demonstration that it is in fact impossible to ‘‘purify’’ narrative fiction
of rhetoric, as if separating the wheat from the chaff – aligns his account with neoAristotelian approaches to literary theory. Given Booth’s affiliation with the Chicago
School (see Crane 1953, Richter forthcoming, Vince 1993), it is not surprising that he
would take issue with the formalist orientation of critics such as Allen Tate (cf. Booth
1961: 28–9), who echoed other New Critics in insisting on the autonomy and selfsufficiency of the work of art, defined as a particular use of (or orientation toward)
language. Booth’s model can be situated instead in the context of the critical
framework developed by neo-Aristotelians such as Crane, Elder Olson, and Richard
McKeon, who used Aristotle’s taxonomy of causes as an approach to analyzing literary
texts – in effect using Aristotle’s own ideas to recontextualize the organicism that the
New Critics had derived from the Poetics. In this framework, causes of verbal art
include the efficient cause (¼ the author), the final cause (¼ effect on readers), the
material cause (¼ the language), and the formal cause (¼ the mimetic content) (Vince
1993: 117).7 Accordingly, from a neo-Aristotelian perspective, New Critical theories
of literary art in general and narrative fiction in particular were guilty of reducing all
the causal forces operating on and manifested in verbal art to the material cause alone.
Rather than consigning to the intentional fallacy a concern with efficient causes, or to
the affective fallacy a concern with final causes, the neo-Aristotelian approach suggested that all of these concerns are legitimate research foci. Analogously, factoring in
efficient and final as well as material causes, Booth studied how particular kinds of
fictional designs prompt specific kinds of rhetorical transactions between authors and
audiences.
Furthermore, Booth’s wide-ranging discussion of narrative types ranging from
Boccaccio’s Decameron to ancient Greek epics to novels and short fictions by authors
as diverse as Cervantes, Hemingway, and Céline encouraged subsequent theorists in
the Anglo-American tradition to explore various kinds of narratives, rather than
focusing solely on the novel. This uncoupling of narrative theory from novel theory
– a process that had been initiated independently by the Russian Formalists some 40
years earlier – culminated in such wide-scope works as Robert Scholes and Robert
Kellogg’s study of The Nature of Narrative (1966). Significantly, Scholes and Kellogg’s
study was published the same year as Communications 8, which effectively launched
structuralist narratology with a similarly transgeneric focus.
The Structuralist Synecdoche: Narratology after Russian
Formalism
In rejecting the New Critics’ exclusive emphasis on the material cause of verbal art,
the neo-Aristotelian Chicago School had, by the time of the so-called ‘‘structuralist
revolution,’’ already taken issue with the narrative ¼ language equation that is at the
heart of structuralist narratology. Nonetheless, when Barthes, Bremond, Genette,
A Genealogy of Early Developments
29
Greimas, and Todorov initiated the project of narratology, their efforts hearkened back
to a major ‘‘theoretical synecdoche’’ (Steiner 1984) used by the Russian Formalists.
Citing a 1928 statement by the Formalist Viktor Zirmunsky – ‘‘Insofar as the
material of poetry is the word . . . the classification of verbal phenomena provided by
linguistics should be the basis for a systematically constructed poetics’’ – Steiner
points out that ‘‘the model described by Zirmunsky is a synecdoche, a pars pro toto
relationship. It substitutes language – the material of verbal art – for art itself, and
linguistics – the science of language – for literary studies’’ (p. 138). The narratologists
repeated this trope, but drew on Saussure’s work to lend the synecdoche hermeneutic
authority as well as quasi-scientific status. Ironically, however, in reinstating the
Formalists’ synecdoche on what they viewed as rigorously structuralist-linguistic
grounds, narratologists exposed the limits of the very linguistic models that they
used to make their case, that is, to reassert the argument that narrative is (preeminently) a kind of language.8
In any event, the use of (Saussurean) linguistics as a pilot-science or ‘‘theoretical
synecdoche’’ shaped the object, methods, and overall aims of structuralist narratology
as an investigative framework (see Herman 2002: 2–4). Narratology’s grounding
assumption is that a common, more or less implicit, model of narrative explains
people’s ability to recognize and interpret many diverse productions and types of
artifacts as stories. In turn, the raison d’être of narratological analysis is to develop an
explicit characterization of the model underlying people’s intuitive knowledge about
stories, in effect providing an account of what constitutes humans’ narrative competence. Hence, having conferred on linguistics the status of a ‘‘founding model’’
(Barthes [1966] 1977: 82), Barthes identifies for the narratologist the same object
of inquiry that (mutatis mutandis) Ferdinand de Saussure ([1916] 1959) had specified
for the linguist: the system (la langue) from which the infinity of narrative messages
(la parole) derives and on the basis of which they can be understood as stories in the
first place.
Further, narratologists like the early Barthes used structuralist linguistics not just
to identify their object of analysis, but also to elaborate their method of inquiry. This
adaptation of linguistic methods was to prove both enabling and constraining.9 On
the positive side, the example of linguistics did provide narratology with a productive
vantage-point on stories, affording terms and categories that generated significant
new research questions. For example, the linguistic paradigm furnished Barthes with
what he characterized as the ‘‘decisive’’ concept of the ‘‘level of description’’ (Barthes
1977: 85–8). Imported from grammatical theory, this idea suggests that a narrative is
not merely a ‘‘simple sum of propositions’’ but rather a complex structure that can be
analyzed into hierarchical levels – in the same way that a natural-language utterance
can be analyzed at the level of its syntactic, its morphological, or its phonological
representation. Building on Todorov’s (1966) Russian Formalist-inspired proposal to
divide narrative into the levels of story and discourse, Barthes himself distinguishes
three levels of description: at the lowest or most granular level are functions (in Propp’s
and Bremond’s sense of the term); then actions (in the sense used by Greimas in his
30
David Herman
work on actants); and finally narration (which is ‘‘roughly the level of ‘discourse’ in
Todorov’’ [1977: 88]) (cf. Genette [1972] 1980).
Yet narratology was also limited by the linguistic models it treated as paradigmatic. Ironically, the narratologists embraced structuralist linguistics as their pilotscience just when its deficiencies were becoming apparent in the domain of linguistic
theory itself (cf. Herman 2001). The limitations of the Saussurean paradigm were
thrown into relief, on the one hand, by emergent formal (e.g., generative-grammatical) models for analyzing language structure. On the other hand, powerful tools were
being developed in the wake of Ludwig Wittgenstein, J. L. Austin, H. P. Grice, John
Searle, and other post-Saussurean language theorists interested in how contexts of
language use bear on the production and interpretation of socially situated utterances.
These theorists began to question what they viewed as counterproductive modes of
abstraction and idealization in both structuralist linguistics and the Chomskyan
paradigm that displaced it. Research along these lines led to the realization that
certain features of the linguistic system – implicatures, discourse anaphora, protocols
for turn-taking in conversation, and so forth – emerge only at the level beyond the
sentence.
Accordingly, Barthes unintentionally reveals the limits of structuralist narratology
when he remarks that ‘‘a narrative is a long sentence, just as every constative sentence
is in a way the rough outline of a short narrative,’’ suggesting that one finds in
narrative, ‘‘expanded and transformed proportionately, the principal verbal categories:
tenses, aspects, moods, persons’’ (Barthes 1977: 84). Other early narratologists shared
with Barthes the assumption that all the categories pertaining to sentence-level
grammar could be unproblematically scaled up to the discourse level, without
compromising the descriptive or explanatory power of the grammatical machinery
involved. Todorov’s study of Boccaccio’s Decameron borrowed categories from traditional grammars to compare narrated entities and agents with nouns, actions and
events with verbs, and properties with adjectives (Todorov 1969). Genette ([1972]
1980) drew on the same grammatical paradigm in using tense, mood, and voice to
characterize the relations between the narrated world, the narrative in terms of which
it is presented, and the narrating that enables the presentation.
Indeed, the narratologists’ appropriation of ideas from structuralist linguistics not
only determined their object and method of inquiry, but also shaped the overall
purpose of the narratological enterprise itself. In Jonathan Culler’s phrase, ‘‘linguistics
is not hermeneutic’’ (Culler 1975: 31); that is, linguistic analysis seeks to provide not
interpretations of particular utterances, but rather a general account of the conditions
of possibility for the production and processing of grammatically acceptable forms
and sequences. Similarly, narratologists argued that the structural analysis of narrative
should not be viewed as a handmaiden to interpretation. The aims of narratology
were, rather, fundamentally taxonomic and descriptive. Structural analysis of stories
concerned itself not with what narratively organized sign systems mean but rather
with how they mean, and more specifically with how they mean as narratives. The
narratologists therefore adhered to Jakobson’s (1960) distinction between poetics and
A Genealogy of Early Developments
31
criticism; in privileging the code of narrative over particular stories supported by that
code, they pursued narrative poetics, not narrative criticism. The critic engages in
serial readings of individual stories; the narratologist studies what allows an assemblage of verbal or other signs to be construed as a narrative in the first place.
Thus the structuralist narratologists made into an explicit methodology the
theoretical synecdoche that was a more or less implicit background assumption for
the Russian Formalists. In essence, the narratologists looked to language theory for
model-building purposes. In the process, they confirmed the fruitfulness of transferring into the domain of narrative theory ideas incubated in other disciplines (Herman
2002). But there is another lesson to be learned from the history of narratology as
well: namely, that when it comes to research on narrative, cross-disciplinary transfers
need to be managed with care, on pain of overextending ideas and methods whose
limits of applicability have already made themselves manifest in other contexts.
Wellek and Warren Revisited: A Genealogical Perspective
Overall, the account offered in this chapter suggests that the history of early developments in narrative theory should be viewed, not as a series of events linked by a
single causal chain and lending themselves to neat periodizations (Russian Formalism,
New Criticism, narratology, etc.), but rather as fields of forces sometimes unfolding in
parallel, sometimes temporally staggered, but emerging at different rates of progression and impinging on one another through more or less diffuse causal networks. In
sketching the beginnings of modern-day research on narrative, I have thus subscribed
to the genealogical method, as characterized by R. Kevin Hill (1998):
History, according to genealogists, is not teleological (as it is for Hegel). They cannot
identify a goal of a historical process, and then go on to show how it gradually emerged
from its embryonic beginnings. Rather, they chart the processes that, by contingent
confluence, produce a contemporary result. Hence the metaphor: no individual is the
goal of a family history. Rather, a family is a vast fabric of relationships, and any one
individual represents only one among many confluences of past lines of descent.
(Hill 1998: 1)
Likewise, no individual approach or school is the ‘‘goal’’ of the history (better,
histories) of narrative theory. The field constitutes, instead, a cluster or family of
related developments with intersecting lines of descent; in this context, earlier
developments have a shaping but not determinative influence on later ones, and
whereas some modes of analysis branch out from and feed back into a shared historical
tradition, others represent theoretical innovations that have not had a larger continuing impact on this research domain.
Viewed from a genealogical perspective, Wellek and Warren’s chapter on narrative
fiction did not ‘‘anticipate’’ structuralist narratology. Rather, it was itself shaped by
32
David Herman
German, Czech, and Russian lines of development that also led, through multiple
routes of historical, biographical, and conceptual influence, to the formation of the
narratologists’ models and methods. Drawing on a common stock of ideas, both
the authors of Theory of Literature and the narratologists parsed narrative structure
into story and discourse, shifted their interest from specific characters to character
types (or actants), highlighted the interconnectedness of causality and chronology in
narrative representations, and pointed (more or less explicitly) to the benefits of a
transgeneric focus on narratives of all kinds.10
But then there are elements of Wellek and Warren’s account that were simply
unassimilable by structuralist narratology as originally constituted. Because of their
own Saussurean heritage, the structuralists were ill-equipped to handle the problem of
fictional reference, that is, the question of what (fictional and other) narratives refer
to. Thus, in broaching the idea of the fictional Kosmos – the storyworld underlying and
evoked by fictional representations – Wellek and Warren’s account highlighted
problems of narrative semantics that would remain undeveloped for some three or
four decades, until another complex of discourses began to reshape narrative theory
and alter its genealogical profile.11
NOTES
1 Even so, given the limited scope of my account, I will not be able to deal with all of
the contexts in which modern-day research on
narrative has developed – in some cases yielding approaches very different in focus and
method from those discussed in this chapter.
For example, since theories of narrative were
entangled until quite recently with theories of
the novel, Georg Lukács’s ([1920] 1971,
[1938] 1990) Hegel- and Marx-inflected studies of the novel constitute relevant sources for a
broader survey of early developments in the
field. Similarly, during the mid twentieth century, Frankfurt School Marxists such as Walter
Benjamin (1969: 28–109) and Ernst Bloch
(Bloch 1988: 245–77) explored ideas that are
important for narrative theory. Of course, recent non-Western approaches to narrative theory would also need to be included in a more
comprehensive survey.
2 Born in Austria in 1903, Wellek was an inaugural member of the Prague Linguistic Circle, which was established in 1926 (Doležel
1995: 34, Erlich 1965: 156). Wellek was familiar with key ideas advanced by the Russian
Formalists, whose work preceded the founding
of the Czech group and (through Roman
Jakobson) helped set its agenda (Steiner
1984: 27–33). After Hitler invaded Czechoslovakia, Wellek emigrated to the United
States and served as the Chair of the Comparative Literature Program at Yale from 1946–72
(see Lawall 1993).
3 See, however, Thompson’s (1971) account of
underlying differences between the Russian
and Anglo-American varieties of formalism.
4 In a controversial essay, Darby (2001) also
traces back to Goethe’s (and Schiller’s) work
tendencies in modern-day research on narrative. But whereas my chapter builds on the
constellation sketched by scholars like Doležel
(1990), Erlich (1965), and Steiner (1984), and
uses the cross-disciplinary idea of morphology
to explore affiliations among German, Slavic,
Anglo-American, and Francophone theories of
narrative, Darby instead proposes a bifurcated
genealogy. For him, there is a radical split
between the German Erzähltheorie that gained
momentum in the 1950s, on the one hand,
and a ‘‘French-American’’ tradition originating
with structuralist narratology, on the other
hand.
A Genealogy of Early Developments
5 In this chapter, I focus mainly on Russian Formalist extensions of the morphological model
and their bearing, in turn, on later French
structuralist narratological theories. However,
in their own analyses of narrative semiotics
Prague structuralists such as Felix Vodička,
who wrote an important 1948 study of Czech
fictional prose, also worked in what can be
construed as a broadly morphological tradition
(see Doležel 1989, 1995: 44–5).
6 A fuller genealogy of German narrative theory
would also need to include discussion of an
important work mentioned by Darby (2001)
but not treated here, namely, Spielhagen’s
(1883) book on the theory and technique of
the novel.
7 David Richter (in an email message dated 4
November 2003) notes that members of the
Chicago School found in Aristotle’s work
grounds for dividing both the efficient and
the final cause into ‘‘external’’ and ‘‘internal’’
subtypes. Whereas the author or maker is the
external efficient cause, the manner in which
something is done can be construed as an
internal efficient cause. Likewise, whereas the
effect produced on an audience constitutes
the external final cause of a work, the work’s
internal final cause might be thought of as
REFERENCES
AND
Aristotle (1971). Poetics. In H. Adams (ed.), Critical Theory Since Plato (pp. 48–66). San Diego:
Harcourt Brace Jovanovich.
Bakhtin, M. M. ([1929] 1984). Problems of
Dostoevsky’s Poetics, ed. and trans. C. Emerson.
Minneapolis: University of Minnesota Press.
Barthes, R. ([1957] 1972). Mythologies, trans.
A. Lavers. New York: Hill and Wang.
Barthes, R. ([1966] 1977). ‘‘Introduction to the
Structural Analysis of Narratives.’’ In S. Heath
(trans.), Image Music Text (pp. 79–124). New
York: Hill and Wang.
Benjamin, W. (1969). Illuminations, ed. H. Arendt,
trans. H. Zohn. New York: Schocken.
Bloch, E. (1988). The Utopian Function of Art and
Literature: Selected Essays, ed. J. Zipes, trans.
J. Zipes and F. Mecklenburg. Cambridge, MA:
MIT Press.
8
9
10
11
33
‘‘a purpose inherent in the thing made, a kind
of entelechy of the artifact.’’
Todorov’s (1965) translation of key Formalist
texts served as a connecting link between
Russian and French versions of the theoretical synecdoche in question. Meanwhile, as
Doležel (1995: 39–40) points out, though
they were developing their ideas several decades before the French, Prague School structuralists refrained from making this sort
of pars pro toto argument (cf. Doležel 1990:
147–75).
See Pavel (1989) for a wide-ranging critique
of structuralist appropriations of ideas from
linguistics.
On the connection between causality and
temporality in stories, see Barthes’s ([1966]
1977: 94) famous characterization of narrative as a systematic application of the logical
fallacy post hoc propter hoc (‘‘after this, therefore
because of this’’).
I am grateful to Brian McHale, Monika Fludernik, Manfred Jahn, James Phelan, Peter
Rabinowitz, and David Richter for their insightful feedback on earlier drafts of this
chapter, and to Lubomı́r Doležel, Uri Margolin, and Peter Steiner for invaluable bibliographic assistance.
FURTHER READING
Booth, W. C. (1961). The Rhetoric of Fiction. Chicago: University of Chicago Press.
Bremond, C. (1964). ‘‘Le message narratif.’’ Communications 4, 4–32.
Bremond, C. (1973). Logique du récit. Paris: Seuil.
Brooks, C. ([1947] 1992). ‘‘The Heresy of Paraphrase.’’ In H. Adams (ed.), Critical Theory Since
Plato, revised edn. (pp. 961–68). Fort Worth:
Harcourt Brace Jovanovich.
Chatman, S. (1978). Story and Discourse: Narrative
Structure in Fiction and Film. Ithaca, NY: Cornell
University Press.
Crane, R. S. (1953). The Languages of Criticism and
the Structure of Poetry. Toronto: University of
Toronto Press.
Culler, J. (1975). Structuralist Poetics: Structuralism,
Linguistics, and the Study of Literature. Ithaca,
NY: Cornell University Press.
34
David Herman
Darby, D. (2001). ‘‘Form and Context: An Essay in
the History of Narratology.’’ Poetics Today 22(4),
829–52.
Doležel, L. (1989). ‘‘Two Narratologies: Propp and
Vodička.’’ In K. Eimermacher, P. Grzybek, and
G. Witte (eds.), Issues in Slavic Literary and
Cultural Theory (pp. 13–27). Bochum: Universitätsverlag Dr. Norbert Brockmeyer.
Doležel, L. (1990). Occidental Poetics: Tradition and
Progress. Lincoln: University of Nebraska Press.
Doležel, L. (1995). ‘‘Structuralism of the Prague
School.’’ In R. Selden (ed.), The Cambridge History of Literary Criticism, vol. 8 (pp. 33–57).
Cambridge, UK: Cambridge University Press.
Dosse, F. (1997). History of Structuralism, vol. 1,
trans. D. Glassman. Minneapolis: The University of Minnesota Press.
Erlich, V. (1965). Russian Formalism: History –
Doctrine, 2nd edn. The Hague: Mouton.
Foucault, M. ([1971] 1984). ‘‘Nietzsche, Genealogy,
History,’’ trans. D. F. Bouchard and S. Simon.
In P. Rabinow (ed.), The Foucault Reader
(pp. 76–100). New York: Pantheon Books.
Genette, G. ([1972] 1980). Narrative Discourse: An
Essay in Method, trans. J. E. Lewin. Ithaca, NY:
Cornell University Press.
Greimas, A. J. (1983). Structural Semantics: An
Attempt at a Method, trans. D. McDowell,
R. Schleifer, and A. Velie. Lincoln: University
of Nebraska Press.
Herman, D. (2001). ‘‘Sciences of the Text.’’ Postmodern Culture, 11, 3. <www.iath.virginia.edu/
pmc/text-only/issue.501/11.3herman.txt>.
Herman, D. (2002). Story Logic: Problems and Possibilities of Narrative. Lincoln: University of
Nebraska Press.
Herman, D. (2004). ‘‘Toward a Transmedial
Narratology.’’ In M.-L. Ryan (ed.), Narrative
Across Media: The Languages of Storytelling
(pp. 47–75). Lincoln: University of Nebraska
Press.
Hill, R. K. (1998). ‘‘Genealogy.’’ In E. Craig (ed.),
The Routledge Encyclopedia of Philosophy, vol. 4
(pp. 1–5). London: Routledge.
Jakobson, R. (1960). ‘‘Closing Statement: Linguistics and Poetics.’’ In T. A. Sebeok (ed.), Style in
Language (pp. 350–77). Cambridge, MA: MIT
Press.
Lawall, S. (1993). ‘‘Wellek, René.’’ In I. Makaryk
(ed.), Encyclopedia of Contemporary Literary Theory
(pp. 484–5). Toronto: University of Toronto
Press.
Lévi-Strauss, C. ([1955] 1986). ‘‘The Structural
Study of Myth,’’ trans. C. Jacobson and B. G.
Schoepf. In H. Adams and L. Searle (eds.),
Critical Theory Since 1965 (pp. 809–22). Tallahassee: University Presses of Florida.
Lubbock, P. ([1921] 1957). The Craft of Fiction.
London: Jonathan Cape.
Lukács, G. ([1920] 1971). The Theory of the Novel,
trans. A. Bostock. London: Merlin Press.
Lukács, G. ([1938] 1990). The Historical Novel,
trans. H. and S. Mitchell. Lincoln: University
of Nebraska Press.
Miller, J. E., Jr. (ed). (1972). Theory of Fiction:
Henry James. Lincoln: University of Nebraska
Press.
Nehamas, A. (1986). Nietzsche: Life as Literature.
Cambridge, MA: Harvard University Press.
Nietzsche, F. ([1887] 1968). On the Genealogy of
Morals. In W. Kaufman (ed.), Basic Writings of
Nietzsche, trans. W. Kaufman (pp. 437–599).
New York: The Modern Library.
Pavel, T. G. (1989). The Feud of Language: A History
of Structuralist Thought, trans. L. Jordan and
T. G. Pavel. Oxford: Blackwell.
Prince, G. (1987). A Dictionary of Narratology.
Lincoln: University of Nebraska Press.
Prince, G. (1995). ‘‘Narratology.’’ In R. Selden
(ed.), The Cambridge History of Literary Criticism,
vol. 8 (pp. 110–30). Cambridge, UK: Cambridge University Press.
Propp, V. ([1928] 1968). Morphology of the Folktale,
trans. L. Scott, revised by L. A. Wagner. Austin:
University of Texas Press.
Richter, D. H. (forthcoming). ‘‘Chicago School,
The.’’ In D. Herman, M. Jahn, and M.-L. Ryan
(eds.), The Routledge Encyclopedia of Narrative
Theory. London: Routledge.
Saussure, F. de ([1916] 1959). Course in General
Linguistics, ed. C. Bally and A. Sechehaye,
in collaboration with A. Riedlinger, trans.
W. Baskin. New York: The Philosophical
Library.
Scholes, R. and R. Kellogg (1966). The
Nature of Narrative. Oxford: Oxford University
Press.
Shklovsky, V. ([1929] 1990). Theory of Prose, trans.
B. Sher. Elmwood Park, IL: Dalkey Archive
Press.
A Genealogy of Early Developments
Spielhagen, F. (1883). Beiträge zur Theorie and
Technik des Romans. Leipzig: Staakmann.
Steiner, P. (1984). Russian Formalism: A Metapoetics. Ithaca, NY: Cornell University Press.
Steiner, P. and Davydov, S. (1977). ‘‘The Biological Metaphor in Russian Formalism: The
Concept of Morphology.’’ SubStance 16, 149–58.
Thompson, E. M. (1971). Russian Formalism and
Anglo-American New Criticism: A Comparative
Study. The Hague: Mouton.
Todorov, T. (1965). Théorie de la littérature: Textes
des formalistes russes. Paris: Seuil.
Todorov, T. (1966) ‘‘Les catégories du récit littéraire.’’ Communications 8, 125–51.
Todorov, T. (1969). Grammaire du ‘‘Décaméron.’’
The Hague: Mouton.
35
Tomashevsky, B. ([1925] 1965). ‘‘Thematics.’’ In
L. T. Lemon and M. J. Reis (eds.), Russian
Formalist Criticism (pp. 61–95). Lincoln: University of Nebraska Press.
Vince, R. W. (1993). ‘‘Neo-Aristotelian School.’’
In I. Makaryk (ed.), Encyclopedia of Contemporary
Literary Theory (pp. 116–19). Toronto: University of Toronto Press.
Wellek, René and Warren, Austin (1949). ‘‘The
Nature and Modes of Narrative Fiction.’’ In
Theory of Literature (pp. 219–34). New York:
Harcourt Brace.
Wittgenstein, L. ([1953] 1958). Philosophical Investigations, eds. G. E. M. Anscombe and
R. Rhees, trans. G. E. M. Anscombe, 3rd edn.
Oxford: Blackwell.
2
Histories of Narrative Theory (II):
From Structuralism to the
Present
Monika Fludernik
The history of narratology has recently been cast in two different plots. The first plot,
entitled ‘‘The rise and fall of narratology,’’ consists in a story that starts with early
beginnings in Todorov, Barthes, and Greimas; finds its climax in Gérard Genette
(with a few adjacent peaks taken up by F. K. Stanzel, Mieke Bal, Seymour Chatman,
Gerald Prince and Susan Lanser); and thereafter plunges to a decline, thus suitably
bracketing the discipline between the ‘‘death of the author’’ and the ‘‘death of
narratology.’’ This account is now widely considered to be outmoded (Herman
1999: 1), even by Carlo Romano (2002) in his recent piece in the Chronicle of Higher
Education. In her introduction to Neverending Stories, Ingeborg Hoesterey (1992)
provides a bridge to an alternative plot of narratological development by distinguishing between an ‘‘ ‘archaic’ phase of narratological scholarship’’ (roughly equivalent to
the period discussed by David Herman in the previous chapter of this volume), a
second phase of ‘‘classical’’ narratology based on the structuralist paradigm (from
which Herman took his term ‘‘postclassical narratologies’’), and a third phase of
so-called ‘‘critical’’ narratology which Hoesterey describes as a ‘‘new Hellenism’’
(Hoesterey 1992: 3). The model is designed, positively, to reassure us that narratology
keeps flourishing, but bears some traces of the first plot. Although meant to be
descriptive on the pattern of styles of Greek architecture and sculpture, the term
‘‘Hellenism’’ – despite the associated label ‘‘critical’’ – hints not only at the current
proliferation of scholarship but also at a dilution of aesthetic standards usually
associated with the Hellenistic period, implying a discipline past its prime that is
experiencing a final flourish before sinking into well-deserved oblivion.
The second, alternative plot, now dominant, sees classical narratology as a stage in
the much more encompassing development of narrative theory. It bears the title ‘‘The
From Structuralism to the Present
37
rise and rise of narrative’’ (cf. Cobley’s ‘‘The Rise and Rise of the Novel’’).1 In this
competing plot, the adolescence of narratology was followed by a reorientation and
diversification of narrative theories, producing a series of subdisciplines that arose in
reaction to post-structuralism and the paradigm shift to cultural studies. (See, for
instance, the telltale plural in the title of David Herman’s 1999 collection Narratologies and the surveys provided by Richardson 2000, Fludernik 2000a, and Nünning
2000.) Among these subdisciplines the most frequently mentioned are the psychoanalytic narrative approach (e.g., work by Ross Chambers 1984 and Peter Brooks
1985); feminist narratology (see especially Lanser 1992, 1999, Warhol 1989, 2003);
cultural studies-oriented narrative theory (e.g., Steven Cohan and Linda Shires 1988
or Nancy Armstrong and Leonard Tennenhouse 1993); as well as more recent work
concentrating on postcolonial (readings of) narrative. This extension, if not proliferation, of paradigms also has terminological consequences. Some critics still identify
themselves as narratologists (especially Herman, Ryan, Lanser, Fludernik, Nünning),
whereas others restrict the use of the term narratology to what I here call ‘‘classical
narratology’’ (Barthes, Todorov, Genette, Prince, Chatman, early Lanser, Cohn, possibly Stanzel). In the spirit of this volume, I will therefore be using narrative theory and
narratology interchangeably (but compare Nünning 2003 for a complex terminological proposal). Plot number two, perhaps to be entitled alternatively ‘‘De pluribus
progressio,’’ moreover ends as a success story. Out of the diversity of approaches and
their exogamous unions with critical theory have now emerged several budding
narratologies which betoken that the discipline is in the process of a major revival.
Among recent achievements one can count developments from possible worlds theory
to information-based and media technology-based approaches (linked especially to the
work of Marie-Laure Ryan 1991, 1999, 2001); the combination of philosophy,
linguistics, and conversation analysis2 in David Herman’s oeuvre (1995, 2002); the
cognitivist and cultural studies-oriented work of Ansgar Nünning (especially 1997,
2000); the marriage of the empirical sciences with narratology in Duchan, Bruder,
and Hewitt (1995) and Bortolussi and Dixon (2003); the organicist and historical
approach represented by Fludernik (especially 1996, 2003a); the extension of rhetorics
and ethics into discrete narratological paradigms in the wake of Wayne C. Booth’s
Rhetoric of Fiction and The Company We Keep (Booth [1961] 1983, 1988, Phelan 1989a,
1996, Phelan and Rabinowitz 1994, Rabinowitz and Smith 1998); as well as the
expansion of narrative theory to cover other media and genres in Wolf (2002, 2003).
In the following I will outline what I see as major aspects of narrative theory in its
various phases since Roland Barthes. It will obviously be impossible to name all
important narrative theorists or to do justice to the diversity and originality of
narratological work written in the past 40 to 50 years. My presentation will moreover
be biased in favor of individual schools, widely influential contributions, and theoretical monograph publications – an emphasis that tends to disadvantage a number of
important scholars.3
38
Monika Fludernik
Structuralist Narratology: The Rage for Binary Opposition,
Categorization, and Typology
Structuralism, based on Ferdinand de Saussure’s linguistic insights into the structure
of language as a system (langue) irrespective of speakers’ faulty performances in their
parole, set the parameters for early literary structuralism of which narratology was one
important offshoot. In imitation of de Saussure and phonology, the most fundamental
building stone of the narratological edifice became the structure of binary opposition.
Thus, in Bremond’s plot analysis (1973), decisions taken between two alternative
courses have to be taken at each juncture of the hero’s quest. Even more prominently,
Gérard Genette’s typology of narrative forms (1980) sports a panoply of binary
oppositions: homodiegetic vs. heterodiegetic narrative (the narrator is or is not a
character in the tale); extradiegetic vs. intradiegetic (the narrator’s act of narration is
situated inside or outside the story); focalization interne vs. externe, and so forth. The
last example is instructive in the way Genette takes three categories and turns them
into a hierarchy of binaries. Although Genette actually has three types of focalization,
focalization zéro is defined as unlimited in contrast to limitation of perspective
(subdivided as external vs. internal). Where binary opposition does not do the job,
three or four categories must be offered. This tradition endures beyond Genette in
F. K. Stanzel’s structuralist reworking of his 1955 (trans. 1971) typology in A Theory
of Narrative ([1979] 1984) where he introduced three axes constituted by binary
oppositions, which in turn constitute his three narrative situations (see below).
A passion for typology and classification also characterizes much of the early narratological work of Mieke Bal, Gerald Prince, and Susan Lanser, and is prominent today in
the work of Ansgar Nünning.
The emphasis on binarism and typology highlights two prevalent features of
narratology – its aspirations to scientificity (via a ‘‘quasi’’-linguistic formalism and
empiricity) and its ultimately descriptive aims. Narratology’s ‘‘geometric imaginary’’
(Gibson 1996) projects the illusion that narrative is knowable and describable, and
therefore that its workings can be explained comprehensively. Narratology promised
to provide guidelines to interpretation uncontaminated by the subjectivism of traditional literary criticism. This attitude presupposes that texts are stable entities and
that readers react to them in foreseeable ways. More recent pragmatically oriented and
post-structuralist proponents of narrative theory, by contrast, have queried the stability of the text, arguing that textual analysis becomes affected by the reader’s
interaction with it,4 and some of them critique the very categories of linguistic
binarism and narratology’s obsession with typology (Gibson 1996).
The major problem of narratology in its early structuralist and typological phase,
however, lay in the difficult relation between theory and practice. On the one hand,
narratology claims to deliver a set of instruments for analyzing texts (this ‘‘toolbox’’
rationale of narratology is most prominently displayed in Nünning’s work); on the
other hand, narratology focuses on the why and the wherefore, the semiotics and
From Structuralism to the Present
39
grammar of narrative. In other words, narratology is both an applied science and a
theory of narrative texts in its own right. As applied narratology it faces the critical
challenge ‘‘So what? – What’s the use of all the subcategories for the understanding of
texts?’’ As a theory, narratology – like deconstruction or Lacanian psychoanalysis
– encounters the criticism that its theoretical proposals do not help to produce
significant readings. Unlike postcolonial or feminist interpretations of literature,
narratological analyses do not in themselves tend to produce entirely new readings
of a text; they frequently highlight how the text manages to have certain effects and
explain why these occur, thus providing arguments for existing interpretations of the
text. This is why narratology has performed so well with postmodernist narrative
where its instruments are eminently suited to demonstrating how mimetic traditions
are being contravened and playfully refunctionalized. Excellent studies in this area
are McHale ([1987] 1996, 1992) and Wolf (1993). Research on postmodernist texts
has also significantly extended narratological theorizing, for instance in relation to the
category of person (Fludernik 1994, Margolin 2000) and tense (Margolin 1999a).
From this perspective, narratology, therefore, is really a subdiscipline of aesthetic and
literary semiotics in so far as these specifically relate to narrative.
Narratology in its classical phase can be grouped into schools, although the
adequacy of a classification relying on mere nationality may be queried. Most
prominent is, obviously, the paradigm instituted by Gérard Genette in Narrative
Discourse (1980), whose international influence was cemented by its early translation
into English and by its adoption on the part of prominent American (Prince 1982,
1987), European (Bal [1985] 1997, 1999) and Israeli (Rimmon-Kenan [1983] 2002)
scholars. Genette’s narratological oeuvre is noteworthy above all for its careful
attention to the temporality of narrative, prominently displayed in Genette’s distinction between order, duration (tempo), and frequency. Thus, Genette’s term for a
flashback, analepsis, has become a household word in literary criticism, and –
especially in work dealing with postmodernist fiction – the term metalepsis, which
refers to a transgression of narrative levels, occurs again and again (e.g., McHale 1992,
[1987] 1996).5 Genette’s temporal model of order, tempo, and frequency decisively
extends Günther Müller’s 1948 distinction between Erzählzeit (narrating or discourse
time) and erzählte Zeit (narrated or story time). The category of tempo (including
pause, [stretch],6 scene, summary, and ellipsis) summarizes and systematizes Müller;
but Genette adds the important aspects of order (i.e., the narrative reordering of story
elements)7 and frequency, noting that events can be told more than once or that
recurring story elements may be condensed into one typical representation. Genette,
moreover, was the first to recognize that retrospective narration is not the only
possible temporal relationship between narration and story, drawing attention to
the existence of simultaneous and intercalated storytelling.
If Genette’s insights into narrative temporality proved attractive, even more so did
his distinction between homodiegesis and heterodiegesis, since this once and for all
eschewed the problem of having to explain why narrators such as in Fielding’s Tom
Jones are ‘‘third person’’ narrators even though they refer to themselves as ‘‘I.’’ Genette’s
40
Monika Fludernik
term heterodiegesis clarifies simply and elegantly that the narrator is not part of the
story world. However, Genette’s use of the root term diegetic in many technical terms
(homo-/heterodiegetic, extra-/intradiegetic, metadiegetic, etc.), though usefully allowing distinctions of person and level, at the same time introduced the problem that
diegesis, in the Greek original, actually referred to the narratorial discourse, that is, to
the act of telling, rather than to the story (the muthos or – in later narratological
parlance – the histoire). It is one of the reasons why German-speaking academics have
largely resisted the Genettean terminology.
Genette’s model, besides the named major insights, moreover was responsible for
establishing the concept of focalization in narrative studies, which has largely replaced
the traditional terms perspective and point of view. Genette’s famous distinction between
‘‘who speaks’’ (the narrator) and ‘‘who sees’’ (what Bal calls the focalizor [1997:
146–9]) helped to promote a narratology that maximizes discreteness and precision
in classification. Genette thus set himself apart from the much earlier model of F. K.
Stanzel ([1955] 1971) which was based on a holistic conception of prototypical
narrative situations in line with organicist and ‘‘morphological’’ frames current in
earlier (Friedemann [1910] 1865, Müller [1948] 1968) and contemporary (Lämmert
1955, Hamburger [1957] 1993) German work on narrative. Stanzel additionally
integrated American new critical insights by Percy Lubbock and Norman Friedman,
themselves inspired by Henry James.
Stanzel’s magnum opus, A Theory of Narrative, first published as Theorie des Erzählens
in 1979 and translated from the German revised 1982 edition in 1984, was, and to
some extent still is, the canonical narratological model in German-speaking countries
and in parts of Eastern Europe. In this book Stanzel revised his original model of three
narrative situations arranged around a ‘‘typological circle’’ along structuralist lines by
introducing three axes based on binary oppositions, with one pole of each axis
constituting the prototype of a narrative situation. Stanzel’s model differs from
Genette’s in its organicist and holistic frame, its emphasis on prototypicality, its
historical outlook, and its attention to textual dynamics.8 In lieu of Genette’s
pervasive binaries, Stanzel proposes a tripartite theoretical setup in three narrative
situations (NRSs) which correspond to prototypical versions of historically influential
novel types. The authorial NRS, figured in Fielding’s Tom Jones, combines an omnicommunicative (Sternberg [1978] 1993) narratorial presence up and above the world
of fiction (extradiegetic and heterodiegetic with zero focalization) with a panoramic
view of the fictional world and easy access to characters’ thoughts and emotions. The
narrator is typically intrusive and indulges in much metanarrative comment. By
contrast, in the prototypical figural NRS (e.g., Joyce’s A Portrait of the Artist as a
Young Man) the narrative conveys the illusion of unmediated access to the main
protagonist’s mind and there is no foregrounded narrator persona. Finally, there is
the first-person NRS (e.g., Dickens’s David Copperfield) in which the narrator looks back
on his previous experience, evaluating it in his function as a narrating self, but often
immersing himself and the reader in his past experiencing self. Since the three NRSs
are merely prototypes, individual texts may combine characteristics of these NRSs. In
From Structuralism to the Present
41
particular, nineteenth-century fiction displays what Stanzel calls the authorial-figural
continuum, the frequent move of the narrative between external and internal perspectives in a given section of the narrative.
In his revision of the original typology in 1979, Stanzel grounded the three NRSs
on three axes. The category person (homo-/heterodiegesis) is founded on the identity
vs. nonidentity of narratorial and fictional worlds with the first-person NRS prototypically situated at the pole of identity, the category perspective (external vs.
internal) defines the authorial NRS as prototypically dominated by external perspective, and the category mode (teller vs. reflector mode) defines the figural NRS as
constituted by the reflector pole. Stanzel placed these NRSs on a typological circle not
only for the sake of categorization but in order to illustrate the continua between
narrative forms and the principal openness of categorization. Despite these confessed
aims, the model provoked heated debates (Stanzel 1978, 1981, Cohn 1981, Genette
1988, Chatman 1990).
In addition to these typological arrangements, Stanzel was particularly concerned
with the history of the novel and the primacy of authorial narrative before first-person
texts and later the figural novel. Stanzel also analyzed the way in which novel
openings tend to foreground narratorial discourse and how later in the book, the
profile of the narrator weakens. He thus implicitly incorporated the reader into his
account and anticipated the pragmatic and historical emphasis of later narratological
studies.
Among the most influential scholars coming after Stanzel and Genette in narratology’s classical phase, three will be discussed specifically at this point.9 Gerald Prince
not only provided readers with the first dictionary of narratological terms (now
complemented by the Schellinger 1998 Encyclopedia of the Novel and the Herman,
Jahn, and Ryan forthcoming Routledge Encyclopedia of Narrative Theory); he was also
responsible for two important refinements of the Genettean model – the introduction
of the narratee (narrataire), the text-internal figure whom the narrator addresses, and,
secondly, the extensive definition of narrativehood10 (what makes a narrative narrative) and of degrees of narrativity (the well-formedness of narratives). The first
innovation has given rise to a spate of communication-oriented models (e.g., Coste
1989, Sell 2000). Prince’s attempt to define the basic requirements for a text to be
considered a narrative inspired major reformulations of the term and concept of
‘‘narrativity’’ in White (1980), Chatman (1990), Ryan (1992), Sturgess (1992),
Fludernik (1996), Prince (1995, 1999), McHale (2001), Sternberg (2001), and
Herman (2002).
The second critic I wish to discuss here is the Dutch scholar Mieke Bal, who
proposed an originally controversial and now quite current extension of Genette’s
focalization theory. According to Bal, focalization properly defined requires both a
focalizor and an object of focalization. She therefore distinguishes between who does
the focalizing (an extradiegetic narrator, a character) and what is being focused on (the
external behavior of a character or the character’s mind – this corresponds to Genette’s
external vs. internal focalization).11 Bal therefore rewrites Genette’s model which was
42
Monika Fludernik
based on the limitation of perspective – zero focalization (no limitation) vs.
limited external or internal focalization – into a neat binarism of focalizors and
focalized. This then results in the introduction of a narrator-focalizor, a narrator
who ‘‘sees.’’ Bal’s second important innovation was her extension of narratology to
cover film (already discussed in Chatman 1978: 96), ballet, and drama (Bal 1997: 5).
Bal – despite her keen interest in the ideological functions of narrative – opposed the
current anthropomorphization of the narrator figure, marking the merely functional
quality of the narrational instance with the impersonal pronoun ‘‘it.’’ In her later
postnarratological research, Bal went on to focus on painting as narrative and
as ideology, particularly in relation to biblical subjects and from a feminist perspective. In this she anticipated later developments in New Historicism and cultural
studies.
The third of the triumvirate to be specifically noted in this section is Seymour
Chatman, who wrote the standard textbook of narratological study in the United
States, Story and Discourse (1978).12 Chatman’s shrewd presentation of the two
fundamental levels of narrative – story (what the narrative is about) and discourse
(the text) – served to reformulate the question of what constitutes a narrative by
extending the definition to cover a variety of narrative media, especially film.
Narrative discourse in Chatman’s model can therefore come in different shapes –
narratorial discourse, filmic sequences, and so on. Chatman’s definition of narrativity in Story and Discourse reposes on the dynamic interrelation of the two levels
of story (plot) and discourse (medial representation). Chatman’s important second
theoretical volume, Coming to Terms (1990), significantly expanded his concept of
narrativity by integrating narrative with other text types (like argument, description, etc.), thus opening the way for a cross-generic and linguistic discussion of
the specifics of narrative.13 Besides introducing the useful distinction between
covert and overt narrators, Chatman has additionally enriched our terminology by
his distinction between narratorial slant and character-related filter to denote
limitations on perspective; this attempt to separate ideological from perceptionbased types of point of view has been extremely influential in studies of focalization and unreliability.14 Moreover, Chatman was one of the first critics to
analyze film as a narrative genre. He thus initiated a line of enquiry into film
as narrative that complements key studies by Metz (1971), Branigan (1984,
1992), Bordwell (1985), and others. More controversially, Chatman also introduced the concept of the ‘‘cinematic narrator’’ (1990: 126–34) to film studies,
thus attempting to supply filmic discourse with an equivalent of verbal narratives’
narrator function.
Classical narratology, as we have seen, developed a terminology to describe textual
diversity and it instituted a number of key categories for a narrative grammar and
poetics. Among these, the story–discourse distinction is perhaps the most fundamental. Secondly, the conception of narrative as communication resulted in a more
extended list of narrative instances besides the author and the narrator (invented as a
separate figure by K. Friedemann and Wolfgang Kayser): the narratee, the implied
From Structuralism to the Present
43
author (Booth [1961] 1983), and the implied reader (Iser [1972] 1990). At the same
time, the pronominal, temporal, and expressive features of narrative came into view
and engendered a plethora of new categories as mentioned above (Stanzel’s narrative
situations; Genette’s homo-/heterodiegesis, order, tempo, frequency; Chatman’s slant
and filter; Bal’s focalizor; Lanser’s communicational and perspective categories, etc.).
The issue of tense in narrative especially gave rise to numerous incisive studies
(Weinrich [1964] 1985, Ricoeur 1984–88, Fleischman 1990). Alongside focalization,
the issue of narrative level and especially the status of characters’ discourse also
attracted extensive attention. A major preoccupation of narratologists has been the
presentation of consciousness in narrative, especially in the shape of free indirect
discourse (Cohn 1978, McHale 1978, 1983, Banfield 1982, Sternberg 1982, Fludernik 1993).
Although classical models started out with a structuralist fixation on plot and
narrative grammars15 in Todorov, Barthes, Bremond, and Greimas, narratology in the
1970s and 1980s was mostly concerned with discourse and narration rather than plot.
Plot only moved back into focus in Peter Brooks’s psychoanalytic study Reading for the
Plot (1985) and, more recently, in the work of Marie-Laure Ryan (1991, 1992) and
David Herman (2002). Curiously, another area of narratology that has remained
somewhat underresearched is that of setting and character, even though there are at
least some superb essays on the latter topic by Philippe Hamon (1972), Uri Margolin
(1990, 1995, 1996) and Ralf Schneider (2001).16 Finally, narrative mimesis and the
issue of fictionality became major focusing points, particularly in the context of
postmodernist literature which seemed to contradict the narratological models of
Genette and Stanzel since these were based on the eighteenth-century to early
twentieth-century realist novel. Whereas classic narratology started out by analyzing
postmodernist texts as deviating from realistic and verisimilar parameters (no plot,
contradictory character, illogical concatenation of action sequences, etc.), studies such
as Wolf (1993) have helped to go beyond the description of defamiliarizing devices to
the analysis of metafictional strategies in the framework of fictionality and the
constitution of aesthetic illusion. Important studies contrasting fiction and historical
narrative were Ricoeur’s Time and Narrative (1984–8) and essays by Gérard Genette
and Dorrit Cohn (republished in Genette 1991 and Cohn 1999). The extension of
narratological analysis to historiography and generally to nonfictional narrative
occurred in the wake of the ‘‘narrative turn’’ in historical studies which is centrally
linked to the name of Hayden White. White, in his analyses of nineteenth-century
historiography, demonstrated the presence of literary generic frames in the narrative
writing of history (especially 1973, 1987). More importantly, he argued that plot
construction (i.e., narrativization) affects historiographic discourse as much as it does
literary narrative. The narratological analysis of twentieth-century historical texts
(Carrard 1992) and research on faction (the nonfiction novel)17 moreover increasingly
emphasized the continuity between literary and nonliterary narrative and advanced
the issue of narrative strategies employed in different ways or combinations in either
realm.
44
Monika Fludernik
Let us now turn to the first major paradigm shift in narrative studies which set in
in the early 1980s and was closely related to contemporary trends in critical theory in
the United States.
Beyond Form: Pragmatics, Gender, and Ideology
Contextual narratology emerged from two main sources. The first source can be
identified as the native American tradition in the wake of Henry James’s The Art of
the Novel (1907–17; James [1934] 1953), followed by work on narrative perspective
(Percy Lubbock 1921 and Norman Friedman 1955), which reached its apogee in
Wayne C. Booth’s studies of irony in The Rhetoric of Fiction ([1961] 1983) and A
Rhetoric of Irony (1974).18 A second major source was the pragmatic revolution in
linguistics which displaced linguistic structuralism and generative grammar and
reintroduced semantics, context-orientation, and textual issues to the study of language. Whereas earlier, the linguistic model had inspired the development of text
grammars (e.g., van Dijk 1972, Petöfi and Rieser 1973, Petöfi 1979) and semiotics
(Lotman 1977), these were now displaced by the models of text linguistics, speech act
theory, sociolinguistics, and conversation analysis which were offering methodologies
and concepts eminently suitable for literary application. In particular, linguistic
pragmatics emphasized the notion of function in ways that allowed for multiple
relations between form and function. Thus, Meir Sternberg’s and other members’
work at the Porter Institute for Poetics and Semiotics in Tel Aviv established that one
form could correspond to several functions, and one function be marked with several
surface structure elements. This research decisively promoted functional analysis and
frame-theoretical concerns, thus laying the foundation for more cognitivist studies in
the 1990s.19
Conversational narrative, analyzed by William Labov in the 1970s and by Deborah
Tannen, Wallace Chafe, and others in the 1990s, had become a key area of narratological research in Germany in the late 1970s (Harweg 1975, Ehlich 1980, Quasthoff
1980, 1999, Stempel 1986) and centrally influenced postclassical narratological work
by David Herman (1997, 2002) and Monika Fludernik (1991, 1993, 1996). At the
same time, the aesthetic and quasi-fictional aspects of everyday storytelling were
increasingly being discussed in linguistics (Tannen 1982, 1984, Norrick 2000, Ochs
and Capps 2001). This research has resulted in an extension of the narratological
corpus, now including oral language narrative as well as the novel and short story.
A second major area of narratological research came to fruition in the context of
feminist literary criticism. In 1988, a heated debate arose between Susan Lanser
(1986, 1988) and Nilli Diengott (1988) about the blindness of narratology regarding
issues of gender. Lanser, who is still the major proponent of gender-oriented narratology (Lanser 1992, 1995, 1999), mainly concentrates on the genderization of
narrator figures (cf. also Fludernik 1999). Thus, the question of a narrator’s properties
(overt/covert, homo-/heterodiegetic, etc.) is extended into the field of sex and gender;
From Structuralism to the Present
45
the explicit naming, description, or actions of narrator figures need to be compared
with implied genderization by means of dress codes, behavioral patterns, and cultural
presuppositions. Such analyses are pragmatic in the original linguistic sense of the
term since they involve the readers’ interpretative strategies, speculations, and guesswork. Instead of seeking to rewrite the major narratological categories, Robyn Warhol
has proposed the concept of the ‘‘engaging narrator’’ in an attempt to discuss different
types of narratorial discourse in male- and female-authored texts (Warhol 1989,
2003). Here, too, narratees and actual readers are signally involved. Besides the
issue of narratorial sex or gender, ‘‘female’’ vs. ‘‘male’’ plot structures (Mezei 1996,
Page 2003) and revisions of the history of narrative from a female perspective
(e.g., Aphra Behn as first novelist – Fludernik 1996) have been additional foci of
research. At the same time, Roof (1996) and Lanser (1995, 1999) have extended
feminist research into queer studies.20
A third vast area of narratological research that developed in the 1980s and 1990s
is ultimately ideological in orientation. The emphasis in this work derives from newer
theoretical disciplines such as postcolonial criticism, New Historicism, or cultural
studies. To take postcolonial criticism as an example, in this kind of narratological
analysis, the main point concerns the question of how the text is imbued with
(neo)colonial discourse that colludes with the oppression of the native population
and how the discourse at the same time ends up undermining this ideology. Postcolonial and feminist theory or gender studies frequently join forces in the analysis of
orientalist, exoticist, and colonial discourses since patriarchal and (neo)colonial patterns can be observed to operate in tandem in literary texts, travelogues, or historical
writing. Postcolonial narratological criticism attempts to describe how the choice of
specific narrative techniques helps to transmit underlying orientalist or patriarchal
structures and how the narrative, by its choice of focalization, plot structure, or use of
free indirect discourse sometimes resists these structures, undermines or deconstructs
them. Finally, postcolonial narratology is concerned with experimental narrative
techniques that correlate with the celebration of cultural hybridity or the symbolic
liberation of the subaltern. (For criticism in this vein, see Pratt 1992, Spurr 1993,
Fludernik 2000b, Richardson 2001a.)
Lack of space prevents me from illustrating similar research for earlier Marxist or
cultural materialist analyses of narrative or to discuss the new-historicist uses of
narrative theory. (For good surveys see Eagleton 1996 and Cobley 2001.) Marxist
readings of the novel and the uses of the concept of ideology by Marx and Engels as
well as Althusser were already crucially concerned with narrative and the novel (as
witness the work of Georg Lukács). What I want to emphasize here, however, is the
increasing turn within feminist, gender-oriented, postcolonial, and ideological criticism in general toward a symptomatic reading of texts. Like earlier psychoanalytic work,
critical discussion of texts and genres attempts to locate textual strategies that signal
unconscious or repressed psychological or ideological ‘‘drives’’ which the critic uncovers, reading ‘‘against the grain’’ of the text. This disclosure of the text’s secret
motivations, of which the author may be entirely unaware, tends to change the
46
Monika Fludernik
pattern of interpretation of narrative techniques as well. There are narrative and
rhetorical techniques used by the author to convey his or her meanings, but these
techniques become split between their ostensible signifying functions and their secret
and insidious conveyings of ideological purport. The scenario is even further complicated by the fact that postcolonial, feminist, and Marxist critics frequently detect
signs in which the text surreptitiously seems to undermine or put in doubt its
ostensible ideological drift – as when, by seemingly praising patriarchal structures,
a criticism of them can be gleaned from the text. Thus, in many of Kipling’s short
stories, the narrative voice ostensibly participates in a colonialist and orientalist
discourse, but the plots of the tales undermine the lessons of British superiority and
contempt for the natives, and the narrator seems to be uttering his commonplaces
tongue in cheek (‘‘Beyond the Pale’’ is a good example of such a constellation). What
has not been systematically analyzed in narratology are the structural patterns to
which such readings resort (but see Chambers 1991, Sinfield 1992): how is it possible
for narrative techniques to mean both one thing and its opposite at the same time?
What types of narrative techniques are at issue in such cases? Most frequently, no
specific formal elements signal ideological symptoms or their subversion; rather, the
frame within which the texts are read opens up avenues of interpretation that within a
different (patriarchal, colonial, capitalist, etc.) frame were not initially salient in the
reading process.
This kind of broadening of the complexity of narrative issues takes me to the next
section in which recent extensions of the term narrative are discussed. Whereas, in the
1960s, everything became a text (including Paris viewed from the Eiffel Tower), we
are now experiencing a ‘‘narrative turn.’’
The ‘‘Narrative Turn’’ and the Media
In this section I want to discuss developments in media studies and the generalization
of the term ‘‘narrative’’ (see Bal 1999) within a wide spectrum of the social sciences,
resulting in the application of narratological paradigms to legal, medical, psychological, or economic discourses. Such extensions are always charged with tension since
the appropriation of narratological frameworks by nonliterary disciplines often results
in the dilution of the narratological basis, in a loss of precision, and the metaphoric
use of narratological terminology. On the other hand, as Rimmon-Kenan has argued
(2001), narrative theory needs to come to terms with the deployment of its concepts
in nonnarratological contexts. In psychoanalytic practice, the example she gives,
patients are encouraged to construct life stories for themselves that allow them to
project a positive identity (cf. Eakins 1999). As Cohn (1999: 38–57) demonstrates,
Freud’s clinical narratives were significantly different from fictional and real-world
storytelling. What is particularly interesting in this context is that odd uses of the
concept ‘‘narrative’’ frequently occur in conjunction with acts of storytelling. Thus,
many social sciences use narratives from interviews as their source material to then
From Structuralism to the Present
47
construct what they call a ‘‘narrative’’ of, say, homelessness or a ‘‘narrative’’ of
criminality. These narratives are rationalizations of the behavior and experience
observed in their subjects but not narratives in any traditional narratological sense.
Even in policy training, narrative strategies are prominent these days (Roe 1994).
Instead of merely rejecting the application of narratological terminology to different
subjects, narrative theory must try to theorize the extension of the concept and
propose theoretical frameworks that counteract the attendant loss of precision (Nünning and Nünning 2002: 3–5).
I would like to start with psychoanalysis, whose narratological heyday goes back to
the 1980s when Ross Chambers (1984) and Peter Brooks (1985) published important
studies using psychoanalytic frameworks. In narrative, three main applications of
psychoanalytic concepts are possible: analyzing the author, analyzing the character(s),
and subjecting the reader’s relation to the text to psychoanalytic interpretation.
Depending on the various paradigms, such analysis can be based on Freudian,
Lacanian, Kleinian, or other theoretical frameworks. Major psychoanalytic literary
studies also focus on specific concepts such as repression, transference, or hysteria, and
many literary critics have been particularly concerned with the treatment of women in
the text, reading the often phallocentric psychoanalytic frameworks against their
grain.21 By contrast, recent work in psychoanalysis not only relies on patients’
storytelling but methodologically utilizes dialogue and narrative for clinical purposes
(Schafer 1992). Such uses of narrative in therapy and in the theoretization of the
therapeutic process deploy narrative concepts in an interdisciplinary manner – not
that literary texts are studied by recourse to the framework of classical psychoanalysis,
but therapist–client interaction is analyzed within a narrative framework and
methodology.
In legal discourse, too, narrative has for a long time played a crucial role. Witnesses
and defendants tell stories; police, prosecutors, defense lawyers, and judges produce
narratives of legitimation that juries need to find convincing. Moreover, narratives of
guilt or innocence are frequently based on the interpretation of evidence similar to
psychoanalytic reconstructions from patients’ symptoms. In the past 20 years, a new
area of research called ‘‘Law and Literature’’ has developed which is concerned with
literary thematizations of legal concerns (Weisberg 1984, 1992, Dimock 1996,
Brooks 2000, Thomas 1987, 2001) and with the influence of legal (especially
metaphorical) language on legal practice (Hyde 1997).22 Much of this research
involves narrative concerns.
Although narrative has become an important term in many other disciplines, the two
discussed above should suffice to make the point that narratology is increasingly
appealed to as a master discipline. Narratology thus finds itself confronted with
nonliterary and even nonlinguistic uses of the concept of narrative among which the
literary and poetic framework of traditional narratology seems to lose in significance.
Besides these extensions of the term narrative, new departures in narrative study
also include possible worlds theory and its recent transfer of narratological analysis
into hypertext and communication studies. Possible worlds theory has a long
48
Monika Fludernik
philosophical tradition starting with Saul A. Kripke. It originally emerged from the
attempt to solve the problem of fictional reference. The concept of possible worlds (in
which reference is no longer a problem) was first utilized in narratological research by
Thomas G. Pavel and Lubomı́r Doležel (Pavel 1986, Doležel 1998a, 1998b)23 and is
now associated with the work of Marie-Laure Ryan (1991, 1999, 2001). Its recent
proposals have been especially important to the elucidation of plot structure, the
reader’s manipulation of alternative plot developments, or the characters’ planned or
fantasized alternative action series. One of the major innovations of this approach is
the concept of transworld identity which characterizes protagonists like Napoleon or
Charlemagne in fictional texts. Ryan has used insights from possible worlds theory to
reformulate the concept of narrativity (Ryan 1992) and to analyze hyperfictions from a
narratological perspective (Ryan 2001). Parallel work touches on the grammar of
virtuality (Margolin 1999b), which is closely linked to possible world scenarios.
Other transgeneric and intermedial extensions of narratology are presented in a
recent survey volume by Nünning and Nünning (2002). Especially noteworthy is the
increasing attention paid in narrative study to drama (Richardson 1987, 1988,
2001b, Jahn 2001) – a genre formerly excluded from narratology by nearly everyone
except Mieke Bal (1997) and Manfred Pfister, whose influential theory of drama
utilizes many categories from narratology (Pfister [1977] 1991). Nünning’s subsumption of drama under the umbrella of transgeneric narratology is indeed partly based on
his adoption of Pfister’s theory of open versus closed perspective structure which was
developed for the analysis of plays but which Nünning and Nünning (2000, 2002)
have now centrally reintegrated into narratology (cf. Fludernik 2003b). Nünning and
Nünning (2002) additionally include an analysis of cartoons, painting, poetry, and
music as narrative (see especially Wolf 1999, 2002). Pictorial narrative and music also
figure prominently in Marie-Laure Ryan’s Narrative Across Media (2004) and are
discussed in separate essays in this volume.
The Present: The Cognitivist Turn and the Resurrection
of the Linguistic Model
Concurrently with the ‘‘narrative turn’’ in the social sciences and narrative theory’s
move into media studies, another major shift was taking place which one could dub
the ‘‘cognitivist turn’’ (Jahn 1997). Increasingly, narrative theory began to reorient
itself toward the cognitive roots already present in the structuralist and morphological traditions (Herman 2002), additionally absorbing insights from cognitive
linguistics and empirical cognitive studies. One way to map the history of narratology
is therefore to see it as adopting linguistic paradigms one by one as they arose in the
twentieth century – structuralism (classical narratology); generativist linguistics (text
grammars); semantics and pragmatics (speech act theory, politeness issues, etc.); text
linguistics (conversation analysis and critical discourse analysis); and now cognitive
linguistics (cognitivist narratology). Cognitive linguistics analyzes the way in which
From Structuralism to the Present
49
language structure is predetermined by human cognition. Cognitive linguistic
approaches cover a wide range of subjects. These include the lexicalization of color
terms and spatial reference, the analysis of constitutive body schemata as they affect
metaphoric and conceptual thinking, and the crucial importance of schemata and
prototypes in human cognition and hence in language structure.
The cognitivist turn in narrative study concerns two basic levels. On the one hand,
it focuses on humans’ perceptions of actions and events from a cognitive viewpoint; on
the other hand, it analyzes narrative structures (as transmitted in texts) and how these
obey fundamental cognitive parameters or frames (see van Dijk and Kintsch 1983).24
Whereas David Herman’s monumental Story Logic (2002) concentrates on the first
approach, Fludernik’s model of the narrative episode and its textual surface structures
tends to focus on the second level. At the same time, the cognitivist paradigm shift
has produced two major lines of methodology, one focusing on conversation analysis
and narrative in the oral language as a prototype for literary and written narrative; the
other focusing on constructivist presuppositions about the reader–text relationship. In
practice, the various studies share aspects of all of these orientations. It is, however,
useful to provide a rough pattern of developments since this enables one to place
individual projects as offshoots of one rather than the other axis of analysis. Many of
these approaches are represented in Herman (2003).
Basically, David Herman’s Story Logic is focused on conversational narrative, which
he sees as a paradigm for narrative in general. Herman’s major cognitivist thrust
concerns the introduction of preference rules that are generically based and could be
expanded historically. Following Herman’s timely collection of essays, Narratologies
(1999) – a survey of current narratological criticism at the turn of the millennium –
his magisterial Story Logic delves into philosophy, linguistics, conversation analysis,
and cognitive theory to propose a system of microdesigns and macrodesigns governing the logic of story(telling). In this important book traditional narratological
categories like plot, perspective, person, or narratees (among others) are integrated
into a larger framework inspired by a wide range of theoretical approaches based in
(philosophical) linguistics. Whereas Herman’s emphasis is on the logic of narrative and
his focus is on story rather than discourse, Marisa Bortolussi’s and Peter Dixon’s
timely Psychonarratology (2003) privileges the empirical analysis of readers’ engagements with texts and demonstrates how ever more sophisticated experiments can help
determine which textual features are responsible for certain ‘‘literary’’ reactions on the
part of readers, or whether narratological categories like focalization are useful in
actual empirical work.
Fludernik (1996, 2003a), by contrast, combines a strong emphasis on the prototypicality of conversational narrative with an emphasis on textual surface structure
and a diachronic perspective. The approach relies on cognitivist paradigms to the
extent that it analyzes narrative patterns whose shape is determined not merely by
linguistic markers on the surface structure but in addition by the active narrativization of the text on the part of readers. Fludernik’s work moreover displays a strong
interest in historical or diachronic issues. Previously, Stanzel had been the only early
50
Monika Fludernik
theoretician with an important historical orientation. The history of narrative forms
also played a role in feminist studies and in a number of German habilitations
(e.g., Korte 1997). Fludernik (1996, 2003a) put the diachronic analysis of narrative
back on the agenda, particularly with its focus on the development of narrative from
the Middle English period to the early novel. The analysis of medieval and early
modern texts promises to have important theoretical repercussions since current
narratological categories are still by and large based on the novel between (maximally)
1700 and the 1990s.
Another major proponent of the cognitivist turn is Ansgar Nünning, who explicitly declares his narratological oeuvre an effort in constructivism.25 Nünning started
out with analyses of the narratorial functions in the nineteenth-century novel (1989)
and quickly came to focus on a critique of the implied author (see Nünning’s chapter
in this volume) and the revision of the concept of unreliable narration (Nünning
1999a, 1999b).26 Nünning’s work tries to steer a middle course between the textual
analysis of signals helping the reader to recognize unreliability and the reader’s active
construction of an unreliable narrator by way of interpretative strategy. Nünning and
Nünning’s more recent work on multiperspectivism (2000) expands this line of
argument to its logical conclusion by tackling the general process of assigning overall
meaning to narratives, replacing the implied author concept by a theory of constructivist attributions of overall textual meaning. This work – despite its criticism of
Booth – actually closely relates to the ethics of narrative which is one of Booth’s major
legacies.
This brings us back to Wayne C. Booth and to the closing of a circle that originated
with classical narratology and its typological preferences and has meanwhile encompassed a decisive broadening of the subject and the methodologies of narrative theory.
Having started out with a focus on the novel, narratology is now held responsible for
explaining narrative in general – and this includes conversational storytelling, narrative representations in medical or legal contexts, historiography, news stories, films,
ballets, plays, video clips, and much more. At the same time, narratology has
absorbed insights from critical theory, molded itself into feminist, psychoanalytic,
and postcolonial shapes, and has adopted text linguistic, cognitivist, constructivist,
and empirical models for its various frameworks. Perhaps this flexibility has earned
the discipline its present vitality and productivity – disciplines chained to one specific
methodology like deconstruction have had a harder time surviving in the competitive
climate of the 1990s. Like David Herman, I too see the developments between the
1970s and the present as a conflux of different outlooks and emphases that combine in
interesting ways in the work of individual narratologists. Nevertheless, the cognitivist turn in narratology also allows one to postulate an entirely different story – namely
that of a resuscitation of linguistic models in recent narratological research. On this
view, classical narratology was unable to prove the empirical relevance of its linguistic
categories in interdisciplinary research; with the move of linguistics into cognitive
studies the research produced by narratologists seems to have acquired a new lease on
life. Thanks to the narrative turn in the humanities even linguists are taking
From Structuralism to the Present
51
narratology more seriously. The cognitivist paradigm shift could thus pave the way
for a much closer companionship of narratology with the empirical sciences and,
perhaps, come a long way towards fulfilling narratology’s original aspirations toward a
scientific image.27
NOTES
1 This is the title of Chapter 3 in Cobley’s
Narrative (2001).
2 Although the term discourse analysis is current, I have avoided it in this presentation
because of its connotations of Foucauldian
discourse analysis and linguistic analogues
in critical discourse analysis (e.g., Fairclough
1995).
3 For introductions to narrative theory and
narratological terms for beginners see Chatman (1978, 1993), Jahn and Nünning
(1994), Bal (1997), Rimmon-Kenan (2002),
and Jahn’s homepage (Jahn 2003) which includes a study course for German MA students.
4 See especially Sternberg (1982, 1983, 1987,
[1978] 1993), Yacobi (1981, 1987, 2000),
Fludernik (1993, 1996), and Herman (2002).
5 In November 2002 there was even a symposium on metalepsis in Paris, at which Gérard
Genette participated (see Pier and Schaeffer
forthcoming).
6 Some narratologists have added a ‘‘slow
motion’’ category (Chatman 1978: 72–3,
Martinez and Scheffel 1999: 43–4).
7 This topic has meanwhile received an even
more extensive treatment by Ireland (2001).
Genette’s analepses and prolepses reformulate
Eberhard Lämmert’s analyses (1955: 100–89).
8 For a good summary of Stanzel’s model see
Stanzel (1978, 1981) and Cohn (1981).
9 Although Lanser (1981) is a major landmark
in narrative theory, I have shifted a discussion
of her work to the feminist section since her
later feminist contributions are the ones that
were most influential.
10 Subsequently renamed ‘‘narrativeness’’ (Prince
1999).
11 See also the incisive reformulations in French
criticism (Vitoux 1982, Cordesse 1988) as
12
13
14
15
16
17
18
19
20
21
22
23
24
well as Lanser (1981), Nelles (1990),
Edmiston (1991), Jahn (1996, 1999), Niederhoff (2001), Shen (2001), and Nieragden
(2002).
Chatman also authored a very handy introduction to narrative theory (Chatman 1993),
which unfortunately is no longer in print.
Compare also Fludernik (1996: 352–8,
2000c) as well as Wolf (1998) on narrative
vs. lyric poetry and narrative in painting and
music (Wolf 2002).
See the essays by Prince and Phelan in van
Peer and Chatman (2001).
On text grammars in the 1970s see Petöfi
(1979) as representative of a group of text
linguists as well as Pavel (1973).
See also Phelan (1989b).
See Hollowell (1977) and Foley (1986).
Booth’s work is part of what is sometimes
called the Chicago School of literary criticism
which displayed a strong emphasis on rhetorical models (see also Phelan 1996) and on
narrative ethics, a subject to which Booth
contributed significantly in his later books.
On ethics in literature see also Gibson
(1999).
See especially Sternberg (1983).
For a good survey of a gender-focused narratology see also Nünning and Nünning
(2004) and Prince (1996).
Relevant studies are too numerous to mention. Note the early work by Shoshana Felman (1982, 1985) and several studies by
Elisabeth Bronfen (1992, 1998).
For a good survey see Kayman (2002).
For a good summary see Doležel (1998b).
Another related strand of cognitive poetics is
closely aligned with cognitive metaphor theory (Stockwell 2002, Semino and Culpepper
2002).
52
Monika Fludernik
25 Nünning and Nünning (2002) emphasizes
the cognitive basis of narrativity.
26 The original volume of essays in German
came out in 1998; the two papers from
1999 modify the earlier model.
27 A first draft of this paper has received extremely generous feedback from J. Alber,
D. Herman, M. Jahn, U. Margolin, and
REFERENCES
AND
Armstrong, N. and Tennenhouse, L. (1993). ‘‘History, Poststructuralism, and the Question of
Narrative.’’ Narrative 1(1), 45–8.
Bal, M. ([1985] 1997). Narratology: Introduction to
the Theory of Narrative, 2nd edn. Toronto:
University of Toronto Press.
Bal, M. (1999). ‘‘Close Reading Today: From
Narratology to Cultural Analysis.’’ In
W. Grünzweig and A. Solbach (eds.), Grenzüberschreitungen: Narratologie im Kontext/Transcending Boundaries: Narratology in Context
(pp. 19–40). Tübingen: Narr.
Banfield, A. (1982). Unspeakable Sentences: Narration and Representation in the Language of Fiction.
Boston: Routledge & Kegan Paul.
Booth, W. C. (1974). A Rhetoric of Irony. Chicago:
University of Chicago Press.
Booth, W. C. ([1961] 1983). The Rhetoric of Fiction, 2nd edn. Chicago and London: University
of Chicago Press.
Booth, W. C. (1988). The Company We Keep: An
Ethics of Fiction. Berkeley: University of California Press.
Bordwell, D. (1985). Narration in the Fiction Film.
Madison: University of Wisconsin Press.
Bortolussi, M. and Dixon, P. (2003). Psychonarratology: Foundations for the Empirical Study of Literary Response. Cambridge, UK: Cambridge
University Press.
Branigan, E. (1984). Point of View in the Cinema:
A Theory of Narration and Subjectivity in Classical
Film. Berlin: Mouton.
Branigan, E. (1992). Narrative Comprehension and
Film. London: Routledge.
Bremond, C. (1973). Logique du récit. Paris:
Seuil.
Bronfen, E. (1992). Over Her Dead Body: Death,
Femininity and the Aesthetic. Manchester, UK:
Manchester University Press.
G. Olson, to whom I wish to express my
grateful acknowledgment for constructive
criticisms of the piece. Revisions were also
made in response to comments from Brian
Richardson. Given the many partly conflicting suggestions I received, I take full
responsibility for this final version of the
paper.
FURTHER READING
Bronfen, E. (1998). The Knotted Subject: Hysteria
and its Discontents. Princeton, NJ: Princeton
University Press.
Brooks, P. (1985). Reading for the Plot: Design and
Intention in Narrative. New York: Vintage.
Brooks P. (2000). Troubling Confessions: Speaking
Guilt in Law and Literature. Chicago: Chicago
University Press.
Carrard, P. (1992). Poetics of the New History:
French Historical Discourse from Braudel to
Chartier. Baltimore: Johns Hopkins University
Press.
Chafe, W. L. (1994). Discourse, Consciousness, and
Time: The Flow and Displacement of Conscious
Experience in Speaking and Writing. Chicago:
University of Chicago Press.
Chambers, R. (1984). Story and Situation: Narrative
Seduction and the Power of Fiction. Theory and
History of Literature, 12. Minneapolis: University of Minnesota Press.
Chambers, R. (1991). Room for Maneuver: Reading
Oppositional Narrative. Chicago: University of
Chicago Press.
Chatman, S. (1978). Story and Discourse: Narrative
Structure in Fiction and Film. Ithaca, NY: Cornell
University Press.
Chatman, S. (1990). Coming To Terms: The Rhetoric
of Narrative in Fiction and Film. Ithaca, NY:
Cornell University Press.
Chatman, S. (1993). Reading Narrative Fiction.
New York: Macmillan.
Cobley, P. (2001). Narrative: The New Critical
Idiom. London: Routledge.
Cohan, S. and Shires, L. M. (1988). Telling Stories:
A Theoretical Analysis of Narrative Fiction. New
York: Routledge.
Cohn, D. (1978). Transparent Minds: Narrative
Modes for Presenting Consciousness in Fiction.
Princeton, NJ: Princeton University Press.
From Structuralism to the Present
Cohn, D. (1981). ‘‘The Encirclement of Narrative:
On Franz Stanzel’s Theorie des Erzählens.’’ Poetics
Today 2(2), 157–82.
Cohn, D. (1999). The Distinction of Fiction. Baltimore: Johns Hopkins University Press.
Cordesse, G. (1988). ‘‘Narration et focalisation.’’
Poétique, 19(76), 487–98.
Coste, D. (1989). Narrative as Communication. Theory and History of Literature, 64. Minneapolis:
University of Minnesota Press.
Couturier, M. (1995). La Figure de l’auteur. Paris:
Seuil.
Diengott, N. (1988). ‘‘Narratology and Feminism.’’ Style, 22(1), 42–51.
Dijk, T. van (1972). Some Aspects of Text Grammars:
A Study in Theoretical Linguistics and Poetics. The
Hague: Mouton.
Dijk, T.A. van and Kintsch, W. (1983). Strategies
of Discourse Comprehension. New York: Academic
Press.
Dimock, W.-C. (1996). Residues of Justice: Literature, Law, Philosophy. Berkeley, CA: University
of California Press.
Doležel, L. (1998a). Heterocosmica: Fiction and Possible Worlds. Baltimore, MD: Johns Hopkins
University Press.
Doležel, L. (1998b). ‘‘Possible Worlds of
Fiction and History.’’ New Literary History 29,
785–810.
Duchan, J. F., Bruder, G. A., and Hewitt, L. E.
(eds.) (1995). Deixis in Narrative: A Cognitive
Science Perspective. Hillsdale, NJ: Erlbaum.
Edmiston, W. F. (1991). Hindsight and Insight:
Focalization in Four Eighteenth-Century French
Novels. University Park, PA: Pennsylvania
State University Press.
Eagleton, T. (ed.) (1996). Marxist Literary Theory:
A Reader. Oxford: Blackwell.
Eakins, P. J. (1999). How Our Lives Become Stories:
Making Selves. Ithaca, NY: Cornell University
Press.
Ehlich, K. (ed.) (1980). Erzählen im Alltag. Frankfurt/Main: Suhrkamp.
Fairclough, N. (1995). Critical Discourse Analysis:
The Critical Study of Language. London: Longman.
Felman, S. (1982). Literature and Psychoanalysis.
Baltimore, MD: Johns Hopkins University Press.
Felman, S. (1985). Writing and Madness: Literature/
Philosophy/Psychoanalysis. Ithaca, NY: Cornell
University Press.
53
Fleischman, S. (1990). Tense and Narrativity: From
Medieval Performance to Modern Fiction. Texas
Linguistics Series. Austin: University of Texas
Press.
Fludernik, M. (1991). ‘‘The Historical Present
Tense Yet Again: Tense Switching and Narrative Dynamics in Oral and Quasi-Oral Storytelling.’’ Text 11(3), 365–98.
Fludernik, M. (1993). The Fictions of Language and
the Languages of Fiction: The Linguistic Representation of Speech and Consciousness. London:
Routledge.
Fludernik, M. (ed.) (1994). Second-Person Narrative.
Special issue, Style 28.3.
Fludernik, M. (1996). Towards a ‘‘Natural’’ Narratology. London/New York: Routledge.
Fludernik, M. (1999). ‘‘The Genderization of Narrative.’’ In J. Pier (ed.), GRAAT 21: Recent
Trends in Narratological Research: Papers from the
Narratology Round Table, ESSE 4, Debrecen,
September 1997 (pp. 153–75). Tours: Publications des Groupes de Recherches AngloAméricaines de l’Université François Rabelais
de Tours.
Fludernik, M. (2000a). ‘‘Beyond Structuralism in
Narratology: Recent Developments and New
Horizons in Narrative Theory.’’ Anglistik
11(1), 83–96.
Fludernik, M. (2000b). ‘‘The Hybridity of Discourses about Hybridity: Kipling’s ‘Naboth’
as an Allegory of Postcolonial Discourse.’’ In
T. Steffen (ed.), Crossover: Cultural Hybridity in
Ethnicity, Gender, Ethics (pp. 151–68). Tübingen: Stauffenberg.
Fludernik, M. (2000c). ‘‘Genres, Text Types,
or Discourse Modes – Narrative Modalities
and Generic Categorization.’’ Style 34(2),
274–92.
Fludernik, M. (2003a). ‘‘The Diachronization of
Narratology.’’ Narrative 11(3), 331–48.
Fludernik, M. (2003b). Review of A. Nünning
& V. Nünning (eds.) (2000) Multiperspektivisches Erzählen: Zur Theorie und Geschichte
der Perspektivenstruktur im englischen Roman des
18. bis 20. Jahrhunderts. GRM 53, 262–7.
Foley, B. (1986). Telling the Truth: The Theory
and Practice of Documentary Fiction. Ithaca, NY:
Cornell University Press.
Friedemann, K. ([1910] 1965). Die Rolle des
Erzählers in der Epik. Darmstadt: Wissenschaftliche Buchgesellschaft.
54
Monika Fludernik
Friedman, N. (1955). ‘‘Point of View in Fiction.
The Development of a Critical Concept.’’
PMLA 70, 1160–84.
Füger, W. (1978). ‘‘Das Nichtwissen des Erzählers
in Fieldings Joseph Andrews: Baustein zu einer
Theorie negierten Wissens in der Fiktion.’’ Poetica 10, 188–216.
Genette, G. (1980). Narrative Discourse: An Essay
in Method. Ithaca, NY: Cornell University Press.
Genette, G. (1988). Narrative Discourse Revisited.
Ithaca, NY: Cornell University Press.
Genette, G. (1991). Fiction et diction. Paris: Seuil.
Gibson, A. (1996). Towards a Postmodern Theory of
Narrative. Edinburgh: Edinburgh University
Press.
Gibson, A. (1999). Postmodernity, Ethics, and the
Novel. London: Routledge.
Hamburger, K. ([1957] 1993). The Logic of Literature, 2nd revised edn, trans. M. J. Rose.
Bloomington, IN: Indiana University Press.
Hamon, P. (1972). ‘‘Pour un statut sémiologique
du personnage.’’ Littérature 6, 86–110.
Harweg, R. (1975). ‘‘Perfekt und Präteritum im
gesprochenen Neuhochdeutsch. Zugleich ein
Beitrag zur Theorie des nichtliterarischen
Erzählens.’’ Orbis 24(1), 130–83.
Herman, D. (1995). Universal Grammar and
Narrative Form. Durham, NC: Duke University
Press.
Herman, D. (1997). ‘‘Scripts, Sequences, and Stories: Elements of a Postclassical Narratology.’’
PMLA 112(5), 1046–59.
Herman, D. (ed.) (1999). Narratologies: New
Perspectives on Narrative Analysis. Theory and
Interpretation of Narrative Series. Columbus:
Ohio State University Press.
Herman, D. (2002). Story Logic: Problems and
Possibilities of Narrative. Lincoln, NE: University of Nebraska Press.
Herman, D. (ed.) (2003). Narrative Theory and the
Cognitive Sciences. Stanford, CA: Publications of
the Center for the Study of Language and Information.
Herman, D., Jahn, M. and Ryan, M.-L. (forthcoming). The Routledge Encyclopedia of Narrative Theory. London and New York: Routledge.
Hoesterey, Ingeborg (1992). ‘‘Introduction.’’ In
A. Fehn, I. Hoesterey, and M. Tatar (eds.),
Neverending Stories: Towards a Critical Narratology (pp. 3–14). Princeton, NJ: Princeton University Press.
Hollowell, J. (1977). Fact & Fiction: The New
Journalism and the Nonfiction Novel. Chapel
Hill, NC: University of North Carolina Press.
Hrushovski, B. (1984). ‘‘Fictionality and Fields of
Reference: Remarks on a Theoretical Framework.’’ Poetics Today 5(2), 227–51.
Hyde, A. (1997). Bodies of Law. Princeton, NJ:
Princeton University Press.
Ireland, K. (2001). The Sequential Dynamics of Narrative: Energies at the Margins of Fiction. Madison,
WI: Fairleigh Dickinson University Press/London: Associated University Presses.
Iser, W. ([1972] 1990). The Implied Reader: Patterns
of Communication in Prose Fiction from Bunyan to
Beckett. Baltimore, MD: Johns Hopkins University Press.
Jahn, M. (1996). ‘‘Windows of Focalization:
Deconstructing and Reconstructing a Narratological Concept.’’ Style 30(2), 241–67.
Jahn, M. (1999). ‘‘More Aspects of Focalization:
Refinements and Applications.’’ In J. Pier (ed.),
GRAAT 21: Recent Trends in Narratological
Research: Papers from the Narratology Round
Table, ESSE 4, Debrecen, September 1997
(pp. 85–110). Tours: Publications des Groupes
de Recherches Anglo-Américaines de l’Université François Rabelais de Tours.
Jahn, M. (1997). ‘‘Frames, Preferences, and the
Reading of Third-Person Narratives: Towards
a Cognitive Narratology.’’ Poetics Today 18(4),
441–68.
Jahn, M. (2001). ‘‘Narrative Voice and Agency in
Drama: Aspects of a Narratology of Drama.’’
New Literary History 32(3), 659–79.
Jahn, M. (2003). ‘‘Narratology: A Guide to
the Theory of Narrative’’ <www.uni-koeln.de/
ame02/pppn.htm>.
Jahn, M. and Nünning, A. (1994). ‘‘Forum:
A Survey of Narratological Models.’’ Literatur
in Wissenschaft und Unterricht 27(4), 283–303.
James, H. (1907–17). The Novels and Tales of Henry
James, Vol. I – Vol. 26. New York: Scribner.
James, H. ([1934] 1953) The Art of the Novel,
introduction R. P. Blackmur. New York:
Scribner.
Kayman, M. A. (2002). ‘‘Law-and-Literature:
Questions of Jurisdiction.’’ REAL: The Yearbook
of Research in English & American Literature 18,
1–20.
Kayser, W. (1957). ‘‘Wer erzählt den Roman?’’
Neue Rundschau 68, 444–59.
From Structuralism to the Present
Korte, B. (1997). Body Language in Literature. Toronto and Buffalo, NY: University of Toronto
Press.
Labov, W. (1972). Language in the Inner City: Studies in the Black English Vernacular. Philadelphia:
University of Pennsylvania Press.
Lämmert, E. (1955). Bauformen des Erzählens.
Stuttgart: Metzler.
Lanser, S. S. (1981). The Narrative Act: Point of
View in Prose Fiction. Princeton, NJ: Princeton
University Press.
Lanser, S. S. (1986). ‘‘Toward a Feminist Narratology.’’ Style 20(3), 341–363.
Lanser, S. S. (1988). ‘‘Shifting the Paradigm:
Feminism and Narratology.’’ Style 22(1),
52–60.
Lanser, S. S. (1992). Fictions of Authority: Women
Writers and Narrative Voice. Ithaca, NY: Cornell
University Press.
Lanser, S. S. (1995). ‘‘Sexing the Narrative: Propriety, Desire, and the Engendering of Narratology.’’ Narrative 3(1), 85–94.
Lanser, S .S. (1999). ‘‘Sexing Narratology: Toward
a Gendered Poetics of Narrative Voice.’’ In
W. Grünzweig and A. Solbach (eds.), Grenzüberschreitungen: Narratologie im Kontext/Transcending Boundaries: Narratology in Context
(pp. 167–84). Tübingen: Narr.
Lotman, J. M. (1977). The Structure of the Artistic
Text. Michigan Slavic Contributions. Ann
Arbor: University of Michigan Slavic Department.
Lubbock, P. (1921). The Craft of Fiction. London:
Jonathan Cape.
McHale, B. (1978). ‘‘Free Indirect Discourse:
A Survey of Recent Accounts.’’ Poetics and Theory of Literature 3, 249–87.
McHale, B. (1983). ‘‘Unspeakable Sentences, Unnatural Acts: Linguistics and Poetics Revisited.’’ Poetics Today 4(1), 17–45.
McHale, B. (1992). Constructing Postmodernism.
London: Routledge.
McHale, B. ([1987] 1996). Postmodernist Fiction.
London: Routledge.
McHale, B. (2001). ‘‘Weak Narrativity: The Case
of Avant-Garde Narrative Poetry.’’ Narrative
9(2), 161–8.
Margolin, U. (1990). ‘‘The What, the When, and
the How of Being a Character in Literary Narrative.’’ Style 24(3), 105–20.
55
Margolin, U. (1995). ‘‘Characters in Literary
Narrative: Representation and Signification.’’
Semiotica 106(3–4), 373–92.
Margolin, U. (1996). ‘‘Characters and Their Versions.’’ In C.-A. Mihailescu and W. Hamarneh
(eds.), Fiction Updated: Theories of Fictionality,
Narratology, and Poetics (pp. 113–32). Toronto:
University of Toronto Press.
Margolin, U. (1999a). ‘‘Of What is Past, is Passing, or to Come: Temporality, Aspectuality,
Modality, and the Nature of Literary Narrative.’’ In D. Herman (ed.), Narratologies: New
Perspectives on Narrative Analysis (pp. 142–66).
Theory and Interpretation of Narrative Series.
Columbus: Ohio State University Press.
Margolin, U. (1999b). ‘‘Story Modalised, or the
Grammar of Virtuality.’’ In J. Pier (ed.) GRAAT
21: Recent Trends in Narratological Research:
Papers from the Narratology Round Table, ESSE
4, Debrecen, September 1997 (pp. 49–62).
Tours: Publications des Groupes de Recherches
Anglo-Américaines de l’Université François
Rabelais de Tours.
Margolin, U. (2000). ‘‘Telling in the Plural: From
Grammar to Ideology.’’ Poetics Today 21(3),
591–618.
Martinez, M. and Scheffel, M. (1999). Einführung
in die Erzähltheorie. Munich: C. H. Beck.
Metz, C. (1971). Langage et cinéma. Paris: Larousse.
Mezei, K. (ed.) (1996). Ambiguous Discourse. Feminist Narratology and British Women Writers.
Chapel Hill, NC: University of North Carolina
Press.
Müller, G. ([1948] 1968). ‘‘Erzählzeit und erzählte
Zeit.’’ In Morphologische Poetik. Darmstadt: Wissenschaftliche Buchgesellschaft.
Nelles, W. (1990). ‘‘Getting Focalization into
Focus.’’ Poetics Today 11(2), 365–82.
Niederhoff, B. (2001). ‘‘Fokalisation und Perspektive: Ein Plädoyer für friedliche Koexistenz.’’
Poetica 33, 1–22.
Nieragden, G. (2002). ‘‘Focalization and Narration: Theoretical and Terminological Refinements.’’ Poetics Today 23(4), 685–97.
Norrick, N. (2000). Conversational Narrative. Storytelling in Everyday Talk. Amsterdam: John
Benjamins.
Nünning, A. (1989). Grundzüge eines kommunikationstheoretischen Modells der erzählerischen Vermittlung: Die Funktionen der Erzählinstanz in den
56
Monika Fludernik
Romanen George Eliots. Horizonte, 2. Trier: Wissenschaftlicher Verlag.
Nünning, A. (1997). ‘‘ ‘But Why Will You Say
That I Am Mad?’: On the Theory, History, and
Signals of Unreliable Narration in British Fiction.’’ Arbeiten aus Anglistik und Amerikanistik
22(1), 83–105.
Nünning, A. (1999a). ‘‘Reconceptualizing the
Theory and Generic Scope of Unreliable
Narration.’’ In J. Pier (ed.), GRAAT 21: Recent
Trends in Narratological Research: Papers from the
Narratology Round Table, ESSE 4, Debrecen,
September 1997 (pp. 63–84). Tours: Publications des Groupes de Recherches AngloAméricaines de l’Université François Rabelais
de Tours.
Nünning, A. (1999b). ‘‘Unreliable, Compared to
What?: Towards a Cognitive Theory of Unreliable Narration: Prolegomena and Hypotheses.’’
In W. Grünzweig and A. Solbach (eds.), Grenzüberschreitungen: Narratologie im Kontext/Transcending Boundaries: Narratology in Context (pp.
53–74). Tübingen: Narr.
Nünning, A. (2000). ‘‘Towards a Cultural and
Historical Narratology: A Survey of Diachronic
Approaches, Concepts, and Research Projects.’’
Anglistentag 1999 Mainz: Proceedings, ed. Bernhard Reitz and Sigrid Rieuwerts (pp. 345–73).
Trier: WVT.
Nünning, A. (2003). ‘‘Narratology or Narratologies?: Taking Stock of Recent Developments,
Critique and Modest Proposals for Future Usages of the Term.’’ In T. Kindt and H.-H.
Müller (eds.), What is Narratology? Questions
and Answers Regarding the Status of a Theory
(pp. 239–75). Berlin: de Gruyter.
Nünning, A. and Nünning, V. (eds.) (2000). Multiperspektivisches Erzählen: Zur Theorie und
Geschichte der Perspektivenstruktur im englischen
Roman des 18. bis 20. Jahrhunderts. Trier: WVT.
Nünning, A. and Nünning, V. (eds.) (2002).
Erzähltheorie transgenerisch, intermedial, interdisziplinär. WVT Handbücher zum literaturwissenschaftlichen Studium, 5. Trier: WVT.
Nünning, A. and Nünning, V. (eds.) (2004).
Erzähltextanalyse und Gender Studies. Stuttgart:
Metzler.
Ochs, E. and Capps, L. (2001). Living Narrative:
Creating Lives in Everyday Storytelling. Cambridge, MA: Harvard University Press.
Page, R. (2003). ‘‘Feminist Narratology? Literary
and Linguistic Perspectives on Gender and Narrativity.’’ Language and Literature 12, 43–56.
Pavel, T. G. (1973). ‘‘Some Remarks on Narrative
Grammars.’’ Poetics 8, 5–30.
Pavel, T. G. (1986). Fictional Worlds. Cambridge,
MA: Harvard University Press.
Peer, W. van and Chatman, S. (eds.) (2001). New
Perspectives on Narrative Perspective. Albany, NY:
State University of New York Press.
Petöfi, J. (ed.) (1979). Text vs Sentence: Basic Questions of Text Linguistics: First Part. Papiere zur
Textlinguistik, 20.1. Hamburg: Buske.
Petöfi, J. and Rieser, H. (1973). Studies in Text
Grammar. Dordrecht: Reidel.
Pfister, M. ([1977] 1991). The Theory and Analysis
of Drama, trans. J. Halliday. Cambridge, UK:
Cambridge University Press.
Phelan, J. (ed.) (1989a). Reading Narrative: Form,
Ethics, Ideology. Columbus: Ohio State University Press.
Phelan, J. (1989b). Reading People, Reading Plots:
Character, Progression, and the Interpretation of
Narrative. Chicago: University of Chicago Press.
Phelan, J. (1996). Narrative as Rhetoric: Technique,
Audiences, Ethics, Ideology. Columbus: Ohio State
University Press.
Phelan, J. (2001). ‘‘Why Narrators Can Be Focalizers – And Why It Matters.’’ In W. van Peer
and S. Chatman (eds.), New Perspectives on Narrative Perspective (pp. 51–64). Albany, NY: State
University of New York Press.
Phelan, J. and Rabinowitz, P. J. (eds.) (1994).
Understanding Narrative. Columbus: Ohio State
University Press.
Pier, J. and Schaeffer, J.-M. (eds.) (forthcoming).
Actes du colloque ‘‘La Métalepse, Aujourd’hui.’’
Paris: Editions du CNRS.
Pratt, M. L. (1977). Toward a Speech Act Theory of
Literary Discourse. Bloomfield, IN: Indiana University Press.
Pratt, M. L. (1992). Imperial Eyes: Travel Writing
and Transculturation. London: Routledge.
Prince, G. (1982). Narratology: The Form and Functioning of Narrative. Berlin: Mouton.
Prince, G. (1987). A Dictionary of Narratology.
Lincoln and London: University of Nebraska
Press.
Prince, G. (1995). ‘‘On Narratology: Criteria, Corpus, Context.’’ Narrative 3(1), 73–84.
From Structuralism to the Present
Prince, G. (1996). ‘‘Narratology, Narratological
Criticism, and Gender.’’ In C.-A. Mihailescu
and W. Hamarneh (eds.), Fiction Updated: Theories of Fictionality, Narratology, and Poetics
(pp. 159–64). Toronto: University of Toronto
Press.
Prince, G. (1999). ‘‘Revisiting Narrativity.’’ In
W. Grünzweig and A. Solbach (eds.), Grenzüberschreitungen: Narratologie im Kontext/Transcending Boundaries: Narratology in Context
(pp. 43–51). Tübingen: Narr.
Prince, G. (2001). ‘‘A Point of View on Point of
View or Refocusing Focalization.’’ In W. van
Peer and S. Chatman (eds.), New Perspectives on
Narrative Perspective (pp. 43–50). Albany, NY:
State University of New York Press.
Quasthoff, U. M. (1980). Erzählen in Gesprächen:
Linguistische Untersuchungen zu Strukturen und
Funktionen am Beispiel einer Kommunikationsform
des Alltags. Tübingen: Narr.
Quasthoff, U. M. (1999). ‘‘Mündliches Erzählen
und sozialer Kontext: Narrative Interaktionsmuster in Institutionen.’’ In W. Grünzweig and A. Solbach (eds.), Grenzüber
schreitungen: Narratologie im Kontext/Transcending
Boundaries: Narratology in Context (pp. 127–46).
Tübingen: Narr.
Rabinowitz, P. J. and Smith, M. W. (1998). Authorizing Readers: Resistance and Respect in the
Teaching of Literature. New York: Teachers College Press.
Richardson, B. (1987). ‘‘ ‘Time is Out of Joint.’:
Narrative Models and the Temporality of the
Drama.’’ Poetics Today 8(2), 299–310.
Richardson, B. (1988). ‘‘Point of View in Drama:
Diegetic Monologue, Unreliable Narrators, and
the Author’s Voice on Stage.’’ Comparative
Drama 22(3), 193–214.
Richardson, B. (2000). ‘‘Recent Concepts of Narrative and the Narratives of Narrative Theory.’’
Style 34(2), 168–75.
Richardson, B. (2001a). ‘‘Construing Conrad’s The
Secret Sharer: Suppressed Narratives, Subaltern
Reception, and the Act of Interpretation.’’ Studies in the Novel 33(3), 306–21.
Richardson, B. (2001b). ‘‘Voice and Narration in
Postmodern Drama.’’ New Literary History
32(3), 681–694.
Ricoeur, P. (1984–88). Time and Narrative,
Vol. I-III, trans. K. McLaughlin and D. Pellauer.
Chicago: University of Chicago Press.
57
Rimmon-Kenan, S. (2001). ‘‘Narrative as Paradigm in the Interface between Literature and
Psychoanalysis.’’ Lecture given at the IAUPE
conference in Bamberg, July 2001.
Rimmon-Kenan, S. ([1983] 2002). Narrative Fiction: Contemporary Poetics. London: Routledge.
Roe, E. (1994). Narrative Policy Analysis: Theory
and Practice. Durham, NC: Duke University
Press.
Romano, C. (2002). ‘‘Is the Rise of ‘Narratology’
the Same Old Story?’’ The Chronicle of Higher
Education June 28, B12.
Roof, J. (1996). Come As You Are: Sexuality and
Narrative. New York: Columbia University
Press.
Ryan, M.-L. (1991). Possible Worlds, Artificial Intelligence, and Narrative Theory. Bloomington:
Indiana University Press.
Ryan, M.-L. (1992). ‘‘The Modes of Narrativity
and Their Visual Metaphors.’’ Style 26(3),
368–87.
Ryan, M.-L. (ed.) (1999). Cyberspace Textuality.
Computer Technology and Literary Theory. Indianapolis: Indiana University Press.
Ryan, M.-L. (2001). Narrative as Virtual Reality:
Immersion and Interactivity in Literature and the
Electronic Media. Baltimore: Johns Hopkins
University Press.
Ryan, M.-L. (ed.) (2004). Narrative Across Media:
The Languages of Storytelling. Lincoln: University
of Nebraska Press.
Schafer, R. (1992). Retelling a Life: Narration and
Dialogue in Psychoanalysis. New York: Basic
Books.
Schellinger, P. (ed.) (1998). Encyclopedia of the
Novel, 2 vols. Chicago: Fitzroy Dearborn.
Schneider, R. (2001). ‘‘Toward a Cognitive Theory
of Literary Character: The Dynamics of MentalModel Construction.’’ Style 35(4), 607–40.
Sell, R. (2000). Literature as Communication: The
Foundations of Mediating Criticism. Amsterdam
and Philadelphia: John Benjamins.
Semino, E. and Culpepper, J. (eds.) (2002). Cognitive Stylistics: Language and Cognition in Text
Analysis. Amsterdam and Philadelphia: John
Benjamins.
Shen, D. (2001). ‘‘Breaking Conventional Barriers:
Transgressions of Modes of Focalization.’’ In
W. van Peer and S. Chatman (eds.), New Perspectives on Narrative Perspective (pp. 159–72). Albany: State University of New York Press.
58
Monika Fludernik
Sinfield, A. (1992). Faultlines: Cultural Materialism and the Politics of Dissident Reading. Oxford:
Clarendon Press.
Spurr, D. (1993). The Rhetoric of Empire: Colonial
Discourse in Journalism, Travel Writing and Imperial Administration. Durham, NC: Duke University Press.
Stanzel, F. K. ([1955] 1971). Narrative Situations
in the Novel: Tom Jones, Moby Dick, The
Ambassadors, Ulysses. Trans. J. P. Pusack.
Bloomington: Indiana University Press.
Stanzel, F. K. (1978) ‘‘Towards a ‘Grammar of
Fiction.’ ’’ Novel 11, 247–64.
Stanzel, F. K. (1981). ‘‘Teller-Characters and Reflector Characters in Narrative Theory.’’ Poetics
Today 2(2), 5–15.
Stanzel, F. K. ([1979] 1984). A Theory of Narrative,
trans. C. Goedsche. Cambridge, UK: Cambridge University Press.
Stempel, W.-D. (1986). ‘‘Everyday Narrative as
‘Prototype.’ ’’ Poetics 15, 203–16.
Sternberg, M. (1982). ‘‘Proteus in QuotationLand: Mimesis and the Forms of Reported
Discourse.’’ Poetics Today 3(2), 107–56.
Sternberg, M. (1983). ‘‘Mimesis and Motivation:
The Two Faces of Fictional Coherence.’’ In
J. Strelka (ed.), Literary Criticism and Philosophy.
Yearbook of Comparative Criticism, 10 (pp. 144–88).
University Park, PA: Pennsylvania State University Press.
Sternberg, M. (1987). The Poetics of Biblical
Narrative: Ideological Literature and the Drama
of Reading. Bloomington: Indiana University
Press.
Sternberg, M. ([1978] 1993). Expositional Modes
and Temporal Ordering in Fiction. Bloomington:
Indiana University Press.
Sternberg, M. (2001). ‘‘How Narrativity Makes a
Difference.’’ Narrative 9, 115–22.
Stockwell, P. (2002). Cognitive Poetics: An Introduction. London: Routledge.
Sturgess, P. J. M. (1992). Narrativity: Theory and
Practice. Oxford: Clarendon Press.
Tannen, D. (ed.) (1982). Analyzing Discourse: Text
and Talk. Georgetown University Round Table
on Languages and Linguistics 1981. Washington, DC: Georgetown University Press.
Tannen, D. (1984). Conversational Style: Analyzing
Talk Among Friends. Norwood, NJ: Ablex.
Thomas, B. (1987). Cross-Examinations of Law and
Literature: Cooper, Hawthorne, Stowe and Melville.
Cambridge, UK: Cambridge University Press.
Thomas, B. (ed.) (2001). Law and Literature.
Tübingen: Narr.
Vitoux, P. (1982). ‘‘Le jeu de la focalisation.’’ Poétique 51, 358–68.
Warhol, R. (1989). Gendered Interventions:
Narrative Discourse in the Victorian Novel. New
Brunswick, NJ: Rutgers University Press.
Warhol, R. (2003). Having a Good Cry: Effeminate
Feelings and Pop-Culture Forms. Columbus: Ohio
State University Press.
Weinrich, H. ([1964] 1985). Tempus: Besprochene
und erzählte Welt. Sprache und Literatur, 16.
Stuttgart: Kohlhammer, 4th edn. based on
2nd revised edition 1971.
Weisberg, R. H. (1984). The Failure of the Word:
The Protagonist as Lawyer in Modern Fiction. New
Haven, CT: Yale University Press.
Weisberg, R. H. (1992). Poethics, and Other Strategies of Law and Literature. New York: Columbia
University Press.
White, H. (1973). Metahistory: The Historical
Imagination in Nineteenth-Century Europe. Baltimore, MD: Johns Hopkins University Press.
White, H. (1980). ‘‘The Value of Narrativity in
the Representation of Reality.’’ Critical Inquiry
7, 5–27.
White, H. (1987). The Content of the Form:
Narrative Discourse and Historical Representation.
Baltimore: Johns Hopkins University Press.
Wolf, W. (1993). Ästhetische Illusion und Illusionsdurchbrechung in der Erzählkunst. Theorie und
Geschichte mit Schwerpunkt auf englischem illusionsstörenden Erzählen. Tübingen: Niemeyer.
Wolf, W. (1998). ‘‘Aesthetic Illusion in Lyric
Poetry?’’ Poetica 30, 251–89.
Wolf, W. (1999). The Musicalization of Fiction:
A Study in the Theory and History of Intermediality. IFAVL, 35. Amsterdam: Rodopi.
Wolf, W. (2002). ‘‘Das Problem der Narrativität
in Literatur, bildender Kunst und Musik: Ein
Beitrag zu einer intermedialen Erzähltheorie.’’
In A. Nünning and V. Nünning (eds.),
Erzähltheorie transgenerisch, intermedial, interdisziplinär (pp. 23–104). Trier: WVT.
Wolf, W. (2003). ‘‘Narrative and Narrativity:
A Narratological Reconceptualization and its
From Structuralism to the Present
Applicability to the Visual Arts.’’ Word & Image
19(3), 180–97.
Wolf, W. (2004) ‘‘Aesthetic Illusion as an Effect of
Fiction.’’ In M. Fludernik and U. Margolin (eds.),
German Narratology, Special issue. Style 38 (3–4).
Yacobi, T. (1981). ‘‘Fictional Reliability as a
Communicative Problem.’’ Poetics Today 2(2),
113–36.
59
Yacobi, T. (1987). ‘‘Narrative and Normative Pattern: On Interpreting Fiction.’’ Journal of Literary Studies 3, 18–41.
Yacobi, T. (2000). ‘‘Interart Narrative: (Un)Reliability and Ekphrasis.’’ Poetics Today 21(4),
711–49.
3
Ghosts and Monsters: On the
(Im)Possibility of Narrating the
History of Narrative Theory
Brian McHale
A Ghost in the Machine
A specter is haunting narrative theory – or at least, it haunts David Herman’s and
Monika Fludernik’s histories of narrative theory. That specter is none other than
Mikhail Bakhtin. The author of (among other things) two landmark works of
narrative theory, and implicated somehow or other in the production of a third,1
Bakhtin (1895–1975) is certainly the most ubiquitous narrative theorist of the last
quarter of the twentieth century, and arguably one of the most influential. He is the
one narrative theorist about whom every graduate literature student is certain to know
something, even if he or she knows nothing else about narrative theory. Nevertheless,
Bakhtin is conspicuous by his near-absence from both Herman’s and Fludernik’s
histories of narrative theory – complete absence in the case of Fludernik, scant
mention in the case of Herman. How did everyone’s favorite narrative theorist all
but vanish from history – or at least, from these histories?
Perhaps the division of labor is to blame. Herman covers early developments in
narrative theory through structuralism, and Fludernik developments since structuralism, but Bakhtin, due to his long mid-career silence and his belated reception in the
West (about which I’ll say more in a moment), straddles that divide. A contemporary of
the Russian Formalists as well as, through the vagaries of his reception, of 1980s and
1990s contextual narratology, Bakhtin belongs in both histories. Perhaps Herman and
Fludernik each thought the other would take responsibility for him. But if so, this would
be a trivial oversight, and easily remedied. What are editors for, if not to catch such slips?
I don’t believe Bakhtin’s absence is a trivial oversight. Rather, it is symptomatic of
an important tension within the history of narrative theory, and ultimately within
narrative theory itself. Bakhtin, it seems to me, has slipped through the cracks
between two approaches to narrating the history of narrative theory.2 Herman and
Ghosts and Monsters
61
Fludernik vacillate between these two approaches, in either of which Bakhtin might
figure, but in each of which he would figure differently. His strange disappearance is
largely attributable to this vacillation, and its consequence is to throw the entire
enterprise of history and theory, of historicizing theory, into sharp relief.
What kind of history might a history of narrative theory be? As I see it, one might
approach such a history from one of two angles. One might begin from the system of
ideas that has come to comprise narrative theory, identifying the sources of its basic
concepts and terms, tracing their refinements and complications at the hands of later
theorists, and juxtaposing different states of the system. This approach is akin to the
‘‘history of ideas’’ in the classic mode of A. O. Lovejoy, the historiographer of
‘‘romantic,’’ ‘‘primitive,’’ ‘‘the Great Chain of Being,’’ and other ‘‘unit-ideas’’ (his
own term; see Lovejoy [1936] 1957 and [1938] 1948). Herman practices this kind
of history when he specifies the ‘‘research foci’’ that Wellek and Warren share both
with the Russian Formalist tradition and with subsequent narrative theory (this
volume, pp. 21), while Fludernik practices the same sort of idea-oriented history
when, for instance, she traces the refinements introduced by Bal into Genette’s
original notion of focalization (this volume, pp. 40–1).
Alternatively, one might approach the history of narrative theory from the perspective of the latter’s institutional existence. I understand ‘‘institutional’’ here in an extended
but, I hope, still recognizable sense, including not only academic institutions, research
collaborations, disciplinary circuits, conditions of publication and dissemination, and
professional career trajectories, but also informal networks and even personal relationships of intellectual affinity, mentorship, role modeling, and so forth – in short, the
whole social life of ideas. Institutional history, in this extended sense, preoccupies itself
not just with unit-ideas and their pedigrees, but with who knew whom, who taught
whom, who published what and when (and who was prevented from publishing), who
read what, when, and under what circumstances, who brought the word from Moscow
to Prague and what happened to it when it got there, and so on. This is the kind of
history that Herman characterizes as ‘‘seek[ing] to uncover forgotten interconnections;
reestablish obscured or unacknowledged lines of descent: expose relationships between
institutions, belief-systems, discourses, or modes of analysis that might otherwise be
taken to be wholly distinct and unrelated’’ (this volume, p. 20). Following Nietzsche
and Foucault, he calls this mode ‘‘genealogy.’’3
From the perspective of the history of ideas, Bakhtin’s place seems to be relatively
secure. We might take as representative the forthcoming Routledge Encyclopedia of
Narrative Theory, which aspires to serve as a sort of roadmap of the system of concepts
in narrative theory. Here Bakhtin’s name appears in connection with a number of the
key concepts that he contributed, including ‘‘dialogism,’’ ‘‘dual-voice hypothesis,’’
‘‘heteroglossia,’’ and no doubt several others. In the perspective of institutional
history, however, the picture looks quite different. Here Bakhtin – or rather the
writings for which the name ‘‘Bakhtin’’ has come metonymically to stand – including,
problematically, those of the other members of the so-called ‘‘Bakhtin Circle,’’ V. N.
Vološinov and P. N. Medvedev, is profoundly displaced.
62
Brian McHale
Initially associated with the intellectual ferment surrounding the Russian Formalists, against whom he polemicized (through his spokesmen or personae, Vološinov and
Medvedev), Bakhtin was consigned to silence and provincial obscurity during the
Stalinist era, the victim, apparently, of guilt by association (his brother Nikolai, a
veteran of the counterrevolutionary White Army, was living in exile in Britain). He
ought to have been a precursor of Wellek’s and Warren’s narrative poetics, derived in
part from the Slavic context, but in fact there is scant evidence that Wellek himself or
his Prague School teachers before him knew much, if anything, about the Bakhtin
Circle. Bakhtin re-emerged in the 1960s, both in his native Russia and among
Western Slavists, who by the early 1970s had made his work widely available in
translation. His integration into Western poetics, however, really only began in the
late 1960s, when he was taken up by the Paris structuralists through the mediation of
Julia Kristeva.4 By the 1980s, Bakhtin was enjoying an extraordinary posthumous
career, especially in the United States, but eventually in Russia as well. The story of
his reception over the past two decades involves serial appropriations and misappropriations, decontextualizations and recontextualizations, as he has migrated from one
institutional context to another, passing ghost-like through the walls separating
paradigms and disciplines, uncannily splitting and multiplying until now we have
a whole host of spectral Bakhtins. There is an American liberal-humanist Bakhtin
(sponsored by Wayne Booth, who wrote the introduction to the 1984 retranslation of
Problems of Dostoevsky’s Poetics), a post-structuralist Bakhtin, a materialist Bakhtin, a
feminist Bakhtin, an African-Americanist Bakhtin (where ‘‘double-voiced’’ becomes
conflated with ‘‘double consciousness’’), a postcolonial Bakhtin, a post-Soviet religious
Bakhtin, and no doubt many others.5
The conspicuous absence (or near-absence) of Bakhtin from both Herman’s and
Fludernik’s histories is mainly attributable to their vacillation between the history-ofideas mode and the genealogical mode, between accounting for the system of narrative
theory and narrating its institutional fortunes. They apparently felt obliged to try to
do both, perhaps in a spirit of nonpartisanship, conscious that, if they opted for one
mode to the exclusion of the other, they would be taking sides in an internecine
division that runs right through narrative theory (as we shall see in the next section).
They also apparently felt that they could reconcile the two modes, which (as we’ll also
see in the next section) is harder to do than one might imagine – perhaps impossible.
Instead of reconciling them, they only mixed them up, producing histories that are
neither all one thing nor all the other, nor even some synthesis of the two, but a little
of this and a little of that, a little history of ideas mixed in with a little genealogy.
Bakhtin, in particular, is placed at a disadvantage by this mixed approach. As we
have seen, his place seems secure as far as the history of ideas is concerned, but
anything but secure in an institutional perspective. Institutionally, Bakhtin is a
nomad, a sort of wandering Jew or hungry ghost, doomed restlessly to crisscross the
intellectual world without ever settling in any particular niche. A contemporary and
antagonist of the Russian Formalists, he ought to figure in histories of Formalism, but
rarely does. Instead, he appears as a fellow-traveler of a range of contemporary
Ghosts and Monsters
63
contextual narratologies – materialist, African-American, feminist, postcolonial, and
so on – with whom his institutional affiliation is not only entirely posthumous, but
largely dependent on various forms of misprision and intellectual press-ganging. One
readily sees how a historian of narrative theory, confronted with the vagaries and
incoherence of Bakhtin’s institutional career, might prefer to treat him instead as a
figure in the history of ideas, where his position seems so much more stable: Bakhtin,
father of ‘‘dialogism.’’
Unfortunately, the same nomadism that makes it hard to integrate him into the
institutional history of narrative theory also tends to undermine his place in the
history of ideas. The very facility with which ‘‘dialogue,’’ ‘‘dialogism,’’ ‘‘heteroglossia,’’
and so on have been taken up across such a wide range of practices, both inside the
disciplinary sphere of narrative theory and well beyond it, is apt to arouse one’s
suspicions. Perhaps these ‘‘traveling concepts’’ don’t really belong to narrative theory
after all; perhaps, after all, Bakhtin is not really a narrative theorist, and ought not to
be covered in a history of the discipline. And so, institutionally unplaceable, disciplinarily suspect, neither here or there, Bakhtin slips through the cracks.
Surely there is something uncanny, or maybe something karmic, in Bakhtin’s
historical displacement. Bakhtin charged the Formalists with having excluded the
contingencies of social experience and social context from their account of how
language operates in verbal art – with having, in effect, excluded the historical
existence of language, with having dehistoricized it.6 Is it some kind of poetic justice,
then, or maybe karmic justice, that Bakhtin should himself undergo dehistoricization,
historical decontextualization? Bakhtin, contemporary of the Formalists, has become
everybody’s contemporary, unmoored in time and space, available for assimilation (or so
it sometimes seems) to anyone’s system of ideas. Of course, it is precisely his insistence
on historicizing language, on restoring it to its place in a historically contingent social
realm, that has made Bakhtin so congenial to so many varieties of historicist and
contextualist theory in our own time. It is this, more than anything else, that has
made him such an attractive candidate for serial reappropriation and recontextualization. Cut loose to drift through history precisely because he sided with historicism
against dehistoricized form, Bakhtin falls prey to so many complex historical ironies
that one cannot help but regard his situation as uncanny. Remember, too, that
Bakhtin’s attacks on the Formalists were conducted under the cover of others’
names, Vološinov’s and Medvedev’s; that is to say, they were (or so some authorities
maintain) ghost-written.
Structure Versus History
The Bakhtin case serves here, as I intimated earlier, as a kind of parable of the tension
in these histories of narrative theory between, precisely, history and narrative theory –
between the obligation incumbent upon these historians to reconstruct the successive
states of the system of narratological concepts, and their obligation to place that
64
Brian McHale
system and those concepts in the context of institutional history, to historicize them.
This tension runs right through these histories, indeed (as I shall argue below) right
through our discipline, where it is replicated at every level, under various terms. It is
reflected in the current tension within the field between, on the one hand, its
‘‘classically’’ structuralist narratological tendencies and its contextualist tendencies (see
Chatman 1990), broadly construed to include feminist narratology and other historicist varieties. Of all the terms that one might use to capture this tension, the most
canonical of all are the ones with which I have titled the present section: structure
versus history.7
It would be nice to be able to say that the two orientations, historicist and
structuralist, were complementary, that narratology in the structuralist lineage complemented contextualist narrative theory and vice versa; unfortunately, the relationship
between them is not as happy or untroubled as that would imply. Quite the reverse, in
fact: under the big tent of narrative theory, structuralism and historicism jockey for
position, each seeking to outflank or overcome the other, to contain the other, and if
that doesn’t work, then to forget or repress the other – a risky strategy since, as we
know, the repressed is apt to return. The accusation, on the part of historicist narrative
theory, that narratology represses history is a familiar one, of course; this was already
the substance of the Bakhtin Circle’s attack on the Formalists, and later of Jameson’s
critique of Formalism and structuralism alike. But the reverse is also arguably true:
that historicist narrative theory represses its narratological Other.
To illustrate how historicism seeks to outflank and contain narratology, one need
look no further than several of the chapters in the present Companion. In essays by, for
instance, Alison Case (‘‘Gender and History in Narrative Theory’’), Alan Nadel
(‘‘Second Nature, Cinematic Narrative, the Historical Subject, and Russian Ark’’)
and Harry Shaw (‘‘Why Won’t Our Terms Stay Put?’’), narratological categories are
revisited from a historical perspective, and rethought as historically contingent and
variable, rather than as permanent features of a system that applies differently at
different epochs. As part of his period-based narratology, for instance, Shaw proposes
to treat the communications model not as a universal ‘‘given’’ of narratives everywhere, but as malleable and adjustable, varying from period to period. On the other
side, structuralist maneuvers to outflank historicism are readily illustrated from the
writings of the Russian Formalists themselves, who from the very outset were subject
to the pressure of a competing Marxist-historicist mode of explanation, and who
responded by seeking to contain or pre-empt the historical in various ways. One such
maneuver can be seen in Jakobson’s and Tynjanov’s thesis on ‘‘Problems in the Study
of Language and Literature’’ ([1928] 1971), dating from very near the end of the
Formalists’ brief flowering, after Jakobson had already decamped for Prague, in which
the model of the literary text as a systematic relation of parts is scaled up to the level
of literary history itself. Reconceived in systemic terms, history now falls within
the Formalists’ purview, instead of being left ‘‘outside,’’ the province of hostile
historicists.8
Ghosts and Monsters
65
Even more powerfully pre-emptive, however, is the concept of motivation, which
was already in place in some of the earliest versions of Formalist narrative poetics
(Tomashevsky [1925] 1965, Shklovsky [1929] 1990). If, as the Formalists posited, a
work of verbal art comprises a series of devices – devices of estrangement, of
retardation, and impeded form, and so on – then the artist must choose either to
lay his or her devices bare (as Sterne notoriously does in Tristram Shandy) or to produce
alibis for them, masking their arbitrariness behind a scrim of naturalness and
inevitability – to motivate the devices. Thus, to adapt one of Shklovsky’s simpler
examples (Shklovksy 1990: 110), if Sherlock Holmes requires an opposite number, a
chronically misguided detective whose incompetence can be played off against Holmes’s own infallibility, then it doesn’t much matter in strictly formal terms whether
the nincompoop is another private detective or a civil servant. Given Conan Doyle’s
historical moment, however, the obvious choice, the one that most satisfactorily
motivates the device, is the civil servant – Inspector Lestrade, of course – allowing
the author to profile the superiority of the bourgeois individual against a background
of state bungling. Shklovsky suggests that, in a proletarian regime, the role assignments might be reversed – the successful detective might be the civil servant, while
the bungler might be the private eye – but the device, and the necessity for some
motivation of it, would be identical. The concept of motivation thus allows historical
contingency to be drawn into the very fabric of formal structure; the historical is
placed at the service of the device. This opening of structure to historical contingency
through the concept of motivation has a long legacy in subsequent narrative theory.
It reappears, with some variations, in Paris structuralist theory under the names
of vraisemblance (Genette [1968] 2001) and naturalization (Culler [1975] 2002:
153–87), and persists as a central concept of Tel Aviv school narratology down to
the present (e.g., Sternberg 1983).
Seeking to contain its narratological Other, historicism represses narratology, just
as, vice versa, narratology represses history – or so I have claimed. Let me see whether
I can justify my claim, and in particular my use of the term ‘‘repress,’’ which will have
struck some readers, I’m sure, as rhetorical overkill. I can make my case most
efficiently by way of example, so let me briefly revisit a pair of undisputed classics
in our field: Henry Louis Gates’s account of free indirect discourse in Zora Neale
Hurston’s Their Eyes Were Watching God, from his book The Signifying Monkey (1988)
and elsewhere, and Gérard Genette’s poetics of Proustian narrative in Narrative
Discourse ([1972] 1980).
Gates’s work is a landmark in the historicizing of narratological description. He
introduces narratological categories, for the first time ever in any systematic way, into
the discussion of the history of novelistic representations of African-American speech.
Nevertheless, in order to interpret and contextualize Hurston’s handling of free
indirect discourse (FID) Gates must, in a sense, ‘‘forget’’ narratology. He insists on
the uniqueness and novelty of FID in Their Eyes Were Watching God – on the way, for
instance, that narration and FID converge stylistically, almost to the point of identity,
66
Brian McHale
and the way FID is used to represent not only individual characters’ speech
and thought, but the collective speech and thought of a whole community.
There’s no disputing the special semantic charge and the extra load of cultural
meaning that these features bear in Their Eyes. Nevertheless, as formal features they
are hardly as unique or unprecedented as Gates makes them out to be, but rather
belong to the range of possibilities that FID encompasses as a quasi-universal category
of narrative discourse. In making his historicist and contextualist case for Hurston’s
innovations, Gates downplays their basis in the system of narrative discourse,
laying himself open to the counterargument that Hurston’s practice isn’t all that
special or unprecedented after all, that her FID is more ‘‘standard,’’ formally and
functionally, than he is willing to admit.9 Structure trumps history; repressed
narratology returns.
Genette, conversely, undertakes to describe a particular novelist’s narrative practice
– Proust’s – against the backdrop of a general system of narratological categories that
aspires to universal and transhistorical applicability. Or rather, he tries to do two
things at once: on the one hand, to capture the idiosyncrasies of Proust’s poetics,
which can only be grasped relative to some ‘‘norm’’ of narrative practice; on the other
hand, to develop a system of general categories using Proust’s Recherche as the primary
example. Genette performs a tricky balancing act between polar opposites – anomaly
and exemplarity, idiolect and system. In the end, however, it’s clearly the system of
norms that holds the upper hand in his approach.
But what if we upended Genette’s approach, and gave anomaly the upper hand
over norm? There are several perspective from which this might be undertaken. First,
we might simply abandon the appeal to transhistorical norms altogether, and develop
a poetics of Proustian narrative that was frankly one-of-a-kind, elevating Proust’s
anomalies into idiosyncratic norms. Alternatively, we might do what Culler
proposes in his foreword to the English translation of Narrative Discourse (Genette
1980: 12–13) and acknowledge the anomalous character of all literary narrative
relative to norms that are based, after all, on models of reality. Between these two
moves – custom-tailoring to the exact specifications of Proustian narrative, on the one
hand, and wall-to-wall literary anomaly, on the other – there is a third possible
move, the historicist one. This is the sort of move that Randall Stevenson (forthcoming) adumbrates when he notes the close correlation between the areas of Genette’s
most valuable contributions to narrative theory – focalization and relations of time –
and the innovations of modernist fiction, not least of all Proust’s fiction, with
respect to perspective and temporality. In other words, suppose we viewed Proust’s
practice, not as anomalous relative to some transhistorical norm, but rather as
typical of a particular historically determined period style – the period style of high
modernism? From this perspective, the categories that Genette imagined to be
universals actually appear to reflect a particular modernist historical practice, and
his supposed transhistorical system comes to share an unexpected kinship with Shaw’s
period-based narratology. With this move, structure is trumped, and repressed history
returns.
Ghosts and Monsters
67
Frankenstein as Narratologist
Again, it would be nice to be able to say that these two orientations, structuralist and
historicist, were, if not complementary, at least reconcilable. My sense, however, is that
their relationship has always been, and is likely to remain, a conflicted one. It may be
that structure and history are finally irreconcilable; or rather, they may be reconciled,
but only on terms congenial to one of the rival orientations and not to the other. To
practice narratology in the structuralist tradition, it would seem, is to strive to get the
better of the historical, to outflank it or trump it or take it in charge, to become its
horizon; and vice versa, to practice contextualist or historicist narratology is to strive
to outflank or trump structure. To practice one is to repress the other, and, as we have
seen, the repressed Other always seems to return.
At the end of her history of narrative theory since structuralism, as in her own
recent narratological work (e.g., Fludernik 2003), Monika Fludernik promises a new
reconciliation of structure and history, in the form of a historicist cognitive narratology.
She advances strong claims for the potential of tools derived from cognitive science to
rejuvenate historicist approaches to narrative. I am a good deal less sanguine than she
is about the chances for any such reconciliation. As a case in point, I might cite a
special issue of Poetics Today in Spring 2002 dedicated to ‘‘Literature and the Cognitive
Revolution,’’ one section of which bears the promising title, ‘‘Cognitive Historicism.’’
The guest editors claim that the papers in this section ‘‘address the complex interaction of evolved neurocognitive structures and contingent cultural environments
with an eye to specific examples of cultural change.’’ These papers demonstrate,
according to the editors, that ‘‘issues in literary history, far from being occluded by
approaches that recognize the validity of human universals and species-specific
cognitive mechanisms, can be productively reopened in ways that have eluded
criticism that relies on purely constructivist notions of the subject’’ (Richardson and
Steen 2002: 5). I remain unpersuaded. I don’t see how ‘‘neurocognitive structures’’
and ‘‘species-specific cognitive mechanisms’’ that have evolved over tens of millennia,
and that are presumably stable and permanent features of modern-day homo sapiens, are
likely to give us much purchase on change occurring at the historical scale of decades
and centuries; and if they don’t capture change and difference at that time-scale, then
I don’t see how the approaches that evoke such structures and mechanisms can claim
to be ‘‘historicist.’’10
In my view, these essays in ‘‘cognitive historicism’’ neither cognitivize historical
difference nor historicize cognitive structures. They don’t after all reconcile cognitivism and historicism, whatever they (or Fludernik) may promise. What they actually
do do is alternate between cognitivist moments and historicist moments, between
structure and history – first one thing, then the other. This is not the same thing as
reconciliation; it’s more like ships passing in the night.
Moreover, I would argue that what these essays do, we all do, whether we call
ourselves cognitive historicists or something else – plain-vanilla historicists, or
68
Brian McHale
cultural materialists, or contextualists, or diehard old-school narratologists, or feminist narratologists, or whatever. I venture to say that many, if not most, of the
contributors to the present Companion believe not only that structure and history
can be reconciled, but moreover that they themselves have done so in their own essays,
seamlessly integrating the tools and insights of narratology with the contingencies of
history. I suspect that, insofar as they hold this belief, they are probably mistaken.
Whatever we may think we’re doing when we study narrative, we’re probably not
reconciling history and structure. On the other hand, neither are we successfully
pursuing one to the exclusion of the other because, as I’ve tried to demonstrate, the
repressed always returns. What we’re really doing, at best, is alternating between
narratological moments and historicist moments. There is no ‘‘pure’’ practice of
narratology to be found, and no ‘‘pure’’ historicist narrative theory either, but neither
is there any stable synthesis or seamless integration of the two – only a messy
patchwork, a little of this and a little of that, first one thing and then the other.
I recently had the opportunity to examine an unpublished doctoral dissertation by a
young Finnish scholar (Samuli Hägg) who has undertaken to do for Pynchon’s
Gravity’s Rainbow what Genette did for Proust’s Recherche – in other words, to produce
a narratology of Pynchon that at the same time reflects critically on the system of
narratological categories, just as Genette produced a narratology of the Recherche that
also elaborated a general system of narrative discourse. He encountered many of the
same difficulties as Genette, in particular the problem of what to treat as anomalous
in Pynchon’s practice, relative to a systematic narratology, and what in Pynchon’s
practice to try to integrate into that system. This in turn entails confronting the
question of what kind of narratology one would have if anomalies such as those
encountered in Gravity’s Rainbow were treated as integral rather than anomalous,
brought ‘‘inside’’ rather than left ‘‘outside.’’ He concludes that the results could only
be, not some coherent ‘‘narratology of Gravity’s Rainbow,’’ but ‘‘a monstrous narratology crudely stitched together from incompatible organs and lumps of flesh.’’11 He
goes on: ‘‘Disturbingly enough . . . this description seems to fit pretty well any
conception of narratology’’ – or of historicist or contextualist narrative theory, for
that matter.
If this is true, and every narrative theory is monstrous, then a history of narrative
theory, such Herman’s or Fludernik’s, could only be an account of how the monster
got stitched together from a little of this and a little of that, here a little structure and
there a little history – in other words, a retelling of Frankenstein. Either that, or it
would have to be a ghost story.12
NOTES
1 See Bakhtin’s Problems in Dostoevsky’s Politics
([1963] 1984), a revision of a work first published in 1929, and his Dialogic Imagination
([1975] 1981). Bakhtin’s role in Vološinov
([1929] 1973) is controversial: was he its
pseudonymous author? Vološinov’s shadow
Ghosts and Monsters
2
3
4
5
6
7
collaborator? His mentor and muse? Nobody
knows for sure.
Nor is he the only the figure to slip through
the cracks in this way. Alan Nadel suggests
(personal communication) that Kenneth Burke
presents a problem comparable to that of
Bakhtin. This is true, but only up to a point;
Bakhtin’s belated currency and astonishing
ubiquity has no parallel in the Burke case.
The distinction between the history-of-ideas
approach to narrative theory and the institutional or genealogical approach roughly parallels the distinction between traditional history
of science and the sociological or cultural studies approach to ‘‘science studies,’’ as practiced
by Bruno Latour, Evelyn Fox Keller, Timothy
Lenoir, Bruce Clarke and others; see, e.g.,
Morrison’s (2002) review-article.
In a parallel development, the Tel Aviv school
of poetics also ‘‘discovered’’ Bakhtin at roughly
the same time, through the mediation of Benjamin Hrushovski (Harshav). However, the
distinctive Tel Aviv synthesis of narrative
poetics would not become widely known outside Israel until almost a decade later, when
the Tel Aviv theorists began publishing in
English.
On Bakhtin in general, see Clark and Holquist
(1984) and Morson and Emerson (1990). On
his Russian reception in particular, see Emerson (1997); on his American reception, see
Hale (1998: 113–27, 197–220). DeMan suggested that the only type of theorist who
might not find any use for Bakhtin would be
one like himself, one ‘‘concerned with tropological displacements of logic’’ (1983: 104) –
in other words, a deconstructionist. But he
might have been wrong about that; others
have found it possible to reconcile Bakhtinian
ideas with deconstruction.
Jameson would launch a similar (and immensely influential) critique against the Formalists in his The Prison-House of Language of
1972, but apparently in ignorance of his precursors in the Bakhtin Circle, whose main
works had only just begun to appear in French
and German, and would only appear in English over the course of the next several years.
Another terminological option, equally canonical, would be synchrony versus diachrony, but
8
9
10
11
12
69
this casts the opposition in terms ‘‘native’’ to
the structuralist camp, one of the parties to the
controversy. Better, surely, to stick with
the somewhat more neutral, ‘‘nonaligned’’
terminology of structure versus history.
This move, involving a change of scale from
text-as-system to literature-as-system, leaves
a long legacy in subsequent literary theory,
from the ‘‘historical structuralism’’ of the
Prague school (Galan 1984) to the literary
semiotics of Lotman and the Tartu school
(Shukman 1977) to the polysystem theory
of Even-Zohar and the Tel Aviv school
(Even-Zohar 1990).
Moreover, as Fludernik has pointed out,
Gates actually misidentifies the discourse
mode in at least one passage from Their Eyes
Were Watching God, remarking on the dualvoice potential of FID apropos of a passage in
which no sentence of FID is to be found.
Fludernik is referring not to The Signifying
Monkey, but to an analysis in an essay on
Hurston and Alice Walker (Gates 1989:
148–9). Gates’s lapse, writes Fludernik,
‘‘shows how grossly narratological discussions of free indirect discourse can be misread
in the interests of metaphor and thematic
speculation’’ (1993: 82) – I would prefer to
say, in the interests of historicizing narratological categories. On African-Americanist
appropriations of narrative theory, compare
Hale (1998: 197–220). See also my parallel
critique (McHale 1994: 60–2) of D. A. Miller’s abuse of the poetics of FID in his The
Novel and the Police.
I may be wrong about this. Work on cognitive approaches presented at the 2004 Society
for the Study of Narrative Literature Conference, especially Ellen Spolsky’s and Lisa Zunshine’s papers on ‘‘theory-of-mind theory,’’
seem to me promising harbingers of what
may prove to be a genuinely historicist cognitivism. Stay tuned.
Hägg presumably has in mind the account of
‘‘narrative and monstrosity’’ in Gibson (1996:
236–74), though he does not allude to it
directly here.
I am grateful to the participants in ‘‘Contemporary Narrative Theory: The State of the
Field,’’ for their feedback, and especially to
70
Brian McHale
Monika Fludernik and David Herman. I have
also profited from the advice of Lubomı́r
Doležel, Bob Griffin, Steve Kern, Randall
REFERENCES
AND
Bakhtin, M. ([1963] 1984). Problems in Dostoevsky’s
Poetics, ed. and trans. C. Emerson. Minneapolis:
University of Minnesota Press.
Bakhtin, M. ([1975] 1981). The Dialogic Imagination: Four Essays, ed. and trans. C. Emerson
and M. Holquist. Austin: University of Texas
Press.
Chatman, S. (1990). ‘‘What Can We Learn from
Contextualist Narratology?’’ Poetics Today 11(2),
309–28.
Clark, K. and Holquist, M. (1984). Mikhail
Bakhtin. Cambridge, MA and London: Harvard
University Press.
Culler, J. ([1975] 2002). Structuralist Poetics: Structuralism, Linguistics and the Study of Literature.
London: Routledge.
DeMan, P. (1983). ‘‘Dialogue and Dialogism.’’
Poetics Today 4(1), 99–107.
Emerson, C. (1997). The First Hundred Years of
Mikhail Bakhtin. Princeton, NJ: Princeton University Press.
Even-Zohar, I. (1990). Polysystem Studies. Special
Issue, Poetics Today 11(1).
Fludernik, M. (1993). The Fictions of Language and
the Languages of Fiction: The Linguistic Representation of Speech and Consciousness. London and
New York: Routledge.
Fludernik, M. (2003). ‘‘The Diachronization of
Narratology.’’ Narrative 11(3), 331–48.
Galan, F. W. (1984). Historic Structures: The Prague
School Project, 1928–1946. Austin: University of
Texas Press.
Gates, H. L. (1988). The Signifying Monkey:
A Theory of African-American Literary Criticism.
New York and Oxford: Oxford University
Press.
Gates, H. L. (1989). ‘‘Color Me Zora: Alice
Walker’s (Re)writing of the Speakerly Text.’’
In P. O’Donnell and R. C. Davis (eds.), Intertextuality and Contemporary American Fiction
(pp. 144–67). Baltimore, MD: Johns Hopkins
University Press.
Stevenson, and the editors of the present
volume.
FURTHER READING
Genette, G. ([1972] 1980). Narrative Discourse,
trans. J. E. Lewin. Ithaca, NY: Cornell University Press.
Genette, G. ([1968] 2001). ‘‘Vraisemblance and
Motivation,’’ trans. D. Gorman. Narrative 9(3),
239–58.
Gibson, A. (1996). Towards a Postmodern Theory of
Narrative. Edinburgh: Edinburgh University
Press.
Hale, D. J. (1998). Social Formalism: The Novel in
Theory from Henry James to the Present. Stanford
CA: Stanford University Press.
Jakobson, R. and Tynjanov, J. (1971 [1928]).
‘‘Problems in the Study of Language and Literature.’’ In L. Matejka and K. Pomorska (eds.),
Readings in Russian Poetics: Formalist and Structuralist Views (pp. 79–81). Cambridge, MA:
MIT Press.
Jameson, F. (1972). The Prison-House of Language:
A Critical Account of Structuralism and Russian
Formalism. Princeton, NJ: Princeton University
Press.
Lovejoy, A. O. ([1938] 1948). ‘‘The Historiography of Ideas.’’ In Essays in the History of Ideas
(pp. 1–13). Baltimore, MD: Johns Hopkins
University Press.
Lovejoy, A. O. ([1936] 1957). ‘‘Introduction: The
Study of the History of Ideas.’’ In The Great
Chain of Being: A Study of the History of an Idea
(pp. 3–23). Cambridge, MA: Harvard University Press.
McHale, B. (1994). ‘‘Whatever Happened to Descriptive Poetics?’’ In M. Bal and I. E. Boer
(eds.), The Point of Theory: Practices of Cultural
Analysis (pp. 56–65). Amsterdam: Amsterdam
University Press.
Morrison, M. (2002). ‘‘Why Modernist Studies
and Science Studies Need Each Other.’’ Modernism/Modernity 9(4), 675–82.
Morson, G. S. and Emerson, C. (1990). Mikhail
Bakhtin: Creation of a Prosaics. Stanford, CA:
Stanford University Press.
Ghosts and Monsters
Richardson, A. and Steen, F. (2002). ‘‘Literature
and the Cognitive Revolution: Introduction.’’
Poetics Today 23(1), 1–8.
Tomashevsky, B. ([1925] 1965). ‘‘Thematics.’’ In
L. T. Lemon and M. J. Reis (eds.), Russian
Formalist Criticism: Four Essays (pp. 61–95).
Lincoln: University of Nebraska Press.
Shklovsky, V. ([1929] 1990). Theory of Prose, trans.
B. Sher. Normal, IL: Dalkey Archive Press.
Shukman, A. (1977). Literature and Semiotics: A
Study of the Writings of Yu. M. Lotman.
Amsterdam and New York: North-Holland
Press.
71
Sternberg, M. (1983). ‘‘Mimesis and Motivation:
The Two Faces of Fictional Coherence.’’ In J. P.
Strelka (ed.), Literary Criticism and Philosophy
(pp. 145–88). University Park, PA: Pennsylvania State University Press.
Stevenson, R. (forthcoming). ‘‘Modernist Narrative.’’ In D. Herman, M. Jahn, and M.-L. Ryan
(eds.), The Routledge Encyclopedia of Narrative
Theory. London and New York: Routledge.
Vološinov, V. N. ([1929] 1973). Marxism and the
Philosophy of Language, trans. Ladislaw Matejka
and I. R. Titunik. Cambridge, MA: Harvard
University Press.
PART I
New Light on Stubborn
Problems
4
Resurrection of the Implied
Author: Why Bother?
Wayne C. Booth
It will surprise no reader here that the author of The Rhetoric of Fiction ([1961] 1983) is
appalled by the number of critics who have embraced the ‘‘death of the author.’’
(Actually it has always been only the implied author;1 no one has claimed that the fleshand-blood author never existed.) How could anyone ever believe that the author’s
intentions about a work are irrelevant to how we read it? The critics were of course
justified in claiming that the author’s expressed intentions, outside the text, could be
in total contrast to the intentions finally realized in the finished text. But does not
that difference dramatize the importance of the implied author/author distinction?
This whole essay could be devoted to refuting this or that absurd assassination
attempt, followed by a report on the best efforts at resuscitation. Some of the latter,
like Jim Phelan’s and Peter Rabinowitz’s, are so good that I’m tempted simply to copy
them. Instead I shall offer one more sermon on why we should all work to keep the
concept alive, regardless of how we define it.
When I first wrote about the implied author (IA), I was driven by at least three
motives, all triggered by anxiety about the critical scene in the 1950s.
1 Distress about the widespread pursuit of so-called objectivity in fiction. Defensible novelists, many were saying, must engage in showing, not telling, thus leaving it to the
reader to perform all judgments. Praiseworthy novelists must kill off all open signs of
the author’s opinions. Authorial commentary is not just frequently dull; it is always a
violation of true ‘‘poetic’’ quality.
Long before Barthes and Foucault and others made that assassination attempt
explicit, critics were proclaiming that genuinely admirable fiction purged itself of
all signs of the writer’s opinions about what was artistically shown. Praiseworthy
fiction had to be presented objectively, with the author’s opinions about characters and
events not merely concealed but fully purged.2
76
Wayne C. Booth
This view often led to a downgrading of the narrative mastery practiced by geniuses
like Joseph Fielding, Jane Austen, and George Eliot – not to mention many great
European and Russian novelists; often it led to outright misreadings. Many of the
‘‘point-of-viewists’’ simply ignored the ways in which openly expressed authorial
rhetoric can in itself be a major aesthetic creation. If space allowed I would quote a
couple of lengthy authorial commentaries by ‘‘George Eliot,’’ the male IA created by
the woman genius, Marian Evans; her/his aggressive ‘‘intrusions’’ had not only aided
me in reading her novels but led me to admire, and even love, the creator. Would I
have fallen in love with Marian Evans if I’d known her? Depends on the circumstances
in which I encountered the flesh-and-blood person (the FBP). But her IAs, employing
manifold self-conscious intrusions, are fabulous.
2 Annoyance over students’ misreadings. Though my college students were not at all
bothered by the telling practices of the great self-conscious narrators (they had read
none of the diatribes), they seemed too often ignorant about the differences between
the narrators and the IAs, and between the IAs and the FBPs creating them. Too many
of them had never learned to distinguish – especially as they read the so-called
objective modern fictions – the narrative voice from that of an author deliberately
creating a partially or totally unreliable voice.
My most troubling example of frequent misreadings, as I now remember it, was
Salinger’s Catcher in the Rye. Students consistently identified so fully with Holden
Caulfield that they missed the ironic clues that Salinger provided about his hero’s
serious faults and weaknesses. Most of Holden’s words were taken as if the author
wanted everything he said to be read as straight. The critical neglect of such
widespread misreading bothered me a lot.
3 A ‘‘moralistic’’ distress about how critics ignored the value of rhetorical ethical effects – the
bonding between authors and readers. This one produced many attacks against my
emphasis on ethics, as critics went on asserting that not only poems but novels should
not ‘‘mean but be,’’ and that novels, like poems, ‘‘make nothing happen.’’ Though
fewer critics made this ‘‘aestheticist’’ assertion about fiction, many did buy into
assertions like Oscar Wilde’s ‘‘There is no such thing as a moral or immoral book.
Books are well written or badly written. That is all.’’3 Though that rejection of ethical
concerns is still widespread, more critics these days than in the 1950s are acknowledging that there is such a thing as an ethics of fiction. In the responses to J. M.
Coetzee’s Nobel Prize in 2003, for example, we encountered again and again open
claims about how the justified winner’s novels, laden as they are with depressing
episodes, have improved our ways of responding to the floods of cruelty and suffering
that we encounter in the shit-laden world. Unlike the ‘‘pure’’ ‘‘poetic’’ critics back
then, most critics are admitting, some only tacitly, that great fiction educates us
ethically – unless we misread it.4 But in the 1950s and 1960s, increasing numbers
embraced a movement soon to be called ‘‘reader response,’’ claiming that interpret-
Resurrection of the Implied Author
77
ation is all a matter of the reader’s responsibility: ‘‘My reading of what Henry James’s
characters are doing is much more important than James’s intention, so what do I have
to learn from him? Nothing.’’5
Those three motives are now reinforced by a fourth, as I’ve done more thinking about
how authorial creation of IAs relates to the universality of our daily, hourly, dependence on constructive and destructive role playing. In every corner of our lives,
whenever we speak or write, we imply a version of our character that we know is
quite different from many other selves that are exhibited in our flesh-and-blood
world. Sometimes the created versions of our selves are superior to the selves we
live with day by day; sometimes they turn out to be lamentably inferior to the selves
we present, or hope to present, on other occasions. A major challenge to all of us is
thus to distinguish between beneficial and harmful masking. And that challenge is
especially strong in literary criticism.
Some decades ago Saul Bellow dramatized wonderfully the importance of authorial
masking, when I asked him, ‘‘What’re you up to these days?’’ He said ‘‘Oh, I’m just
spending four hours each day revising a novel, to be called Herzog.’’ ‘‘What does that
amount to, spending four hours every day revising a novel?’’ ‘‘Oh, I’m just wiping out
those parts of my self that I don’t like.’’
In almost all of our utterances we imitate Bellow, consciously or unconsciously–
especially when we have time for revision. We simply wipe out those selves that we
don’t like, or that at least seem inappropriate for the moment. If everything we said
were unrevised, simply a blurting out of ‘‘sincere’’ blasts of undoctored feelings
and thoughts, wouldn’t life become intolerable? Would you want to go to a restaurant
where the boss instructs waiters never to smile unless they sincerely feel like smiling?
Would you want to go on teaching if your administration required you never to
perform a more cheerful, knowledgeable persona than you felt you were as you walked
toward class? Would you want to read poems by Yeats if they were nothing but
undoctored records of his deeply troubled life? Our lives would be disastrous, hour by
hour, if everyone swore to be ‘‘sincere’’ at every moment.
Even more important is the fact that if we don’t think about the differences between
good and bad masking, we are much more likely to get sucked in destructively by the
bad kinds – such as politicians’ dishonest rhetrickery. And won’t we also fail to
appreciate our ethical dependence on the good kinds: the IAs created by our spouses,
at their best, our bosses, at their best, our journalists, at their best, and . . . well, name
your favorite beneficial posers. For me, as this essay will demonstrate, they are the
authors who know how to wipe out the selves they do not like.
The plain ethical fact is that we all, at least to some degree, derive our models of how
to live by taking in, absorbing, the better selves, the IAs, created by others – especially
authors. It’s of course true that, for most of us, the people we live with – our parents and
siblings and friends – have even stronger effects on us as models – good and bad – than
any literary IAs we meet. But even those friends and relatives have influenced us by
managing, at least some of the time, to wipe out those selves they do not like. And for
78
Wayne C. Booth
most of us, they are the ones who first managed to engage us in literary engagement –
filling our lives from infancy on with IAs of amazing diversity.
What is good and what is questionable about the inescapable universality of such
influence by projected – often totally faked – selves? They are always to some degree
masks covering much more complex, and too often much less admirable, selves.
Should we celebrate what some people would call mere hypocrisy, such wearing of
false masks? Is it good for us, because it provides models for living better than would
be provided by the authors’ actual lives, or should we condemn all of it as if it were
Schwarzenegger’s kind of hypocritical performances?
Saving such threatening questions for a possible book about it, I for now
retreat into the corner that seems to me the most important one: literary masking of
the kind Bellow reports. I am certain that we should all feel deeply grateful to Bellow for
the wiping out he did. Through my years of knowing him personally I encountered
other versions that nobody would like – Bellow-versions that, if allowed to dominate
his novels, would have totally destroyed them. What’s more, those who have studied his
manuscripts (as I have not) find confirmation for that claim: he chose, from thousands of
pages, a few hundreds that presented IAs he really liked, and that I love.
Obviously every judgment of authorial character must be tentative. But only the
cynic could ever claim that there are no real moral distinctions: ‘‘Since it’s all
chicanery, why bother?’’ When seriously engaged authors grant us their works, the
FBP has created an IA who aspires, consciously or unconsciously, for our critical
joining. And the IAs are usually far superior to the everyday lives, the FBPs.
Since most of my work about that point has concentrated on fiction, I’ll shift here to
poetry. The point will be not only that IAs are valuable to us as models but that our
knowledge of the difference between cleansed IAs and contemptible FBPs can actually
enhance our admiration of the literary works presenting the IAs. (So far as I know, this
point has never before been dramatized.)
What differences from fictional IAs do we meet when we read poems? Most
obviously, we less often meet ironic portraits of deliberately flawed, intentionally
unreliable narrators. Only rarely does a poet create a consistently dubious voice like
that of Huck Finn or Kazuo Ishiguro’s butler in The Remains of the Day. What we more
often meet are, more reliably speaking, thoroughly cleansed personae who imply a total
identity with the IA whom the FBP hopes we will join: ‘‘I, the poet, speak to you
directly in my true voice.’’ But if we probe the biographies and autobiographies of any
great poet from ancient times to the present, we discover that the poetic self has
emerged dressed up elegantly, exhibiting a sensitivity to life’s woes and blisses that
careful readers find themselves longing to possess – but that the FBP has often violated
in everyday behavior. To paraphrase Yeats’s point about it: it is by struggling with our
selves that we create poetry, in contrast with the rhetoric we produce when we struggle
with others. (I would of course expand the word ‘‘rhetoric’’ to include the internal
quarrels.)
Resurrection of the Implied Author
79
It’s true that some poets, like Browning, have created deliberately unreliable poetic
voices.6 And there are many others, in recent times, who, like Sylvia Plath, have
beautifully revealed and recreated their self-destructive faults and miseries – as if
practicing total undoctored honesty. But in doing so, they are still realizing, during
the very act of creating the poems, selves far superior to those who cursed their
spouses over breakfast.
Some readers are deeply troubled when biographies reveal just how awful was the life
of the FBP. When a biographer of a highly admired poet uncovers ‘‘warts and all,’’ many
feel that the poetry has been sullied. How can we admire T. S. Eliot’s poems as much as we
did formerly, now that we know how hard he worked to clean out from them aspects of
his emotional life that he thought would ‘‘overpersonalize’’ them? Poetry, Eliot claimed,
is ‘‘not a turning loose of emotion, but an escape from emotion; it is not the expression
of personality, but an escape from personality. But, of course, only those who have
personality and emotions know what it means to want to escape from such things’’
(Eliot [1917] 1932: 10). In escaping emotion and personality of one kind, while creating
emotions and personality of another kind, isn’t Eliot’s masking contemptible?
Wouldn’t he have been an even greater poet if he had decided to ‘‘let it all burst out’’?
Not surprisingly, some would answer ‘‘yes,’’ attacking Eliot and other poets for thus
producing poetry less ‘‘honest’’ than poetry ought to be. There is by now almost a
whole industry devoted to ‘‘outing’’ authors’ various suppressions of their questionable selves, often suggesting that the poems would be better if the poets had been
more honest. It should not surprise you that my answer is ‘‘No!’’ Most of the poetry
we admire would be much worse if the authors had failed to put on a mask. Shouldn’t
we be grateful for their cleansing?7 And should we not admire them for their ability
in performing that laundry work?
As the first of only two examples here, consider Robert Frost, a poet who has been
almost viciously ‘‘outed’’ by some biographers. Who is the Robert Frost we meet in
his poem, ‘‘A Time to Talk’’?
When a friend calls to me from the road
And slows his horse to a meaning walk,
I don’t stand still and look around
On all the hills I haven’t hoed,
And shout from where I am, ‘‘What is it?’’
No, not as there is a time to talk,
I thrust my hoe in the mellow ground,
Blade-end up and five feet tall,
And plod: I go up to the stone wall
For a friendly visit.
(Frost [1916] 1939: 156)
First, what kind of person is the narrator, the speaker here?
80
Wayne C. Booth
He’s a man whom I cannot resist admiring and do not want to resist: a dutiful,
hard-working farmer who, though facing the pressing task of hoeing the hills, cares so
much about friendship that he’ll drop his important work and chat. He’s a man who
loves the mellow farmland, but leaves it for the sake of a friendly visit; good talk, in a
rural scene with a friend, is for him a higher value.
But who, behind that narrator, is the IA creating him? Well, in one sense they
are identical; the IA obviously intends no ironies against the speaker. But he is not
only a friendly farmer; though sharing the virtues of the speaker, he is a much more
complex man devoted to poetic form, working hard – probably for hours or days – to
achieve effective rhymes that obey his rule that no reader should be able to claim that
a rhyme was determined only by the rhyming: road, hoed; walk, talk; around, ground;
tall, wall; and – separated by six lines – ‘‘What is it’’ and ‘‘friendly visit.’’ He’s also
working hard with meter and line length, so that he can surprise us with the only
short line, the final one: ‘‘For a friendly visit.’’ To me that’s a lot harder work than
plowing a field (I’ve worked at both), and it’s wonderful to meet a man who, though
he rightly loves farming, considers friendly talk even more important, and yet
considers most important of all the writing of a beautiful poem; for all we know he
hasn’t touched a hoe for years. The result: though the poem is by no means Frost’s
greatest, his IA is moving in the direction of greatness. He is a weirdly richer
character than the reliable narrator, though not in shocking contrast: after all, a
devoted farmer could also be a poet, and both a farmer and a poet could love to
chat with a neighbor.
But where meanwhile is FBP Frost? You may want to join those who refuse to ask
that question, because the answer may diminish the pleasure of the poem. The
contrast between those first two Frosts with the Frosts portrayed in his biographies
is shocking.8 The first and most influential of the negative exposures called him ‘‘an
appalling man, petty, vindictive, a dreadful husband and parent . . . a monster, a man
of systematic cruelty. . . a man who pretended to be a rustic peasant type but who was
always well-to-do and urbane’’ (Thompson 1966: 27).9 The more recent biographers
don’t portray him as quite that bad, and some, without denying appalling faults,
portray him much more favorably (see Parini 1999).10 But none of them reveal a man
I would like to have as a close neighbor or relative or lunch companion. The man I’d
love to have as a next door neighbor or brother is the one who emerges from that
poem, or from the even better ones, or the wonderful poser whose lecture I attended
when a college student.
Does this contrast diminish my admiration for the poems, or, as I claim here,
actually somehow raise it a bit?
An even more complex collection of contrasting selves is presented by Sylvia Plath. Of
all the relatively small number of poets whose lives I’ve studied, she is top contender
for prize as ‘‘FBP With Largest Collection of Contradictory IAs.’’ As her husband, and
her journals, and her many biographers, have revealed, she herself felt divided about
just which of her poems really fit the person she wanted to appear to be. She simply
Resurrection of the Implied Author
81
did not trust any one of her many voices, even when she felt that the poems and stories
they created were good. As she said when explaining why, after many other changes of
title for her first collection of poems, she chose The Devil of the Stairs: ‘‘ . . . this title
encompasses my book and ‘explains’ the poems of despair, which is as deceitful as hope
is’’ (Plath 1981: 13).
She had a terrible time deciding not just between the voice of despair and the voice
of hope but the voices of anger, of physical violence, of revenge, of sexual bliss and
disappointment. Ted Hughes explains that when she finally settled on the title Ariel,
not long before she committed suicide, she ‘‘omitted some of the more personally
aggressive poems from 1962, and might have omitted one or two more if she had not
already published them herself in magazines’’ (Plath 1981: 15). Responding to Plath’s
mother’s distress at how she is portrayed in the poems, Janet Malcolm rightly
explains:
It seems never to have occurred to Mrs. Plath that the persona of Ariel and The Bell Jar
was the persona by which Plath wished to be represented and remembered – that she
wrote this way for publication because this was the way she wished to be perceived, and
that the face she showed her mother was not the face she wished to show the reading
public.
(Malcolm 1994: 15)
I would put it even more strongly, as ‘‘the self her mother knew was not the self she
hoped really to be’’ (Malcolm 1994: 18).
When Hughes decided, after Plath’s death, to publish a collection, he weeded out
some of the contradictions, explaining in his Introduction to the book: ‘‘Several
advisers had felt that the violent contradictory feelings expressed in those [late] pieces
might prove hard for the reading public to take. . . . This apprehension showed some
insight’’ (Plath 1981: 15). Her journal entries (only a fraction of which I’ve read) are
full of stories, many perhaps true and some obviously jazzed up with the thought of
turning them into publishable versions, her disliked selves now wiped out. But in
most of them we can detect a genuine effort to find and project this or that superior
‘‘self,’’ and particularly the self that will know how to deal with being not just a
woman but – too often – a miserable woman.
Throughout her poetry, Plath does often reveal that she is trying hard to give an
honest portrait of some kind of real damaged self. But we can obviously thank our
good fortune that most of those selves were escaped when she sat down to write the
poems about them. Should we feel any regret that she did not write a poem about
the time when, furious with Ted Hughes for his philandering, she tore ‘‘into small
pieces all his work in progress of the winter of 1961, as well as his copy of
Shakespeare’’ (Malcolm 1994: 18)?
Even her latest poems about the approaching suicide imply an author who is still
very creatively alive, while contemplating death. Here is the concluding moment,
from what was probably her last poem, ‘‘Edge’’:
82
Wayne C. Booth
The woman is perfected.
Her dead
Body wears the smile of accomplishment,
The illusion of a Greek necessity
Flows in the scrolls of her toga,
Her bare
Feet seem to be saying:
We have come so far, it is over.
Each dead child coiled, a white serpent,
One at each little
Pitcher of milk, now empty.
She has folded
Them back into her body as petals
Of a rose close when the garden
Stiffens and odors bleed
From the sweet, deep throats of the night flower.
The moon has nothing to be sad about,
Staring from her hood of bone.
She is used to this sort of thing.
Her blacks crackle and drag.
(Plath 1981: 272–3)
After reading that poem aloud several times, I find myself not just admiring it but in
effect loving the author implied by every stroke: a wonderfully different person from
the one I have met in her diaries and in some of her more careless poems. She is of
course thinking about suicide, contemplating it, even planning it. But for the time
being, she is creating a beautiful poem about how contemplating suicide feels. She is
thinking about the awful way in which to commit suicide will for her be a metaphorical killing of her children: her past life with them will disappear, the loveliness
somehow disappearing into – no, not sadness: the moon (the world that transcends
human emotion) has nothing to be sad about.
At the end she – the creator she chooses to project, no matter what Plath had actually
felt five minutes before or yesterday – is attempting to place her coming death, with
poetic force, into the general truth about the universality of death. And meanwhile,
probing for that truth, she is also aspiring for poetic beauty, excellence of poetic structure.
Thinking as a dying self, she is creating a self capable of writing a beautiful poem while
feeling suicidal. At the risk of boring those of you who have by now read the poem aloud,
note how she handles the rhymes: ‘‘in the scrolls of her toga,’’ and on through rose, close,
odors, throats, bone. It’s as if she were sighing: Oh, oh, oh, death where is thy sting?
But the flowing ‘‘o’’s change to nastier vowel rhymes and harsh, explosive alliteration and half alliteration:
Resurrection of the Implied Author
83
Each dead child coiled, a white serpent . . .
And finally a concentration of rhyme and alliteration:
blacks crackle and drag.
Where do her blacks drag Plath and her readers?11 Into a powerful confrontation
with death!
We are told that Plath wrote ‘‘Edge,’’ and others of her best poems, in those final
hours, waking at 5 a.m. in a freezing unheated apartment, financially desperate,
miserably angry against Ted for his affair, overwhelmed with the care of the two
children, desperate for financial help from friends. She must have felt, for a few hours
each morning as she labored with the poems, that she had finally found the true self
that she wanted to express. And she did it. But as we’ve seen, that creative self was
still a mask – a mask that was torn off each morning as soon as the children awoke.
Throughout her life she had been troubled by self-consciousness about her masking –
sometimes as a totally passive, dutiful domesticated woman, sometimes as almost a
whore, sometimes as . . . who knows?
Meanwhile, who are the diverse selves you and I enact as readers of such a poem?
The full value of a poet’s masking can be understood only when we acknowledge
the importance of what it does to our own diverse masks – what we gain from the
encounter. (At the risk of overdramatizing, I’ll label them as diverse ‘‘Booths,’’ but
you as reader may want to substitute your own name.)
Well, first there is Booth1 , the totally engaged reader who initially tried to follow
the poem faithfully and accurately at every point, laboring to join the IA in every
word, hoping to become the Implied Reader the IA hoped for.
Booth1 attempts to exercise Coleridge’s total suspension of disbelief, not allowing
thoughts to intrude about the complex critical matters labored on in an essay like
this. He attempts simply to enter the poem, to join the IA, even wishing that he
could himself, when the time comes to face death, emulate Plath by writing something creative about it.
In that sympathetic reading, Booth1 is consciously or unconsciously joining all
lovers of poetry, and especially lovers of the special type who love to read thoughts
about death. They are not those old-fashioned readers who abhor free verse, or those
optimists who abhor poems that fail to be ‘‘nice’’ and cheerful, or those hyper‘‘objective’’ folks who detest obvious rhyming. The implied readers – Peter Rabinowitz’s ([1987] 1998) ‘‘authorial audience’’ – must have read a good deal of other
modern poetry or they’ll fail in reading this one. Yet they are also lovers of prosodic
richness as amazing as Alexander Pope’s, with his commandment: ‘‘The sound must
seem an echo to the sense’’: close attenders to details, appreciating the subtle free-verse
lyrical effects I have only partially traced.
Meanwhile there is a version of Booth1 , call it Booth1a , who happens to know
throughout the time of reading – unlike some other possible poetry lovers reading
84
Wayne C. Booth
carefully – that this poet committed suicide shortly after writing this poem; the implied
‘‘Plath’’ must have sensed that such a reader will be especially touched, as I am, by
her managing this creative moment at such a desperate time of life. Just think of
how differently we would read this poem if we did not know that she committed
suicide.
While aspiring to join that authorial audience, thinking of myself as sincerely
putting on those quite different masks, I must of course confess to the presence here
of several additional Booths. There is, most embarrassingly, the intrusive critic, Booth2 ,
who is exploiting the poem in order to write this essay. Motivated by his book-effort
grappling with good and bad masking, sometimes even labeling it ‘‘hypocrisy
upward and downward,’’ he has imposed his critical interests in a way that would
no doubt feel offensive, or at least irrelevant, to any one version of Plath’s selves. Booth2
has partially crippled Booth1 , deflecting him somewhat from a full joining:
Booth2 stands cruelly above – or should I say ‘‘below’’? – the IA, the creating, suffering,
persona. He would of course like to claim that he is himself an example of constructive
masking or posing: except for this sentence, his mind, his soul, his self is totally
occupied with the honest critical pursuit of truth about posing and masking, and its
relation to this poem.
Meanwhile, Booth3 , the FBP, takes a break, goes to the toilet, reads a page or two of
the daily paper – and thinks a bit about how to revise the not-yet-good-enough
section of this essay dealing with Plath’s poem. He is troubled by an aching back as he
sits too long at his computer: ‘‘what the hell, just abandon it and mail it in!’’ Actually
he’s writing under some pressure because in a few moments friends will come to play
string quartets, and he really ought to be practicing the cello part for Beethoven’s Op.
59 #3 that they’ll play this morning; he’s thus almost totally distracted from any full
engagement with the poem or with thinking about it. He’s especially troubled by
having stumbled earlier today on a passionately negative review of one of his books.
Next day, back to the draft, feeling good about the Beethoven, Booth4 , the lifetime
moralizer, finds himself thinking not about the poem but about suicide. Unlike many
other readers in the audience that Plath must have expected, this Booth has never
personally contemplated or attempted suicide. His feelings will be very different from
those of any reader who happens to have failed in an attempt last year. He actually has an
uncertain but real conviction that suicide is an immoral act that everyone should resist
to the last possible moment. He thus has an irrelevant, aesthetically destructive,
temptation to lecture the FBP-Plath for what her act did to the world, by ennobling
suicide. Yet meanwhile he should be aware that the whole thesis of this essay – thought
about IAs must not be abandoned, because IAs improve us – moves into Booth4 ’s
territory.
As the other Booths have read and reread the poem, as they have revised and
rerevised these sentences about the reading, Booth2 , the critic, goes on masking
here as a totally scholarly/critical persona: the objective pursuer of nothing but the
truth about how poets enrich us with their maskings. Thus he dissociates himself
uneasily from the various other implied or actual audiences: he becomes the
Resurrection of the Implied Author
85
critic anxious about getting his own essay into good shape, thinking of scores of
trivial matters that not only would Plath consider irrelevant but that he knows are
irrelevant – to the poem. He knows that he thus risks harming the poem, imposing
on it the overstanding that can diminish poetic understanding. Only Plath’s poem can
rescue him from that kind of unfair intrusion. And meanwhile, at this very moment,
he is working hard to imply an author of total honesty.
No reader, of whatever moral or intellectual persuasion, can avoid objecting to at least
some of Plath’s maskings, once they are revealed. I (that is, Booth3 ) cannot help
thinking that if she somehow had resisted the ‘‘modernist individualism’’ that her
culture – and family, her friends, her English teachers, the books she read – had
imposed on her, if she could somehow have managed to diminish the anguished search
for the one true self, she might have avoided that suicide. She had the gifts to become,
as some critics have claimed (though Ted Hughes has denied it) a great novelist. All
we can do, at the end of the story, is thank our fate that she finally found the mask that
freed her to write those final poems. That mask, alas, did not save her from turning on
the oven gas. Booth3 is tempted to say (but he’s probably wrong): we would all be a
lot better off if, at the end, she could just have gone on grappling with life, saying, ‘‘In
celebrating suicide, I was just masking as a true poet absorbed in making beautiful
poetry about it.’’
It is important to stress, as we move toward conclusion, that none of this suggest
that masking as a poetic self is a hoax. The created IA is not – as some biographers
have suggested and others have often implied – someone whose very creation should
lead us to condemn the creator as a deplorable phony. The IAs are not only as genuine
versions of Frosts and Plaths as are the FBP sinners. They are in one sense more
genuine, and of course far more admirable and influential: in wiping out the selves
they do not like, the poets have created versions that elevate both their worlds and
ours. Just think how impoverished our lives would be without such acting out of
superior versions.
Meanwhile a chorus of past assassinators of the IA shout at us: why bother about all
these irrelevant – or fake – distinctions? Why not just live with the text, interpreting
it in whatever way seems right? Forget about FBPs, kill off the IAs, and interpret the
poem! Another chorus, though believing in the notion of IAs, will accuse me of having
besmirched Frost and Plath by dwelling on the warts. Why even mention the
gruesome facts about the author’s life? Just enjoy the poem.
I can only repeat the one simple answer that has been implicit throughout: I find
my admiration for the works of Frost and Plath and other effective maskers actually
rising a bit when I learn some contemptible details about their FBPs. How could
creatures with such faults and miseries manage to produce such beautiful moving
works? Well, obviously, they can do it because they have aspired not just to appear
better but to be better than those parts of their FBPs that they deplored. They create a
realer, truer, more genuine version of their selves than those selves who plagued the
86
Wayne C. Booth
world with base behavior. As they sit polishing their works, they either wipe out
the parts of their selves that they do not like, or, when the darker selves get
dramatized, as happens so often in Plath, the superior IA conquers the other versions
of FBP by polishing the poem or novel or play: ‘‘That’s who I really am, the person
able to exhibit those values and brilliant strokes.’’
Another chorus, much smaller, intrudes now to express annoyance at my having
dodged the problem of the difference between the implied author and the actual text.
How can I go on talking as if the actual person implied by the text is the same as the
text itself ? Well, that’s because I still don’t think there is a real difference, at any
moment of reading of the kind Booth1 has attempted. It is only when we start
thinking (Booth2 ) about different time frames and cultural contrasts that we create
versions of the IA that the author would never have dreamt of as he or she created
the text. Of course the IA I recreate by reading the text now is not identical with the
IA I would have recreated 40 or 20 years ago. But my claim is that what I see as
the actual text implies, in every stroke as I read now, the choices I believe were made
by the author then, which in turn imply the chooser. It’s true that the text is always in
one sense ‘‘out there,’’ divorced from the creator, subject to innumerable readings and
misreadings. But at the time of creation, and at the time of my recreation, unless I am
imposing overstanding, the two have been identical. (I resist offering examples of the
appalling flood of current distortions of texts by readers who ‘‘know’’ in advance
what they are going to find in them – and therefore find it, regardless of what the IA
wanted them to find.)
Others may be troubled by my obsession with ethical effect. To that objection I can
only reply, a bit rudely: are you sure that you are not echoing those author-assassinators who were in fact overstanding misreaders? As they pursued only theoretical or
structural questions, they in fact failed to experience the work as intended. Having
experienced no emotional bonding with the author through the characters, they could
thus dismiss ethical effect. Only readers who have known the thrill of joining authors
in their full engagement, their full achievement, their full cleansing or purification,
leaving abstract critical questions to one side until the poem has been fully experienced – only such fully hooked readers ever discover how that joining changes
one’s life.
As we merge ourselves with the created self who has created the work, as we
recreate the work as intended, we resemble more and more the IA who achieved the
creation. And when we learn of the shoddy selves behind the creation, we not only can
admire the creations even more than previously; we observe models of how we
ourselves might do a bit of creating of better selves.
Though we may later decide that some of our joinings were harmful, even
disastrous, the subject of that joining cries out for persistent study, as we encounter
the daily flood of IAs – constructive and destructive – not only in literature but in
politics, in journalism, in scholarly works, and in the classroom.
Resurrection of the Implied Author
87
NOTES
1 See Booth 1988, especially index entries under
‘‘Author, implied.’’
2 For a selection from innumerable attacks on
authorial intrusions, before the death of the
author movement, see Lubbock (1921), Ford
Madox Ford (1932), and Caroline Gordon and
Allen Tate (1950).
3 From the Preface to The Picture of Dorian Gray.
For a persuasive argument that Wilde was not
an extreme aestheticist but was always actually
in pursuit of moral or ethical effects – under
idiosyncratic definitions – see Richard Ellmann’s (1988) Oscar Wilde.
4 For a first-class demonstration of why ethical
matters cannot be escaped in narrative, see
Yehoshua (2000).
5 There were innumerable excesses in this
shocking elevation of reader over author,
often by brilliant critics like Stanley Fish
(e.g., Fish 1980). I resist documenting their
absurdities – partly because I am these days
more willing to agree that some of their
claims were sound. The best treatment of the
proper balance between reader-response concerns and full respect for intention is to me
REFERENCES
AND
Beach, J. W. (1932). The Twentieth Century Novel:
Studies in Technique. London: The Century
Company.
Booth, W. C. ([1961] 1983). The Rhetoric of
Fiction. Chicago: The University of Chicago
Press, 2nd edn., with added final chapter.
Booth, W. C. (1968). ‘‘The Rhetoric of Fiction
and the Poetics of Fictions.’’ Novel 1 (Winter),
105–17.
Booth, W. C. (1988). The Company We Keep: An
Ethics of Fiction. Berkeley, CA: The University of
California Press.
Crane, R. S. ([1957] 1967). ‘‘Criticism as Inquiry; or,
The Perils of the ‘High Priori Road,’ ’’ In The Idea
of the Humanities, vol. II (pp. 25–44). Chicago:
The University of Chicago Press.
Donoghue, Denis (1999). ‘‘Lives of a Poet.’’ New
York Review of Books 46 (18), 55–7.
6
7
8
9
10
11
Louise Rosenblatt’s (1995) Literature as
Exploration.
See for example Browning’s brilliantly ironic
‘‘My Last Duchess’’ and ‘‘Mr. Sludge, the
Medium.’’
For example, in his work in progress Modernism and the Ethics of Impersonality, Tim Dean
argues that The Waste Land would not be as
great as it is if Eliot had spent his time
laboring with his homosexual Self.
Well summarized by Denis Donoghue
(1999).
No doubt because of the flurry such an indictment produced, Thompson’s many later
accounts were somewhat less severe. But they
still revealed ways in which Frost’s ‘‘arrogance’’ produced suffering in those around
him.
A small number portray Frost, somewhat
dishonestly in my view, as almost a saint: as
pretty much the man who is implied by the
wonderfully complex, honest poetry.
I can’t figure out why she chooses the word
‘‘blacks,’’ except for the rhyme – and of course
the suggestion of gloom.
FURTHER READING
Eliot, T. S. ([1917]1932). ‘‘Tradition and Individual Talent.’’ In Selected Essays: 1917–1932
(pp. 3–11). New York: Harcourt Brace.
Ellmann, R. (1988). Oscar Wilde. New York:
Alfred Knopf.
Fish, S. (1980). Is There a Text in This Class? The
Authority of Interpretive Communities. Cambridge,
MA: Harvard University Press.
Ford, F. M. (1932). The Twentieth-Century Novel:
Studies in Technique. London: Duckworth.
Frost, R. ([1916] 1939). Mountain Interval. New
York: Henry Holt.
Gordon, C. and Tate, A. (1950). The House of
Fiction. New York: Scribner.
Lubbock, P. (1921). The Craft of Fiction. London:
J. Cape.
Malcolm, J. (1994). The Silent Woman: Sylvia Plath
and Ted Hughes. New York: Alfred Knopf.
88
Wayne C. Booth
Parini, J. (1999). Robert Frost: A Life. New York:
Henry Holt.
Phelan, J. (1989). Reading People, Reading Plots:
Character, Progression, and the Interpretation of Narrative. Chicago: The University of Chicago Press.
Phelan, J. and Rabinowitz, P. J. (eds.) (1994).
Understanding Narrative. Columbus: Ohio State
University Press.
Plath, S. (1981). Sylvia Plath: The Collected Poems,
ed. with introduction T. Hughes. London:
Faber and Faber.
Rabinowitz, P. J. ([1987] 1998). Before
Reading: Narrative Conventions and the Politics of
Interpretation. Columbus: Ohio State University
Press.
Rosenblatt, L. ([1965] 1995). Literature as Exploration. New York: The Modern Language Association.
Thompson, L. (1966). Robert Frost. New York:
Holt, Rinehart, and Winston.
Wimsatt, W. K., and Beardsley, M. C. (1954).
‘‘The Intentional Fallacy.’’ In W. K. Wimsatt,
The Verbal Icon: Studies in the Meaning of Poetry
(pp. 3–18). Lexington: University Press of Kentucky.
Yehoshua, A. B. ([1998] 2000). The Terrible Power
of a Minor Guilt, trans. Ora Cummings. Syracuse, NY: Syracuse University Press.
5
Reconceptualizing Unreliable
Narration: Synthesizing
Cognitive and Rhetorical
Approaches
Ansgar F. Nünning
Introducing the Unreliable Narrator
Ever since Wayne C. Booth first proposed the unreliable narrator as a concept it has
been considered to be among the basic and indispensable categories of textual analysis.
Booth’s well-known formulation, ‘‘I have called a narrator reliable when he speaks for
or acts in accordance with the norms of the work (which is to say the implied author’s
norms), unreliable when he does not’’ (1961: 158–9), has become the canonized
definition of the term and was only challenged quite recently. According to Booth,
the distinction between reliable and unreliable narrators is based on ‘‘the degree and
kind of distance’’ (p. 155) that separates a given narrator from the implied author of a
work. A comparison of the definitions provided in standard narratological works, in
scholarly articles, and in glossaries of literary terms shows that the great majority of
narratologists have followed Booth, providing almost identical definitions of the
unreliable narrator.
What most critics seem to have forgotten, however, is that Booth himself freely
admitted that the terminology for ‘‘this kind of distance in narrators is almost
hopelessly inadequate’’ (1961: 158). There is indeed a peculiar discrepancy between
the importance generally attributed to the question of reliability in narrative and
the unresolved issues surrounding the concept: ‘‘There can be little doubt about the
importance of the problem of reliability in narrative and in literature as a whole. . . .
[But] the problem is (predictably) as complex and (unfortunately) as ill-defined as it is
important’’ (Yacobi 1981: 113). While critics and theorists working within the
tradition of rhetorical approaches to narrative consider the implied author to be
among the important and indispensable categories of textual analysis, structuralist
and cognitive narratologists have argued for the abandonment of the implied
90
Ansgar F. Nünning
author and for a radical reconceptualization of the unreliable narrator. In contrast to
many other narratological categories (e.g., such well-defined and familiar terms as
events and actions, homodiegetic and heterodiegetic narrators, order, duration, and
frequency) both of these key concepts of narrative theory have become the subject of
intense debates and even heated controversy.
As the debates surrounding unreliable narration have shown, the topic attracts a
great deal of attention from most of the main current approaches to narrative, ranging
all the way from rhetorical and ethical approaches, from which the concept originates,
over narratological and feminist approaches to cognitive narratology. There are many
reasons why unreliable narration has become such a central issue in contemporary
narrative theory. First, the topic of narratorial unreliability is an extremely rich
theme, involving a number of intriguing theoretical problems (e.g., the
vexed question of whether or not we need the implied author and the equally intricate
question of how readers negotiate textual inconsistencies and ambiguities). Second,
since unreliable narration is situated at the interface of aesthetics and ethics as well as
of description and interpretation, it combines important theoretical and interpretive
enquiries, with any decision about a narrator’s (un)reliability having far-reaching
interpretive consequences. Third, the fact that unreliable narration is very widespread
in both modern and postmodern fiction provides a great challenge both to every critic
and to our usual definition of an unreliable narrator (cf. Wall 1994). Last but not
least, unreliable narration as a phenomenon is, of course, not confined to narrative
fiction, but can be found in a wide range of narratives across the genres, the media,
and different disciplines.
As the title of this chapter suggests, it attempts to survey recent work on unreliable
narration, showing that narrative theory has made considerable progress in dealing
with this very slippery and complex topic. Narratologists have not only identified the
main problems in traditional accounts, but have also provided a number of useful
terminological and taxonomic distinctions as well as theoretical refinements. Attempts to reconceptualize unreliable narration mainly involve four areas: (1) the
theory and definition of unreliable narration; (2) typological distinctions of different
kinds of unreliability; (3) the textual clues and frames of reference involved in
projections of unreliable narrators; (4) the respective roles of the reader, the text,
and the (implied) author.
Realigning the relation between the cognitive and the rhetorical approaches, this
chapter attempts to show that recent work in cognitive narratology and rhetorical
theory provides the basis for reconceptualizing unreliable narration. The next section
outlines the cognitive reconceptualization of unreliable narration, which can shed
more light on the usually unacknowledged presuppositional framework on which
theories of unreliable narration have hitherto been based. This is followed by a section
that combines the insights of cognitive and rhetorical approaches and argues that the
whole notion of unreliable narration only makes sense when we bear in mind that
ascriptions of unreliability involve a tripartite structure that consist of an authorial
agency, textual phenomena (including a personalized narrator and signals of un-
Reconceptualizing Unreliable Narration
91
reliability), and reader response. The final section then provides a brief summary,
suggesting that much more work needs to be done in this fascinating field of narrative
theory.
A Critique of Conventional Theories of Unreliable Narration
and a Cognitive Reconceptualization: The Role of the Reader
and His/Her Frames of Reference
Unreliability is an effect that most readers intuitively recognize. Though Booth’s
attempt to explain that effect in the particular way outlined above provided a
landmark contribution to narrative theory, that way creates as many problems as it
sets out to solve. The definition of an unreliable narrator provided by Gerald Prince
in his Dictionary of Narratology provides a good starting point for understanding
why this is so: ‘‘A narrator whose norms and behavior are not in accordance
with the implied author’s norms; a narrator whose values (tastes, judgments, moral
sense) diverge from those of the implied author; a narrator the reliability of whose
account is undermined by various features of that account’’ (Prince 1987: 101).
Despite the good job Prince does in summarizing the communis opinio, this definition
is marred by vagueness, because the only yardstick it offers for gauging a narrator’s
unreliability is the implied author, whose status and norms are more difficult to
ascertain than one might think. Nonetheless, most theorists and critics who have
written on the unreliable narrator take the implied author both for granted and for the
only standard according to which unreliability can be determined. In what are
arguably some of the best critiques of classical theories of unreliable narration to
date, Tamar Yacobi (1981, 1987) and Kathleen Wall (1994) hold on to the implied
author as though he or she, or rather it, was the only possible way of accounting
for unreliable narration. Critics who argue that a narrator’s unreliability is to be
gauged in comparison to the norms of the implied author just shift the burden of
determination onto a critical passepartout that is itself notoriously ill-defined
(cf. Nünning 1997b).
Some narratologists have pointed out that the concept of the implied author does
not provide a reliable basis for determining a narrator’s unreliability. Not only are
‘‘the values (or ‘norms’) of the implied author [ . . . ] notoriously difficult to arrive at,’’
as Rimmon-Kenan ([1983] 2003: 101) observes, but the implied author is itself a
very elusive and opaque notion. From a theoretical point of view, the concept of the
implied author is quite problematic because it creates the illusion that it is a purely
textual phenomenon. But it is obvious from many of the definitions that the implied
author is a construct established by the reader on the basis of the whole structure of a
text. If the implied author is conceived of as a structural phenomenon that is voiceless,
one should look at it not as a speaker involved in the structure of narrative transmission, but as a component of the reception process, as the reader’s idea of the author, or
‘‘as a construct inferred and assembled by the reader from all the components of the
92
Ansgar F. Nünning
text’’ (Rimmon-Kenan [1983] 2003: 87). When Chatman (1990: 77) writes that ‘‘we
might better speak of the ‘inferred’ than of the ‘implied’ author,’’ he implicitly
concedes that one is dealing with something that has to be worked out by the reader.
From a Boothian perspective, however, the notion of an inferred author is quite
different from an implied author: the latter is a creation by the author, which the
reader may or may not gauge in practice, whereas an inferred author is a creation by
the reader, which may or may not correspond to the implied author projected by the
flesh-and-blood person who wrote the text (see Booth’s contribution to the present
volume). As Phelan (2005: 41–2) persuasively argues, this difference means that
Chatman’s (1990: chap. 5) ‘‘Defense of the Implied Author’’ is not really a defense
of Booth’s concept but a redefinition masquerading as a defense.
The controversy about the concept of the implied author matters because it carries
far-reaching theoretical implications. First, the concept of the implied author reintroduces the notion of authorial intention, though through the back door, by providing
a terminological link to the sphere of the actual author and authorial values. As
Chatman (1990: 77) has pointed out, ‘‘the concept of implied authorship arose in the
debate about the relevance of authorial intention to interpretation.’’ For many critics,
the implied author provides a terminologically acceptable way of talking about the
author and his or her intention, under the guise of talking about textual phenomena.
Second, allegedly representing the work’s norms and values, the implied author is
intended to serve both as a yardstick for an ethical kind of criticism and as a check on
the potentially boundless relativism of interpretation. Third, the use of the definite
article and the singular suggest that there is only one correct interpretation. In short,
the concept of the implied author appears to provide the critic again with a basis for
judging both the acceptability of an author’s moral position and the correctness of an
interpretation.
The main objections to the concept of the implied author involve its lack of clarity
and theoretical incoherence. Structuralist narratologists have pointed out that it is a
contradiction in terms to define the implied author as the structure of the text’s norms
and to thus conflate it with the text as a whole, while also casting it in the role of the
addresser in the communication model of narrative. They have argued that an entity
cannot be both a distinct agent in the sequence of narrative transmission and the text
itself; furthermore, if the implied author is equivalent to the whole text, and if his or
her counterpart the implied reader is also presumed to be a textual function, then the
implied author is equivalent to or a subsumption of the implied reader (Nünning
1997b). About the only thing that is clear, then, is that such an incoherent concept
cannot provide a basis for determining unreliability.
Conventional theories of unreliable narration are also methodologically unsatisfactory because they leave unclear how the narrator’s unreliability is apprehended in the
reading process. The metaphors that Chatman uses in order to explain how the reader
detects a narrator’s unreliability are a case in point. He resorts to what is arguably one of
the two most popular metaphors in this context, that of ‘‘reading between the lines.’’
Chatman (1978: 233) argues that readers ‘‘conclude, by ‘reading out,’ between the
Reconceptualizing Unreliable Narration
93
lines, that the events and existents could not have been ‘like that,’ and so we hold
the narrator suspect.’’ Such observations, though vivid, fail to shed much light on how
a narrator’s unreliability is actually determined by the reader. The second common
metaphor is that something is going on ‘‘behind the narrator’s back’’ (cf. Riggan 1981:
13, Yacobi 1981: 125). Chatman (1978: 233), for instance, suggests that the implied
author establishes ‘‘a secret communication with the implied reader.’’ Riggan (1981:
13) not only uses almost exactly the same phrase but he also states quite unequivocally
that ‘‘the presence of the implied author’s hand is always discernible behind the
narrator’s back’’ (p. 77). In contrast to Booth (1974), whose four steps for reading stable
irony, outlined in part I of A Rhetoric of Irony, constitute a specific method for discerning
much unreliable narration, Riggan does not, however, enlighten the uninitiated as to
how the hand of the omnipresent implied author behind the narrator’s back may in fact
be discerned.
Despite what common sense would appear to tell us, definition apparently is
a problem with the unreliable narrator because most theories leave unclear what
unreliability actually is and fail to distinguish between moral and epistemological
shortcomings. Most definitions in the wake of Booth have emphasized that unreliability consists of a moral distance between the norms of the implied or real author
and those articulated by the narrator. But other theorists have pointed out that what
is at stake is not a question of moral norms but of the veracity of the account a narrator
gives. Thus Rimmon-Kenan’s ([1983] 2003: 100) definition leaves open whether
unreliability is to be gauged in comparison to the accuracy of the narrator’s account
of the story or to the soundness of his or her commentary and judgments: ‘‘An
unreliable narrator [ . . . ] is one whose rendering of the story and/or commentary on
it the reader has reasons to suspect.’’ The ‘‘and/or’’ construction sounds very open and
flexible but actually it is a bit too nonchalant. Most would agree that it does make a
difference whether we have an ethically or morally deviant narrator who provides
a sober and factually veracious account of the most egregious or horrible events,
which, from his point of view, are hardly noteworthy, or a ‘‘normal’’ narrator who is
just a bit slow on the uptake and whose flawed interpretations of what is going on
reveal that he or she is a benighted fool. It is thus unclear whether unreliability is
primarily a matter of misrepresenting the events or facts of the story or whether it
results from the narrator’s deficient understanding, dubious judgments, or flawed
interpretations.
Aware of these problems, some narrative theorists have suggested terminological
refinements and typological distinctions. Lanser (1981: 170–1) was among the first to
suggest that one should distinguish between an unreliable narrator, that is, a narrator
whose rendering of the story the reader has reasons to suspect, and an untrustworthy
one, that is, a narrator whose commentary does not accord with conventional notions
of sound judgment. Pursuing the same line of inquiry, Olson (2003) distinguishes
between factual unreliability associated with a fallible narrator, that is, a narrator
whose rendering of the story the reader has reasons to suspect, and normative
unreliability displayed by an untrustworthy narrator whose commentary and
94
Ansgar F. Nünning
interpretations do not accord with conventional notions of sound judgment, emphasizing that the two types elicit different responses in readers.
Moreover, there may be a number of different reasons for unreliability, including
‘‘the narrator’s limited knowledge, his personal involvement, and his problematic
value-scheme’’ (Rimmon-Kenan [1983] 2003: 100). Depending on the reason for
unreliability, one can distinguish different types of unreliable narrators such as the
madman, the naı̈ve narrator, the hypocrite, the pervert, the morally debased narrator,
the picaro, the liar, the trickster, or the clown (cf. Riggan 1981). Such a typology
‘‘corresponds to an already semanticized classification of unreliable narrators’’
(Fludernik 1999: 76) which is based on social and literary conventions.
In one of the most sophisticated recent articles on the subject, Phelan and Martin
(1999) have developed the most systematic and useful classification of kinds of
unreliability to date, while also addressing some of the most important theoretical
issues regarding the problem of narrative unreliability. Their heuristic typology is
based on the fact that narrators tend to perform three main functions: (1) they report
on characters, facts, and events; (2) they evaluate or regard the characters, facts, and
events; (3) they interpret or read the characters, facts, and events. Each of these
functions or roles can be thought of as existing along one axis of communication,
resulting in a different kind of unreliability: (1) unreliable reporting occurs along the
axis of facts/events; (2) unreliable evaluating occurs along the axis of ethics/evaluation;
(3) unreliable reading or interpreting occurs along the axis of knowledge/perception.
Distinguishing these three axes of unreliability, Phelan and Martin have also pointed
out that narrators can be unreliable in two different ways along each axis, either by
falling short or by distorting. Consequently, they distinguish six main types of
unreliability: underreporting and misreporting, underregarding and misregarding
(or misevaluating), underreading and misreading.
In contrast to typologies like Riggan’s, which are based on real-life parameters, the
taxonomy proposed by Phelan and Martin is not only much more systematic, it also
has the great merit of being based on a rhetorical model, focusing as it does on the
relations among authorial agency, narrator, and authorial audience. Moreover,
Phelan and Martin emphasize that, regardless of the axis, all deviations require the
authorial audience to infer an understanding of the report, the evaluations, or the
interpretations different from those offered by the narrator. Nevertheless, their
rhetorical approach fails to provide a satisfactory answer to the question of how readers
actually recognize an unreliable narrator (in his or her role as reporter, as evaluator,
and/or as interpreter or reader) when they see one.
Narrative theorists working within a cognitive and constructivist approach to
understanding unreliable narration have argued that the link that has been forged
between the unreliable narrator and the implied author deprives narrative theory of
the possibility of accounting for the pragmatic effects of unreliability. Focusing on the
interactivity between textual modes of representation and readers’ choices in constructing narrative worlds, some theorists (e.g., Yacobi 1981, 1987, 2001, Nünning
1998, 1999) have located unreliability in the interaction of text and reader. Indeed,
Reconceptualizing Unreliable Narration
95
they have argued that unreliability is not so much a character trait of a narrator as it
is an interpretive strategy of the reader. This move has led them to propose a
conceptualization of the relevant phenomena in the context of frame theory as a
projection by the reader who tries to resolve ambiguities and textual inconsistencies
by projecting an unreliable narrator as an integrative hermeneutic device. The reader
or critic accounts for whatever incongruity he or she may have detected by reading the
text as an instance of dramatic irony and by projecting an unreliable narrator. Culler
(1975: 157) has clarified what is involved here: ‘‘At the moment when we propose
that a text means something other than what it appears to say we introduce, as
hermeneutic devices which are supposed to lead us to the truth of the text, models
which are based on our expectations about the text and the world.’’ Similarly, I regard
unreliable narration as not only a structural or semantic aspect of the text but also a
phenomenon that involves the conceptual frameworks readers bring to it.
Determining whether a narrator is unreliable is not just an innocent descriptive act
but a subjectively tinged value judgment or projection governed by the normative
presuppositions and moral convictions of the critic, which as a rule remain unacknowledged. Recent work on unreliable narration confirms Culler’s hypothesis about the
impact of realist and referential notions for the generation of literary effects. Culler
(1975: 144) argues that ‘‘most literary effects, particularly in narrative prose, depend
on the fact that readers will try to relate what the text tells them to a level of ordinary
human concerns, to the actions and reactions of characters constructed in accordance
with models of integrity and coherence.’’ Riggan’s book provides a case in point.
Riggan distinguishes four types of such narrators, which he designates ‘‘picaros,’’
‘‘madmen,’’ ‘‘naı̈fs,’’ and ‘‘clowns.’’ These typological distinctions can best be understood as a way of relating texts to accepted cultural models or to literary conventions.
Riggan (and critics like him) integrate previously held world-knowledge with textual
data or even impose pre-existing conceptual models on the text. The models used for
accounting for unreliable narration provide a context which resolves textual inconsistencies and makes the respective novels intelligible in terms of culturally prevalent
frames.
The information on which the projection of an unreliable narrator is based derives
at least as much from these models and the conceptual schema in the mind of the
beholder as from textual data. In other words: whether a narrator is regarded as
unreliable not only depends on the distance between the norms and values of the
narrator and those of the text as a whole (or of the implied author) but also on the
distance that separates the narrator’s view of the world from the reader’s or critic’s
world-model and standards of normalcy, which are themselves, of course, subject to
change. This process of course raises the question of what particular schema are
deemed relevant.
An analysis of the presuppositional framework on which most theories of unreliable
narration rest is overdue since research into unreliable narration has been based on a
number of questionable conceptual presuppositions, which as a rule remain
implicit and unacknowledged. These underlying presuppositions include certain
96
Ansgar F. Nünning
(1) epistemological and ontological premises; (2) assumptions that are rooted in a
liberal humanist view of literature; and (3) psychological, moral, and linguistic
norms, all of which are based on stylistic and other deviational models. An analysis
of the presuppositional framework on which most theories of unreliable narration are
based reveals that the orthodox concept of the unreliable narrator is a curious
amalgam of a realist epistemology and a mimetic view of literature.
The epistemological and ontological premises consist of realist and by now doubtful notions of objectivity and truth. More specifically, the traditional notion of
unreliability presupposes that an objective view of the world, of others, and of oneself
can be attained. The concept of unreliable narration also implies that human beings
are principally taken to be capable of providing veracious accounts of events, proceeding from the assumption that ‘‘an authoritative version of events’’ (Wall 1994: 37)
can in principle be established or retrieved. Theories of narrational unreliability also
tend to rely on realist and mimetic notions of literature. The concept of the unreliable
narrator is based on what Yacobi (1981: 119) has aptly called ‘‘a quasi-human model
of a narrator,’’ and, one might add, an equally anthropomorphic model of the implied
author.
In addition, theories of narrational unreliability are also heavily imbued with a
wide range of unacknowledged notions that are based on stylistic deviation models or
on more general notions of deviation from some norm or other. The notion of
unreliability presupposes some default value that is taken to be unmarked ‘‘reliability.’’ This is usually left undefined and merely taken for granted. Most critics agree,
however, that reliability is indeed the default value. Lanser (1981: 171), for instance,
argues that ‘‘the conventional degrees zero [are] rather close to the poles of authority.’’
Wall is the first theorist of unreliable narration to shed some light on the presuppositions on which this ‘‘reliable counterpart’’ of the unreliable narrator rests when she
argues that the reliable narrator ‘‘is the ‘rational, self-present subject of humanism,’
who occupies a world in which language is a transparent medium that is capable of
reflecting a ‘real’ world’’ (1994: 21). Vague and ill-defined though this norm of
reliability may be, it supplies the standard according to which narrational unreliability is gauged.
Probing further, we can distinguish some other presupposed norms underlying
determinations of unreliability: (1) all those notions that are usually referred to as
‘‘common sense,’’ (2) those standards that a given culture holds to be constitutive of
normal psychological behavior, (3) some conception of linguistic norms, and
(4) culturally agreed-upon moral and ethical standards. One problem with all of
these tacit presuppositions is that the establishment of norms is much more difficult
than critics acknowledge.
Furthermore, in both critical practice and in theoretical work on unreliable
narration, these different sets of norms are usually not explicitly set out but merely
introduced in passing, and they seldom if ever receive any theoretical examination. Let
me give one typical example: in what is the only book-length study of the unreliable
first-person narrator, Riggan (1981: 36) suggests that the narrator’s unreliability may
Reconceptualizing Unreliable Narration
97
be revealed by the ‘‘unacceptability of his [moral] philosophy in terms of normal
moral standards or of basic common sense and human decency.’’ By saying this, he lets
the cat out of the bag: it is not the norms and values of the implied author that
provide the critic with the yardstick for determining how abnormal, indecent,
immoral, or perverse a given narrator is, but ‘‘normal moral standards,’’ ‘‘basic
common sense,’’ and ‘‘human decency.’’ The trouble with seemingly self-explanatory
yardsticks like ‘‘normal moral standards’’ and ‘‘basic common sense’’ is that no
generally accepted standard of normality exists which can serve as the basis for
impartial judgments. In a pluralist, postmodernist, and multicultural age like ours
it has become more difficult than ever before to determine what may count as ‘‘normal
moral standards’’ and ‘‘human decency.’’ In other words: a narrator may be perfectly
reliable compared to one critic’s notions of moral normality but quite unreliable in
comparison to those that other people hold. To put it quite bluntly, a pederast
would not find Humbert Humbert, the fictitious child molester and narrator of
Nabokov’s Lolita, unreliable, though even pederasts may have moral codes that
Humbert violates; a male chauvinist fetishist who gets his kicks out of making love
to store mannequins is unlikely to detect any distance between his norms and those of
the mad monologist in Ian McEwan’s (1979) ‘‘Dead As They Come’’; and someone
used to watching his beloved mother disposing of unwelcome babies would not even
find the stories collected in Ambrose Bierce’s The Parenticide Club in any way
objectionable.
In addition, there are a number of definable textual clues to unreliability, and what
is needed is a more subtle and systematic account of these signals. Unreliable narrators
tend to be marked by a number of textual inconsistencies, including paratextual
elements as well as conflicts between story and discourse. Other textual elements that
signal a narrator’s unreliability may range from internal contradictions within the
narrator’s discourse over discrepancies between their utterances and actions (cf.
Riggan 1981: 36, who calls this ‘‘a gaping discrepancy between his conduct and the
moral views he propounds’’), to those inconsistencies that result from multiperspectival accounts of the same event (cf. Rimmon-Kenan [1983] 2003: 101). The range of
clues to unreliability that Wall (1994: 19) simply refers to as ‘‘verbal tics’’ or ‘‘verbal
habits of the narrator’’ (p. 20) can and should be further differentiated, for example by
specifying the linguistic expressions of subjectivity. Due to the close link between
subjectivity on the one hand and the effect called unreliability on the other the
virtually exhaustive account of categories of expressivity and subjectivity that Fludernik (1993: 227–79) has provided is useful for drawing up a list of grammatical
signals of unreliability, which can be further differentiated in terms of the linguistic
expressions of subjectivity. The establishment of a reading in terms of unreliable
narration frequently depends on the linguistic and stylistic evocation of a narrator’s
subjectivity or cognitive limitations.
To develop a more viable theory of unreliability, we need a pragmatic and cognitive
framework that takes into consideration the world-model, values and norms, and
conceptual information previously existing in the mind of reader or critic, and the
98
Ansgar F. Nünning
interplay between textual and extratextual information. To put it another way, we
need an interactive model of the reading process and a reader-oriented pragmatic or
cognitive framework. Fludernik’s (1993: 353) explanation of irony illuminates how
this might be conceptualized: ‘‘textual contradictions and inconsistencies alongside
semantic infelicities, or discrepancies between utterances and action (in the case of
hypocrisy), merely signal the interpretational incompatability [ . . . ] which then
requires a recuperatory move on the reader’s part – aligning the discrepancy with an
intended higher-level significance: irony.’’ An interactive model of the reading process
alerts theorists of unreliable narration that the projection of an unreliable narrator
depends upon both textual information and extratextual conceptual information
located in the reader’s mind. As Culler, Yacobi, Wall, and others have remarked,
this view of unreliable narration sees it as one kind of naturalization, that is, a way of
bringing the text ‘‘into relation with a type of discourse or model which is already, in
some sense, natural or legible’’ (Culler (1975: 138).
Noticing and clarifying those unacknowledged frames of reference that allow us to
naturalize unreliability also provides the clue to reconceptualizing the phenomenon.
One might begin by distinguishing between frames of reference derived from everyday experience – what we can call referential frames – and those that result from
knowledge of literary conventions. A first referential frame should be based on the
readers’ empirical experience and criteria of verisimilitude. This frame depends on the
assumption that the text refers to or is at least compatible with the so-called real
world and allows us to determine reliability according to the narrator’s behavior in
relation to the norms of that world. A second referential frame depends on the reader’s
knowledge of the social, moral, or linguistic norms relevant for the period in which a
text was written and published (cf. Yacobi 1987) and a third on knowledge of relevant
psychological theories of personality or implicit models of psychological coherence
and normal human behavior.
A second set of models brought into play in order to gauge a narrator’s possible
unreliability involves a number of specifically literary frames of reference. These
include, for example, general literary conventions; conventions and models of literary
genres (cf. Yacobi 1981: 115f.); intertextual frames of reference, that is references to
specific pretexts; stereotyped models of characters such as the picaro, the miles
gloriosus, and the trickster; and last but not least the structure and norms established
by the respective work itself. The generic framework determines in part which criteria
are used when a narrator’s potential unreliability is gauged (cf. Yacobi 1987: 20f.).
A narrator who is considered to be unreliable in psychological or realistic terms may
appear quite reliable if the text belongs to the genre of science fiction.
In addition to such a cognitive turn in the theory of unreliable narration, Zerweck
(2001: 151) has called for a ‘‘second fundamental paradigm shift, one toward greater
historicity and cultural awareness.’’ Since the development of the narrative technique
known as unreliable narration and such cultural frames of reference as norms and
values are subject to historical change, the whole notion of (un)reliability needs to be
historicized and be seen in the context of broader cultural developments. V. Nünning
Reconceptualizing Unreliable Narration
99
and Zerweck argue that, because the ascription of (un)reliability involves interpretive
choices and strategies, it is culturally and historically variable. Using Oliver Goldsmith’s novel The Vicar of Wakefield as a test case of a cultural-historical narratology,
Vera Nünning ([1998] 2004) has demonstrated that only when the historical variability in the construction of meaning and the values of the period in which the work
was written, read, and reviewed are taken into consideration will a narratological
analysis of unreliable narration become valid and historically meaningful. Readers and
critics who do not take cultural and historical differences into account are apt to
misread a sentimental novel like The Vicar of Wakefield, including misreading the
(un)reliability of a narrator like Dr Primrose.
Detecting Unreliable Narration in Practice: The Role of the Text
and the Role of the Author
Proponents of rhetorical approaches to narrative have taken cognitive narratologists
to task for throwing out the textual baby with the bathwater of the implied author.
They have criticized the cognitive theory of unreliable narration for overstating
the role of the reader at the expense of the author’s agency and the textual signals
of unreliability. Moreover, Phelan (2005: 48) has rightly pointed out that the radically
constructivist and cognitive conceptualization of unreliable narration fails to identify
the multiple constraints imposed not just by texts and conventions of reading but also
by those who design those texts, namely (implied) authors. The interpretive move to
read textual inconsistencies as a signal of unreliability after all does not make much
hermeneutic sense if it does not proceed from the assumption that someone designed
the inconsistency as a signal of unreliability (cf. Phelan 2005: 48).
In contrast to cognitive narratologists who seek to relocate unreliability only in the
interaction of reader and text (e.g., Nünning 1998, 1999, Zerweck 2001), Phelan
(2005: 38–49) has re-examined the concept and reconsidered the location of unreliability in light of the recent debates surrounding the implied author and unreliable
narration. Reminding us of Booth’s notion of continuity without identity between the
real and the implied author, he retains the implied author, but moves him or her
outside the text, thus re-establishing a closer link between the flesh and blood author
and the implied author (see also Booth’s essay in the present volume). Rejecting both
the reader-response version of an ‘‘inferred author’’ and the conflation of the implied
author with the text, Phelan (2005: 45) stresses the continuity that pertains between
the real author and his or her implied counterpart by redefining the latter as a
construction by and a partial representation of the real author, as ‘‘a streamlined version
of the real author, an actual or purported subset of the real author’s capacities, traits, attitudes,
beliefs, values, and other properties that play an active role in the construction of the particular
text’’ (italics in original). According to this account, the implied author is not a
product or structure of the text but rather the agent responsible for bringing the text
into existence. Phelan convincingly argues that the notion of unreliable narration
100
Ansgar F. Nünning
presupposes both a rhetorical view of narrative communication and the assumption
that authors fashion their texts in a particular way in order to communicate sharable
meanings, beliefs, attitudes, and values and norms.
Phelan and Martin as well as Greta Olson have reminded us that the different
models of unreliable narration all have ‘‘a tripartite structure that consists of (1) a
reader who recognizes a dichotomy between (2) the personalized narrator’s perceptions
and expressions and (3) those of the implied author (or the textual signals)’’ (Olson
2003: 93). Phelan’s rhetorical model in particular provides a timely reminder that
meaning arises from the recursive relations among authorial agency, textual phenomena, and reader response, and that not only readers but also authors draw on
conceptual and cultural schema: ‘‘But if readers need conceptual schema to construct
interpretations, authors also need conceptual schema to construct structural wholes’’
(Phelan 2005: 49). While acknowledging that the same textual phenomena can and
often will be construed in different ways by different readers, his rhetorical approach
is much better suited to accounting for the many ways in which readers might indeed
share understandings, values, and beliefs with authors and with each other, thus
opening up a useful way of exploring the ethical dimensions of narratives.
In contrast to a radically constructivist and cognitive theory of unreliable
narration, Phelan’s rhetorical and ethical approach to unreliable narration focuses on
the interplay between authorial agency, text-centered phenomena or signals, and
reader-centered elements in the reading process. This approach leads him to
argue that ‘‘while a text invites particular ethical responses through the signals it
sends to its authorial audience, our individual ethical responses will depend on the
interaction of those invitations with our own particular values and beliefs’’ (Phelan
and Martin 1999: 88–9). The concept of unreliable narration presupposes the existence of a constructive agent who builds into the text explicit signals and tacit
assumptions for the authorial or hypothetical ideal audience in order to draw readers’
attention to an unreliable narrator’s unwitting self-exposure or unintentional betrayal
of personal shortcomings. From the point of view of a rhetorical approach
to unreliable narration, Zerweck’s (2001: 156) thesis that the ‘‘unintentional selfincrimination of the personalized narrator is a necessary condition for unreliability’’
thus needs to be supplemented by the insight that the narrator’s unintentional
self-incrimination in turn presupposes an intentional act by some sort of higherlevel authorial agency, though it may be open to debate whether we should
attribute the constructive and intentional acts to ‘‘the implied author’’ or ‘‘the real
author.’’
A brief look at Ian McEwan’s macabre and grotesque short story ‘‘Dead As They
Come’’ (1978) may serve as a convenient example to show how the cognitive and
rhetorical approaches can be synthesized to solve many of the problems outlined
above and to shed more light on the questions faced by any critic doing interpretive
analysis: What textual and contextual signals suggest to the reader that the narrator’s
reliability may be suspect? How does an implied author (as redefined by Phelan)
Reconceptualizing Unreliable Narration
101
manage to furnish the narrator’s discourse and the text with clues that allow the critic to
recognize an unreliable narrator when he or she sees one? In short: how does one detect a
narrator’s unreliability? McEwan’s story is told by a 44-year-old, rich and egotistic mad
monologist (and misogynist) who, after three failed marriages, falls madly in love with a
‘‘fashionable woman’’ who turns out to be a dummy (store mannequin), which he
decides to buy and to call Helen. After a couple of months of what the narrator describes
as emotional and sexual bliss and ‘‘perfect harmony’’ (p. 71) he suddenly begins to
suspect that ‘‘Helen’’ is having an affair with his chauffeur Brian. What makes him
more and more suspicious is that ‘‘Helen was not listening at all’’ (p. 72), that ‘‘she
said nothing, absolutely nothing’’ (p. 73), and that what he believes to see when he
looks into her eyes is ‘‘quiet, naked contempt’’ (p. 76). The story reaches its horrible
climax, alluded to in the title, when in a frenzied fit of passionate madness the narrator
conceives ‘‘two savage and related desires. To rape and destroy her. [ . . . ] I came as she
died’’ (p. 76).
To begin with, while the narrator’s factual reliability, that is, his rendering of
the story, is only impaired by his highly idiosyncratic view of the world and his
deranged and disintegrating mind, the reader has plenty of reasons to suspect both the
nameless narrator’s commentary on and evaluation of the details of the events, and
the way in which the narrator reads and interprets, for example, what he deems to
be Helen’s feelings. The main reason for this is that the narrator violates both
many of the standards that today’s culture holds to be constitutive of
normal psychological behavior and widely accepted norms and values. Right
from the very beginning the implied author leaves the reader in no doubt that
the narrator’s view of the world is radically separated from any sane reader’s worldknowledge, which will immediately tell the reader that the narrator is merely
walking past a store window and that he is stopped in his tracks by a well-dressed
dummy:
I do not care for posturing women. But she struck me. I had to stop and look at her. The
legs were well apart, the right foot boldly advanced, the left trailing with studied
casualness. She held her right hand before her, almost touching the window [ . . . ] Head
well back, a faint smile, eyes half-closed with boredom or pleasure. I could not tell. Very
artificial the whole thing, but then I am not a simple man.
(McEwan 1979: 61–2)
The narrator’s explicit self-characterization includes a number of opaque statements
like ‘‘I am a man in a hurry’’ (pp. 61, 62) and ‘‘I am not a simple man’’ (p. 62) but he
also provides the reader with plenty of information about himself, unwittingly
exposing many of his personal shortcomings:
I must tell you something about myself. I am wealthy. Possibly there are ten men resident
in London with more money than I. Probably there are only five or six. Who cares? I am
rich and I made money on the telephone. I shall be forty-five on Christmas Day. I have
102
Ansgar F. Nünning
been married three times, each marriage lasting, in chronological order, eight, five and
two years. The last three years I have not been married and yet I have not been idle. I have
not paused. A man of forty-four has no time to pause. I am a man in a hurry.
(McEwan 1979: 62)
As in most other cases, the structure of unreliable narration underlying
McEwan’s story can be explained in terms of dramatic irony or discrepant awareness
because it involves a contrast between the narrator’s deranged view of the fictional
world and the divergent state of affairs which the reader can grasp. In the case of
McEwan’s unreliable narrator, dramatic irony results from the discrepancy between
the highly unusual intentions and questionable value system of the narrator
and the general world-knowledge, values, and norms of the average reader. For
the reader, both the internal lack of harmony between many of the statements and
acts of the narrator and contradictions between the narrator’s perspective and the
reader’s own concept of normality suggest that the narrator’s reliability is
indeed highly suspect. The reader interprets what the narrator says in two quite
different contexts. On the one hand, the reader is exposed to what the narrator wants
and means to say, that is, the narrator’s version of his tragic and fatal love story with
Helen. On the other hand, however, the statements of the narrator take on an
additional meaning for the reader, a meaning the narrator is not conscious of and
does not intend to convey. Without being aware of it, McEwan’s unreliable narrator
continually gives the reader indirect information about his idiosyncrasies and
deranged state of mind.
In addition to the peculiar characteristics, strange beliefs, and perverse behavior
explicitly attributed to the narrator in the text, the implied author has also endowed
the story with a wide range of signs and signals that invite the reader to make
inferences pertaining to the narrator beyond what is stated in the text. These
‘‘inference invitations’’ (Bortolussi and Dixon 2003: 80–1) include, for instance, the
bookkeeping manner in which the narrator reviews his marriages ‘‘in chronological
order,’’ the breathless and self-centered quality of his narratorial effusions, the excess
and incoherence of the information he provides, his disdain for others, and his
predilection for ‘‘silent women’’ (p. 63). Readers are thus invited to draw inferences
pertaining to the narrator and his questionable values, constructing him as a complete
egotist, misogynist, and monologist who has no respect for others, and who, as the
decreasing lengths of his marriages indicates, has apparently become increasingly
intolerable, and who is only interested in satisfying his own needs, interests, and
carnal pleasures.
It is thus not just the distance that separates the narrator’s highly idiosyncratic
view of the (fictional) world from the reader’s or critic’s world-knowledge, standards
of normalcy, and norms and values that indicates to the reader that the narrator is
highly unreliable, but also a wide range of textual features that serve as signals of
unreliability. Like many other texts featuring unreliable narrators, the narrative
of McEwan’s monologist is marked by a number of definable textual inconsistencies
Reconceptualizing Unreliable Narration
103
which function as clues to unreliability. Two of the most prominent of these
are internal contradictions within the narrator’s discourse and discrepancies between
his utterances and actions. McEwan’s narrator provides an especially amusing
example:
My ideal conversation is one which allows both participants to develop their thoughts
to their fullest extent, uninhibitedly, without endlessly defining and refining premises
and defending conclusions. . . . With Helen I could converse ideally, I could talk to her.
She sat quite still . . . Helen and I lived in perfect harmony which nothing could
disturb. I made money, I made love, I talked, Helen listened.
(McEwan 1979: 70–1)
The implied author has furnished the story with many other textual signals of the
narrator’s unreliability such as conflicts between story and discourse, between
the narrator’s representation of events and the explanations, evaluations, and interpretations of them that the narrator gives. In such cases as the description of the
scene in which the narrator actually buys ‘‘Helen,’’ his commentary ‘‘is at odds with
the evidence presented in the scene he comments upon’’ (Wall 1994: 25). The
reader or critic can establish such a difference by analyzing those utterances in
which the narrator’s subjective bias is particularly apparent and comparing the
worldview these imply with the story itself. In ‘‘Dead As They Come,’’ for example,
the narrator’s expressive statements such as subjective comments, evaluations, and
general remarks are completely at odds with the view of the events and characters that
is projected by such narrative modes as description, report and scenic presentation, as
well as by numerous small dramatic details. In his factually accurate report of how
he managed to get Helen, the narrator, for instance, mentions that the five female
salesclerks ‘‘avoided my eye’’ (p. 65) and that ‘‘[t]hey smiled, they glanced at
each other’’ (p. 65) after he has made his strange request to buy ‘‘the dummy
(ah my Helen)’’ (p. 65), but he completely fails to interpret correctly why they are
doing this.
In addition to such internal contradictions the implied author (once again as
redefined by Phelan) has carefully equipped the narrator with idiosyncratic verbal
habits which also serve as clues to unreliability. The narrator’s stylistic peculiarities
and his violation of linguistic norms and of Grice’s conversational postulates play an
important role in detecting the narrator’s unreliability. There are, for instance,
pragmatic indications of unreliability such as frequent occurrences of speaker-oriented
and addressee-oriented expressions. One does not need to take a word-count or employ
ponderous statistical methods to show that the unreliable narrator of McEwan’s story
as well as those of Martin Amis’s Money or Julian Barnes’s Talking It Over are
compulsive monologists as well as egotists. The vast majority of their utterances are
indeed speaker-oriented expressions beginning with their favorite word, ‘‘I.’’ Similarly, it is virtually impossible not to notice the plethora of addressee-oriented
expressions that these and many other unreliable narrators tend to use. There
104
Ansgar F. Nünning
are also syntactic indications of unreliability such as incomplete sentences, exclamations, interjections, hesitations, and unmotivated repetition. McEwan’s ‘‘Dead as
They Come’’ is full of them, and so are Patrick McGrath’s novels. One could also
mention such lexical indications of unreliability like evaluative modifiers, expressive
intensifiers, and adjectives that express the narrator’s attitudes, all of which feature
prominently in McEwan’s short stories and McGrath’s novels. All of these stylistic
expressions of subjectivity indicate a high degree of emotional involvement and they
provide clues for the reader to process the narrator as unreliable along the axis of facts/
events, the axis of ethics/evaluation, and/or the axis of knowledge/perception (see
Phelan and Martin 1999).
As these examples may serve to show, the projection of an unreliable narrator does
not hinge upon the reader’s frames of reference or on conventions of reading alone, as
cognitive approaches suggest, because texts and those who design those texts, namely
(implied) authors, impose multiple constraints on the ways in which narrators are
processed. Rhetorical approaches to narrative remind us that the projection of an
unreliable narrator, far from being hit or miss, presupposes the existence of a creative
agent who furnishes the text and the narrator with a wide range of explicit signals and
inference invitations in order to draw readers’ attention to a narrator’s unwitting selfexposure and unreliability.
Conclusion and Suggestions for Further Research
To sum up: by synthesizing concepts and ideas from both cognitive and rhetorical
approaches, this essay has attempted to advance our understanding of unreliable
narration and of how readers negotiate and process texts featuring an unreliable narrator, creating a somewhat more detailed (though by no means exhaustive) inventory
of the presuppositions, frames of reference, and textual signals involved in the
projection of unreliable narrators. If the rhetorical approach with its emphasis on
the recursive relations among (implied) author, textual phenomena or signals, and
reader response encompasses the cognitive narratologist’s emphasis just on reader and
text, then the cognitive approach can nevertheless provide more finely nuanced tools
for recognizing an unreliable narrator. Though the suggested synthesis of the two
approaches still leaves several questions unanswered (e.g., what is the respective
degree of importance of the various items in the inventory outlined above?), it can
arguably yield new insights into unreliable narration and open up productive avenues
of inquiry for narrative theory, the more so because it is just as relevant for the ways in
which, for example, literary characters, events, and plots are constructed (by implied
authors) and processed (by readers) and for the role conceptual schema play on the
production and reception side.
Though agreement has been reached that ascriptions of unreliability involve the
recursive relationship among the author, whether implied or not, textual phenomena,
and reader response, accounts of unreliable narration still differ significantly with
Reconceptualizing Unreliable Narration
105
regard to the respective degree of importance they attribute to each of these three
factors. While cognitive narratologists single out reader response and the cultural
frameworks that readers bring to texts as the most important basis for detecting
unreliability, narrative theorists working in the tradition of rhetorical approaches to
narrative have redressed the balance. Most theorists agree, however, that to determine
a narrator’s unreliability one need not rely merely on intuitive judgments, because a
broad range of definable signals provides clues to gauging a narrator’s unreliability.
These include both textual data and the reader’s pre-existing conceptual knowledge of
the world and standards of normality. In the end it is both the structure and norms
established by the respective work itself and designed by an authorial agency, and the
reader’s knowledge, psychological disposition, and system of norms and values that
provide the ultimate guidelines for deciding whether a narrator is judged to be
reliable or not.
The suggested synthesis of cognitive and rhetorical approaches is chiefly offered as a
means to rethink, and to stimulate further debate on, the intricate problem of
explaining how readers and critics intuitively consider narrators to be instances of
unreliable narration. Much more work, however, needs to be done if we want to come
to terms with the complex set of narrative strategies that ever since Booth’s early work
have been subsumed under the wide umbrella of the term ‘‘unreliable narration.’’
There are at least six important areas that have yet to be adequately explored. One of
them is the development of an exhaustive and full-fledged theory of unreliable
narration integrating the insights recently provided by cognitive and rhetorical
narrative theorists. Second, what is needed is a more subtle and systematic account
of the clues to unreliable narration, including more sophisticated analyses of the
interplay between textual data and interpretive choices. Third, the different uses of
the unreliable narrator in the works of both contemporary novelists and authors from
earlier periods, and the ways in which they reflect or respond to changing cultural
discourses, are just waiting to be explored. Fourth, the history of the development of
the narrative technique known as ‘‘unreliable narration’’ has yet to be written because
no one has dared to provide an historical overview spanning the period from the
eighteenth century to the twentieth (for brief sketches see Nünning 1997a, Zerweck
2001). Fifth, since the generic scope of unreliable narration has as yet neither been
properly defined nor even gauged, unreliability across different genres, media, and
disciplines provides a highly fertile area of research. Notwithstanding a small number
of articles on the subject (see Bennett 1987, Richardson 1988), the use of the
unreliable narrator in genres other than narrative fiction – for instance in dramatic
genres like the memory play or in the dramatic monologue – as well as in other media
and domains (including law and politics) deserves more attention than it has hitherto
been given. Lastly, taking a new look at the development of narrative techniques like
unreliable narration and of the history of the reception of individual unreliable
narrators (see V. Nünning 2004) could be an important force in the current attempts
to historicize narrative theory.
106
Ansgar F. Nünning
REFERENCES
AND
Bennett, J. R. (1987). ‘‘Inconscience: Henry James
and the Unreliable Speaker of the Dramatic
Monologue.’’ Forum 28, 74–84.
Booth, W. C. (1961). The Rhetoric of Fiction. Chicago: University of Chicago Press.
Booth, W. C. (1974). A Rhetoric of Irony. Chicago:
University of Chicago Press.
Bortolussi, M. and Dixon, P. (2003). Psychonarratology: Foundations for the Empirical Study of
Literary Response. Cambridge, UK: Cambridge
University Press.
Chatman, S. (1978). Story and Discourse: Narrative
in Fiction and Film. Ithaca, NY and London:
Cornell University Press.
Chatman, S. (1990). Coming to Terms: The Rhetoric
of Narrative in Fiction and Film. Ithaca, NY and
London: Cornell University Press.
Culler, J. (1975). Structuralist Poetics: Structuralism,
Linguistics and the Study of Literature. London:
Routledge & Kegan Paul.
Fludernik, M. (1993). The Fictions of Language and the
Languages of Fiction: The Linguistic Representation of
Speech and Consciousness. London: Routledge.
Fludernik, M. (1999). ‘‘Defining (In)Sanity: The
Narrator of The Yellow Wallpaper and the Question of Unreliability.’’ In W. Grünzweig and
A. Solbach (eds.), Grenzüberschreitungen: Narratologie im Kontext/Transcending Boundaries: Narratology in Context (pp. 75–95). Tübingen: Narr.
Lanser, S. S. (1981). The Narrative Act: Point of
View in Prose Fiction. Princeton, NJ: Princeton
University Press.
McEwan, I. (1979 [1978]). ‘‘Dead As They
Come.’’ In In Between the Sheets (pp. 61–77).
London: Pan Books.
Nünning, A. (1997a). ‘‘‘But Why Will You Say
That I Am Mad?’ On the Theory, History, and
Signals of Unreliable Narration in British Fiction.’’ Arbeiten aus Anglistik und Amerikanistik
22, 83–105.
Nünning, A. (1997b). ‘‘Deconstructing and
Reconceptualizing the ‘Implied Author’:
The Resurrection of an Anthropomorphicized
Passepartout or the Obituary of a Critical
‘Phantom?’.’’ Anglistik. Mitteilungen des Verbandes Deutscher Anglisten 8(2), 95–116.
FURTHER READING
Nünning, A. (ed.) (1998). Unreliable Narration:
Studien zur Theorie und Praxis unglaubwürdigen
Erzählens in der englischsprachigen Erzählliteratur.
[Unreliable Narration: Studies in the Theory and
Practice of Unreliable Narration in English Narrative Fiction]. Trier, Germany: Wissenschaftlicher
Verlag Trier.
Nünning, A. (1999). ‘‘ ‘Unreliable, Compared
to What? Towards a Cognitive Theory of
Unreliable Narration: Prolegomena and
Hypotheses.’’ In W. Grünzweig and A. Solbach
(eds.), Grenzüberschreitungen: Narratologie im
Kontext/Transcending Boundaries: Narratology in
Context (pp. 53–73). Tübingen: Narr.
Nünning, V. ([1998] 2004). ‘‘Unreliable Narration
and the Historical Variability of Values and
Norms: The Vicar of Wakefield as Test-case for a
Cultural-Historical Narratology.’’ Style 38.
Olson, G. (2003). ‘‘Reconsidering Unreliability:
Fallible and Untrustworthy Narrators.’’ Narrative 11, 93–109.
Phelan, J. (1996). Narrative as Rhetoric: Technique,
Audiences, Ethics, Ideology. Columbus: Ohio State
University Press.
Phelan, J. (2005). Living To Tell About It: A
Rhetoric and Ethics of Character Narration.
Ithaca, NY and London: Cornell University
Press.
Phelan, J. and Martin, M. P. (1999). ‘‘ ‘The Lessons
of Weymouth’: Homodiegesis, Unreliability,
Ethics and The Remains of the Day.’’ In
D. Herman (ed.), Narratologies: New Perspectives
on Narrative Analysis (pp. 88–109). Columbus:
Ohio State University Press.
Prince, G. (1987). A Dictionary of Narratology.
Lincoln: University of Nebraska Press.
Richardson, B. (1988). ‘‘Point of View in Drama:
Diegetic Monologue, Unreliable Narrators, and
the Author’s Voice on Stage.’’ Comparative
Drama 22, 193–214.
Riggan, W. (1981). Picaros, Madmen, Naifs,
and Clowns: The Unreliable First-Person Narrator.
Norman: University of Oklahoma Press.
Rimmon-Kenan, S. ([1983] 2003). Narrative Fiction: Contemporary Poetics. London, New York:
Methuen.
Reconceptualizing Unreliable Narration
Wall, K. (1994). ‘‘The Remains of the Day
and its Challenges to Theories of Unreliable Narration.’’ Journal of Narrative Technique 24, 18–42.
Yacobi, T. (1981). ‘‘Fictional Reliability as a Communicative Problem.’’ Poetics Today 2, 113–26.
Yacobi, T. (1987). ‘‘Narrative and Normative Patterns: On Interpreting Fiction.’’ Journal of Literary Studies 3(2), 18–41.
107
Yacobi, T. (2001). ‘‘Package Deals in Fictional
Narrative: The Case of the Narrator’s (Un)reliability.’’ Narrative 9, 223–9.
Zerweck, B. (2001). ‘‘Historicizing Unreliable
Narration: Unreliability and Cultural Discourse
in Narrative Fiction.’’ Style 35, 151–78.
6
Authorial Rhetoric, Narratorial
(Un)Reliability, Divergent
Readings: Tolstoy’s Kreutzer
Sonata
Tamar Yacobi
My subject brings together a famous story, an ongoing debate on its interpretation,
and a theoretical common denominator that, I would argue, underlies its divergent
readings as alternative ways to resolution. The combination of theory, history, and
reading, especially in the service of accounting for disagreements about the narrator’s
reliability, demonstrates the explanatory power of the Tel-Aviv school of narratology
to which I subscribe. I’ll begin with an overview of the critical controversy triggered
by the Kreutzer Sonata and of the proposed theoretical framework. The greater part of
the essay will show how the various readings of the tale, even where incompatible, are
all embedded and explicable within the framework of this theory.
Tolstoy’s The Kreutzer Sonata (first published 1891) is the story of Pozdnyshev, a
Russian nobleman who has murdered his wife because he suspected her of adultery.
After his trial and acquittal, he tells it all to a chance fellow traveler on a train. His
account is anything but typical. Instead of claiming, say, temporary insanity, he
admits premeditation and minutely details the murder scene. Moreover, unlike his
murderous jealousy at the time, the teller now pities his dead wife and represents
them both as equally helpless victims of the socioeconomic system. In self-defense,
perhaps, he mounts a direct attack on it, with a radical and extraordinary reform as a
solution: since sexual intercourse generates all social evils, it should be abolished, even
between husband and wife. The consequent extinction of the human race does not
deter him. Both tale and reform plan are set within the framing written report of
Pozdnyshev’s original addressee and traveling companion.
The controversy over The Kreutzer Sonata started with its first readers, who either
applauded or denounced its radical message. The censor banned it but couldn’t
prevent handwritten copies from being circulated all over Russia. Peter Ulf Møller,
Tolstoy’s Kreutzer Sonata
109
in his meticulous survey of the early reception, also delineates the parties to the
debate: conservative vs. liberal critics; religious as against secular thinkers; abused
wives, who embraced, and feminists, who repudiated, the doctrine at issue; or those
who agreed with Tolstoy about the problem of sexual life but not with his solution, as
opposed to those who disputed his very premises (Møller 1988: 39–162). In many
cases, the divergence among readers on the author’s ideology came together with their
views of the speaker’s reliability: ‘‘a typical modern husband or a particularly debauched deviant,’’ and ‘‘intended as a character whose misapprehensions were to be
exposed or as the author’s spokesman’’? (Møller 1988: 134–5).
A hundred years later, the debate goes on, and with it, in effect, the quarrel over
Pozdnyshev’s mediation between author and reader. The sheer wonder at his incongruity, even counterproductiveness in this role, would appear inexhaustible. Among
most critics, as one slavicist generalizes, Tolstoy’s ‘‘views . . . on chastity are said to be
presented so unsubtly as to be beyond explication, while the rhetorical strategy of
placing them in the mouth of a preachy and unrepentant murderer is held to be
unfathomable’’ (Herman 1997: 16, n.4). As we’ll find below, commentators roughly
divide into three basic positions: (1) those who take the narrator as Tolstoy’s reliable
spokesman; (2) others who assert his unreliability; and (3) still others who locate the
problem not (or not only) in the narrator’s but in the author’s own unacceptable
ideology and/or confused discourse. The first two views, which judge the narrator by
his (dis)harmony with the author, are symmetrical and in principle mutually exclusive; the third complicates the division in preferring or proceeding to judge the
author himself as thinker, artist, persuader, and always, it so happens, to negative
effect at that. How this third position subdivides, relative to the first two, inter alia,
will appear below.
It is not my intention, or the task of narratology, to resolve ‘‘once and for all’’ the
crux of The Kreutzer Sonata. Instead, I want to show that the diverse interpretive
positions are best mapped and correlated as alternative hypotheses under the umbrella
of a proper theory of interpretation, wide-ranging yet specifically narrative-oriented.
Most wanted here is a theory that can bridge the apparent gulf between readings that
explain the story by pointing to the narrator’s unreliability and those that cast the
blame on the author. To anticipate my fundamental thesis: even contrasted readings of
speaker and author, speaker vis-à-vis author, operate, knowingly or otherwise, by the
same deeper principle, namely, explaining ostensible discordance by reference to an
appropriate integration mechanism.
‘‘Unreliability’’ famously comes from Wayne Booth, who defines it as the narrator’s
distance from the implied authorial norms of the text (1961: 158–9). Booth elaborates on the ‘‘variation’’ in both ‘‘distance’’ and ‘‘norms,’’ suggesting the flexibility of
our conception of a speaker’s (un)reliability. But why, how, and where does a ‘‘distance’’ arise, or fail to arise, between a speaker and the implied author?
Hence my operational theory, influenced by Meir Sternberg’s idea of fictional
discourse as a complex act of communication (e.g., 1978: 254–305, 1983) to be
motivated or otherwise integrated. I have accordingly long defined unreliability as a
110
Tamar Yacobi
reading-hypothesis: one that is formed in order to resolve textual problems (from
unaccountable detail to self-contradiction) at the expense of some mediating, perceiving, or communicating agent – particularly the global speaker – at odds with the
author (e.g., Yacobi 1981, 1987, 2001). Let me briefly re-emphasize the hypothetical
nature of such an interpretive move, which, like any conjecture, is open to adjustment, inversion, or even replacement by another hypothesis altogether. Fictional
unreliability is not a character trait attaching to the (probabilistic) portrait of the
narrator but a feature ascribed (or lifted) ad-hoc on a relational basis, depending on
the (equally hypothetical) norms operative in context. What is deemed ‘‘reliable’’ in
one context, including reading-context, as well as authorial and generic framework,
may turn out to be unreliable in another, or even explained outside the sphere of a
narrator’s failings.
By ordinary standards, for instance, the teller of Bashevis-Singer’s ‘‘Gimpel the
Fool’’ may indeed look like the fool that his neighbors see and trick; yet his apparent
folly is readable in context as genuine innocence, hence as the very merit that qualifies
Gimpel to speak for his creator’s deeper norms. A mirror image of this hard-won
reliability would be the suspicion that Thackeray’s showman of Vanity Fair lapses at
times into inconsistencies that betray an untrustworthy judgment (on this comparison, see Yacobi 2001, with further references). Likewise with tensions in matters of
fact, rather than ethics or ideology. A paradigm of suppressive omniscient narration,
Fielding’s Tom Jones thus goes to show how a narrator’s deliberate concealment of the
truth – false statements included – may yet be accordant with his reliability:
explicable as an authorized tactic by appeal to the text’s rules and goals (Stenberg
1978: 248, 265–8, 1983: 172ff., 2001: esp. 150ff.).
In short, there is no automatic linkage between textual incongruities and narratorial unreliability. The less so because the same perceived tensions, difficulties, incompatibilities – down to linguistic oddities – are always open to alternative principles or
mechanisms of integration. Among them, my earlier work has singled out the
existential, the functional, the generic, and the genetic mechanisms as rivals to the
perspectival way, specifically to (un)reliability judgments. The fact that all five are
relevant to The Kreutzer Sonata debate (though in varying degrees) qualifies the story
for a test case of my theoretical framework. Let me briefly outline the set of
mechanisms as they operate on this ground to diverse interpretive effect.
1 The existential mechanism refers incongruities to the level of the fictive world,
notably to canons of probability that deviate from those of reality. The worlds of
fairy tale, science fiction, or Kafka’s ‘‘Metamorphosis’’ are extreme examples. In our
novella, part of the debate concerns the applicability of Pozdnyshev’s generalizations on issues as significant as love, education, sexual morality, and women’s
emancipation. Pozdnyshev asserts throughout that his marital crisis is typical,
with implications for our judgment of him. To the extent that the claim holds
true of the world, his tale reliably embodies a widespread problem in need of
Tolstoy’s Kreutzer Sonata
2
3
4
5
111
solution; if, on the other hand, his crisis is unique, let alone morbid, his insistence
on its prevalence betrays unreliability.
The functional mechanism imposes order on the deviant in terms of the ends
requiring or justifying that deviance. Whatever looks odd – about the characters,
the ideas, the structure – can be motivated by the work’s purpose, local or overall,
literary or otherwise. Here, Tolstoy’s religious worldview and iconoclastic theory of
art may explain not just the text’s normative hierarchy but the problematic manner
of its representation. Apropos the Sonata, Tolstoy emphasizes the low status of
‘‘artistry,’’ to which he has ‘‘ ‘only given . . . just enough room for the terrible truth
to become visible’ ’’ (quoted in Møller 1988: 10). Such a defiant admission might
warn us against expecting a high norm of unity here. For Tolstoy, Pozdnyshev is a
useful mouthpiece as long as he delivers ‘‘the terrible truth.’’ Seemingly apart from
(un)reliability, then, functionality, like existence (and, next, genre), interacts with it.
The generic principle appeals to a certain encoded model or simplification of reality,
like the causal freedoms of comedy vis-à-vis the stricter tragic plot. Here, J. M.
Coetzee analyzes our tale within the generic frame of secular confession, defined as ‘‘a
mode of autobiographical writing’’ with ‘‘an underlying motive to tell an essential
truth about the self’’ (Coetzee 1985: 194). Coetzee accordingly juxtaposes Pozdnyshev’s claim to tell the truth with the telling’s numerous incongruities, wondering
in turn how reliable his confession is, how persuasive his conversion. The generic
context and ‘‘motive’’ possibly explain away such internal tensions, or, failing to do
so, raise the question of whose responsibility it is: the confessing agent’s or his
creator’s?
The perspectival or unreliability principle enables The Kreutzer Sonata reader to
explain a variety of incongruities – in matters of fact, action, logic, value, aesthetics
– as symptoms of narrator/author discord: clues to Pozdnyshev’s unknowing
misrepresentation of Tolstoy, Tolstoy’s organized exposure of Pozdnyshev. Such a
reading, insofar as it appeals to the teller’s unreliability, presupposes a further
hypothesis about the implied norms that not only govern fact, action, and so forth,
but also determine the choice of this kind of teller.
Finally, the genetic mechanism relegates fictive oddities and inconsistencies to the
production of the text; above all, where unresolved otherwise, they are blamed on
the (e.g., wavering, negligent, or ideologically fanatic) author. This explanation
most differs from all others. When incongruities are shifted from any of the four
discourse contexts to the context of the creator, they become more intelligible, but
not necessarily more acceptable. Just consider ‘‘one inconsistency [that] seems like a
slip on Tolstoy’s part: in chap. 14, Pozdnysheva is said to have borne six children,
while elsewhere it is five’’ (Isenberg 1993: 167, n.29). On the other hand, evidence
of an author’s writing process (diaries, letters), as well as related genetic circumstances, may support other integration mechanisms to favor one interpretive
hypothesis over its alternative(s). Here, too, the well-documented genesis of the
story, including its nine versions and ‘‘Sequel,’’ will be found to play a large role.
112
Tamar Yacobi
Again, like the perspectival mechanism, the genetic explains textual oddities at the
expense of someone associated with the story as told. The difference in the identity of
the responsible party, however, opposes the descriptive concept of unreliable narrator
to the sheer evaluative implications of the inept or offensive or self-contradicting
author. A genetic explanation thus often imputes some loss of control – whether over
a minor detail like ‘‘how many children had Mrs Pozdnyshev,’’ or over significant
issues such as character, plot, and ideology. Yet, though arising from and judged as
incongruity, the consequent irony – now directed against the author himself – is
unintended on every level of the text. By contrast, to hypothesize a fallible narrator
is to assume the ironic mastery of a deliberate communicator behind the scenes. On
such a reading, the fictive message divides between the narrator’s surface oddities and
the underlying coherence of the implied communication that we hypothesize as both
their source and point.
The perspectival and the genetic mechanisms are also the most relevant to the
explanation of the Kreutzer Sonata debate. The others, particularly the existential
and the functional, serve there to support or reject or complement the main
readings. Generally, the first set of readings (Pozdnyshev is reliable) all imply that
there are no significant inner tensions in the telling as such, and those that
still remain should be integrated between the existential mechanism (his problems
represent a common human condition) and the functional principle (apparent discordances are effective in context). For if the narrator is indeed the author’s spokesman, then all his choices, selections, and combinations must somehow operate for
reliability. On the other hand, readings that reject Pozdnyshev’s viewpoint emphasize
the incongruities, but disagree about the responsible agent: the teller (within
a perspectival or unreliability mechanism) and/or his creator (a genetic solution).
These two models of integration produce the second and third sets of the story’s
readings.
The Selling of a Madman: Rhetoric in the Service of Reliability
First, to judge Pozdnyshev reliable means that all significant inner tensions in his
discourse have been contrived to promote the author’s like-minded thematic and
rhetorical goals. In turn, such promotion against, amid, and through tension, demands an authorial rhetoric in the promoter’s favor. Every reliability hypothesis
entails persuasion, but selling the end to sexual intercourse, from the mouth of a
wife killer, too, escalates the need along with the difficulty. What, then, are the
rhetorical countermeasures, and how do the various integration mechanisms work or
substitute for this thrust?
In text order, the first rhetorical step is a multiform extension of the normative crux
and thesis beyond the idée-fixe of one individual, who himself claims their extension.
Even before Pozdnyshev’s monologue, two devices work to this validating effect – the
epigraph and the opening scene (compare Isenberg 1993: 80–2).
Tolstoy’s Kreutzer Sonata
113
The epigraph, like the title, has been chosen by an independent party (the
anonymous frame-narrator, and/or the author). Moreover, Tolstoy appeals here to a
high authority, quoting two of Christ’s sayings on sexual morality (Matthew 5: 28,
19: 10–12): the second explicitly recommends celibacy ‘‘for the sake of the kingdom
of heaven.’’ Christ’s approval of the hero’s thesis endows it in advance with ideological
weight and prestige as well as temporal depth. Pozdnyshev himself duly (re)quotes
and explicates one of these verses (Tolstoy 1963: 313) to suit his new ideology. And as
Møller notes, ‘‘in the eighth draft version,’’ he also ‘‘concludes . . . by repeating his
interpretation of Matthew 5: 28’’ (Møller 1988: 28; Tolstoy 1960: 448–9, n.100). In
further genetic support for the commonality, hence reliability of the thesis, the same
Matthew quotes recur in Tolstoy’s ‘‘Sequel to the Kreutzer Sonata’’ (1929: 163).
Next, the long opening dialogue scene brings out the topical relevance of the
problem. Although the murderer’s lecture dominates the story, it is set within a
framing tale and presented as a response to an exchange of views by an apparently
random group of passengers. Yet the group’s ‘‘choice’’ and ‘‘discussion . . . are by no
means fortuitous’’ (Møller 1988: 20; also Isenberg 1993: 86–9). There arise issues like
the frequency of divorce, women’s rights and education, the institution of marriage,
the deterioration of family ethics. The crisis becomes manifest in the growing
abandonment of the orthodox code of behavior for imported new ideas, and in the
spread of marital conflict to remote villages. Moreover, the two chief debaters, a
conservative old merchant and an emancipated lady, ‘‘represent . . . two fairly easily
datable stages in the history of sexual morality in Russia . . . but both still very much
present as poles in the range of contemporary attitudes’’ (Møller 1988: 23). Hence the
advance support of Pozdnyshev’s claim that his predicament is not exceptional but a
symptom of what was known as ‘‘the woman question’’ (Mandelker 1993: 21–30). As
the opening contestants all agree that marriage has become an insoluble problem for
all classes, they enable the author to expand the scope even beyond the fictive to the
real Russian world of his time. Who would disagree? Mirroring an actual state of
affairs, the ‘‘existential mechanism’’ supports the speaker’s claim for generality.1
While dramatizing the issue’s scope, the dialogue also looks ahead to the extraordinary value-judgment to be made. First, via the negation of rival worldviews. Those
attacked by Pozdnyshev (and his creator, on this reading) are already espoused by such
speakers as defeat their own claims. On the side of modernity, education for women,
and marriage for love, we find the lady whose ‘‘mannish’’ appearance bespeaks in
contemporary context an object of ridicule, or mixed response, while her illogical
interventions are supposed to misfire even outside the fictive debate. Her conservative
antagonist is an old merchant whose practice (he has taken part in an orgy) contradicts
his preaching traditional religion and a hard line with promiscuous females. Genetically, moreover, Tolstoy transformed him from a decent man in the earlier versions to
a hypocrite (Møller 1988: 20–3).
Finally, the dispute is resolved dramatically. The merchant evades the anonymous
narrator’s clear allusion to his sexual practice; his adversary, the lady, under Pozdnyshev’s provocative questions, likewise retires without an answer. The failure of the two
114
Tamar Yacobi
‘‘representative’’ voices infects their respective ideologies. ‘‘Unfair’’ rhetoric, perhaps,
but integral – or at least available – to fiction. The resemblance to a real-life
controversy is only illusive, the realistic facade of a fixed fight. In life, interlocutors
are directly responsible for their attitudes and arguments, for their power relation and
the resolution. Here, as a fictive character, Pozdnyshev confronts the rival ideologues
and ideologies in what is, for him, the reality of his life. But it is the implied author
who – for his thematic and rhetorical purposes – reduces these antagonists to ‘‘straw
people,’’ confused or hypocritical, and accordingly authorizes his own spokesman’s
line of fundamental reform (cf. Isenberg 1993: 83, 87, 88).
Those choices are doubtless functional. Tolstoy could have created an intelligent
female antagonist, or an attractive and happily married one, like Natasha Rostov at
the end of War and Peace. Similarly, the religious ‘‘delegate’’ could be modeled on
Levin from Anna Karenina, and so would practice what he teaches (as indeed his earlier
versions do). In short, by leading the group discussion to the marriage crisis and by
first giving the floor to such inferior doctrinal exponents, Tolstoy would persuade us
of the need for reform. Narratively, at least, he both raises our interest in a suitable
new doctrine and motivates Pozdnyshev’s response.
Not satisfied, however, with the rhetoric of negation, Tolstoy multiplies positive
devices. On the plot level, Pozdnyshev’s ‘‘marriage functions paradoxically as an
example of a completely normal marriage’’ (Møller 1988: 12). Its normalcy (e.g., the
trivial causes and the frequency of marital conflict) ensures its existential scope or
generality; its tragic ending is supposed to act as a warning and an incentive for a
radical change. In between, Tolstoy needed an exciting force strong enough to push
the couple from the normal hell of marriage to an explosion beyond the pale. And
among all forces, he picked Beethoven’s Sonata, ‘‘the titillating, exciting’’ catalyst ‘‘of
ethically unengaged art’’ (Møller 1988: 16–17), to trigger the husband’s murderous
fury.
On the level of the monologist’s own discourse, rhetorical figures abound: appeal to
statistics, rhetorical questions, anaphoric repetitions, analogies between sexual and
other addictions. Pozdnyshev’s shock tactics include the deautomatizings of family
life: love assimilated to lust, or wives equated with whores as kept women. In
marriage, ‘‘an innocent girl is sold to a dissolute man and the sale is attended by
fitting rites’’ (Tolstoy 1963: 309). Or observe how the speaker’s emotional upheaval
goes with his reasoned argument and chronological tale (Møller 1988: 36). Tolstoy
needs, and values, both extremes: the marks of Pozdnyshev’s sincerity and the clear
narrativized reasoning. How to join such persuasive forces? Their incongruity might
undermine the reliability hypothesis: on its own, Pozdnyshev’s monologue could look
like a mixture of a religious tract and nervous tics (Møller 1988: 35–7).
Hence the crucial rhetorical function that is assigned to Pozdnyshev’s inside
addressee and future recorder. That he is impressed by Pozdnyshev’s sincerity mediates
it much better than would any self-declaration. Likewise, his quoting of the long
monologue foregrounds (e.g., via the added chapter divisions) its (chrono)logical
order, while his comments on the original way of speaking ‘‘document Pozdnyšev’s
Tolstoy’s Kreutzer Sonata
115
agitation’’ (Møller 1988: 36–7). Further, as is widely agreed, the anonymous narrator
serves as the reader’s delegate within the fictive world: the global framing communicator (the implied author, ‘‘Tolstoy’’) persuades his addressee (the implied reader) by
dramatizing the process along which the inset speaker (Pozdnyshev) affects his
auditor. The latter thus serves as ‘‘a kind of Directions for Use attached to the
work’’ (Møller 1988: 38). To enhance this symmetrical device, the dramatized
addressee, unlike the opening disputants, is a reasonable person. Neither quick to
change his mind nor deaf to new ideas, he qualifies as our deputy.
This rhetorical process accordingly falls into two parts. During the first,
Pozdnyshev’s addressee frequently interrupts his amazing monologue with questions,
objections, and surprised exclamations.
‘‘What do you mean by the domination of women?’’ I said. ‘‘The law gives the advantage
to man.’’
(Tolstoy 1963: 306)
‘‘Vice, you say?’’ I put in. ‘‘But you are speaking of the most natural human function.’’ . . .
‘‘But how,’’ said I, ‘‘is the human race to be perpetuated?’’
(p. 310)
I found his ideas new and shocking.
‘‘But what is to be done?’’ I said. ‘‘If what you say is true, then a man can make love to
his wife only once in two years; but men –’’
(p. 318)
Such responses verbalize the reader’s expected wonder and possible reservations about
Pozdnyshev’s value system and no-sex thesis. Consistently built up throughout the
opening half (chaps. 3–15), the listener’s dependability as persuasively modulates in
the closing part (chaps. 16–28), when his responses undergo a change. He stops
asking questions, yet continues to express his interest in the new ideas, even when
addressing himself to the reader. At one point, when left alone in the carriage – thus
the narrator about his experiencing self – ‘‘I’’ got ‘‘so lost’’ in ‘‘going over in my mind
all he had said . . . that I did not notice him come back’’; at parting, he is even moved
‘‘to tears’’ (Tolstoy 1963: 330–1, 369).
Moreover, unlike the satiric treatment of the inside speakers in the opening scene,
the author takes care to moderate the addressee’s rhetorical role. Hence, as in the above
quote, we never know if he was actually converted by Pozdnyshev or has only become
convinced that the severity of the marriage crisis and the uselessness of other solutions
at least give the extraordinary proposal a claim to a wider public hearing.2 So even a
skeptical reader should find the lower limit hard to resist and even pushed up toward
assent by the rest of the author’s rhetoric, mediated as well as (like the epigraph)
wholly implicit.
116
Tamar Yacobi
Thus, reconsider the story’s communicative structure. As the official narrator who
quotes and frames and publishes the murderer’s tale, the former listener to it is
responsible for the selection and combination of the given text. This position lends
greater authority to, inter alia, his own quoted responses. Were Pozdnyshev the global
teller, he might have mistaken them for conversion or even fabricated them for
persuasion by example. But as it is, we take these pregnant responses at face value
and as a guide to our own in authorial reading. Inversely with the turning of the
unreliability mechanism on others to support the counterspeaker’s reliability. It is not
Pozdnyshev, the would-be reformer, but the narrator himself who chose to juxtapose
the early contestants’ failures (of appeal, logic, consistency, honesty, etc.) with the
protagonist’s viewpoint.
Finally, Tolstoy’s own ‘‘Afterword’’ or ‘‘Sequel’’ leaves no doubt concerning his likeminded authorial intent. It begins with five points on sexual relations and their
outcome, which condense Pozdnyshev’s argument. Then, even more radically, it
explicates Christ’s teaching on sexual morality (notably the epigraph verses) counter
to the rulings of the Church. As Ernest Simmons (1960: 128) sums up, the ‘‘Afterword’’ displays the basic agreement between creator and creature, except that Tolstoy’s
‘‘idealistic but logically developed thought’’ varies from ‘‘the extravagant conviction
of the deranged Pozdnyshev.’’ In effect, Simmons qualifies the spokesman’s reliability
on the level of presentation alone.
The Sequel’s genetic evidence is reinforced by further data about the underlying
drive. The story’s eight earlier versions and the work in progress on What Is Art?, as
well as the writer’s diary and correspondence, show how Tolstoy gradually shaped his
unconventional ideology along with the communicative and rhetorical strategy
(Møller 1988: 1–38, 181–99). Such evidence would appear to confirm what the
finished novella implies: that the reliability hypothesis best explains an array of
puzzling textual features. Why the New Testament epigraph? Why the debate
prior to Pozdnyshev’s monologue, including the given debaters, traits, ideologies,
ironies, undignified withdrawal? What is the function of Beethoven’s Sonata? Why
the inside addressee, complete with his early wonder and changing attitude?
Why assign the overall narration to this anonymous figure rather than to Pozdnyshev?
And, of course, why did Tolstoy, contrary to his habit, append a ‘‘Sequel,’’ and to this
like-minded effect?
So the hypothesis that the author endows a deviant speaker with reliability looks
the likeliest in context because it provides the fullest and most interconnected
answers: even the oddest parts make joint sense within a rhetorical strategy devised
to create a bridge between Tolstoy’s ideological representative and the presumable
beliefs of the common reader.3
Yet the controversy over The Kreutzer Sonata still goes on; and it will, I suspect,
persist in face of the case made here for reliability by appeal to overall difficult
coherence. J. M. Coetzee comments on the text’s ‘‘lack of armament against other,
unauthorized readings, other truths’’ (1985: 204). Reviewing ‘‘the critical tradition,’’
Isenberg concludes that ‘‘from a narratological perspective, the most remarkable thing
Tolstoy’s Kreutzer Sonata
117
about ‘The Kreutzer Sonata’ is its power to evoke readings against its own grain’’
(1993: 107). Both scholars destabilize the reliability of the protagonist and the
control of his author alike. But then, my interest in the conflicting interpretations
is the theoretical basis that they share with the reliable-mediator alternative. Even
diametrically opposed readings of speaker vis-à-vis author operate, knowingly or
otherwise, by the same deeper principle, namely, accounting for ostensible discordance by reference to an appropriate integration mechanism.
Opposing the Reliable Mediator: The Perspectival
Hypothesis Inverted
One line of counterreading deems Pozdnyshev an unreliable communicator. Its
exponents (predictably, on my account of unreliability) detect tensions among the
data that seem unproblematic or authorized to the first set of readers. For Keith Ellis,
‘‘the moralizing in the novel or the statement of purpose in the ‘Afterword’ . . . cannot
fully account for the action’’; and ‘‘interpretation . . . may be detrimentally restricted
by stated authorial purpose’’ (1971: 892). Isenberg finds a ‘‘conflict between surface
meaning and latent significance,’’ namely, ‘‘between the theory of gender proposed by
the text and the vision of gender performed by it’’ (1993: 93).
For comparison with the above argument for reliability, it is worth noting what
kinds of tensions these inverters allege or emphasize and what kinds get downgraded
or passed over altogether. Ellis thus ignores the epigraph, barely mentions the
opening discussion as ‘‘setting,’’ and regards the anonymous frame-narrator as an
‘‘attentive though unobtrusive audience,’’ who ‘‘limits his comments to occasional
requests for clarification’’ and is ‘‘properly uncritical’’ (1971: 893, 894).4
Instead, the argument for unreliability focuses on the problematic speaker and
speech. As Coetzee rightly observes, his ‘‘agitation,’’ ‘‘funny little sounds,’’ odd ideas,
and odious crime produce a bad impression from the very start, contributing to an
expectation of an unreliable narrator, one whom ‘‘we are all too easily able to read . . .
against himself’’ (Coetzee 1985: 196; emphasis in original). The monologist’s character thus becomes both a cause and a signal of unreliability, ironically targeted by the
implied communication. Accordingly, others highlight Pozdnyshev’s jealousy (‘‘once
its distorting presence is noticed we cannot wholly trust the narrator’s judgments’’
[Ellis 1971: 898]); still others dwell on his mental disorder (‘‘a man who sees the
phallus everywhere’’ [Coetzee 1985: 198]; others see him as a latent homosexual who
is sexually drawn to his wife’s lover [Isenberg 1993: 96–9]); and so forth.
As a reflex of his psychological problems, many have predictably found (or
imagined) Pozdnyshev’s theory to be distorted or marred by contradictions. Those
found between ‘‘theory’’ and ‘‘vision’’ of gender, ‘‘moralizing’’ and ‘‘action,’’ have
specific equivalents. For instance, at one point he defines sex as an acquired, artificial
habit; at another, as an unconscious bestiality (Isenberg 1993: 92–3). Or, as Ellis
(1971) suggests, he ‘‘blames, first men for pursuing women, and then women for
118
Tamar Yacobi
capturing husbands.’’ Again, ‘‘[i]n attacking women’s subjection to men’s sexual
pleasure he gives a strong and sensitive defense of women’s rights, but criticizing
elsewhere their cunning, he declares that ‘the result of all this is the ascendancy
of women, from which the whole world is suffering.’ ’’ He oscillates between the
insistence that every marriage leads to divorce, suicide, and murder and the opposite
claim that the horror of his deed marks his own experience as unique. Further,
the perpetrator’s ‘‘untidy generalizations . . . may well be interpreted as a desperate
attempt at rationalizing the act’’ (Ellis 1971: 894–6). Symptoms of his unacknowledged guilt-complex are detected, for instance, in his misunderstanding of
the lawyer’s reference to ‘‘critical moments in married life’’ as an allusion to his
own crime (Gustafson, in Isenberg 1993: 166, n.23). In Ellis’s account, such incongruities are explicitly reconciled via a psychopathology beyond Pozdnyshev’s rational
control: ‘‘by regarding jealousy as the motivating force in The Kreutzer Sonata
and antithesis as the dominant stylistic trait[,] an interpretation is offered that
would seem to reveal the novel’s unity and coherence’’ (Ellis 1971: 899). So the search
for a principle of integration drives this reading type as well – but away from
reliability. Accepting neither the protagonist’s existential claim (the prevalence
of the marriage crisis), nor his drastic solution, the unreliability hypothesis (over)focuses and subjectivizes the monologue’s infelicities. Their repatterning ensues
from the inverted perspective: both the given account of past events and the
scheme for a reform are diagnosed as symptoms of delusion. In turn, the workings
of the wife-killer’s mind become the central object of implied, ironic interest. The
text’s norms and designs are supposedly concerned not with moral ideology
(as transmitted by an unusual spokesman) but with the speaker’s ongoing psychopathology.
Opposing Authorial Intention or Composition: The Genetic
Alternative
Yet the problematic ideology, not surprisingly, has given rise to a third, mainly
genetic, set of readings. Thereby, tensions and failures are again diagnosed in the text,
yet now found unresolvable within its own coordinates, and so blamed on the author
himself. The shift of responsibility from the explicit to the implicit communicator
intersects not just (as in the reliable vs. unreliable polarizing of the narrator) with an
alternative model of interpretation, but with questions of evaluation and canonicity.
Here are some variants.
1 Some readers refuse to become Tolstoy’s implied audience because of their violent
antagonism to his proposed ideology in the Sonata – whether his iconoclastic
religion or his attitude to women. On either issue, the clash arises not inside the
text but between the advocated thesis and a reader’s imported counterthesis.
The issue and the import favored often express group interests: they jointly oppose
Tolstoy’s Kreutzer Sonata
119
the thesis from diverse, even mutually hostile, viewpoints. Thus the acrimonious
responses of contemporary church dignitaries (in Møller 1988: 145–62) as against
the ‘‘counter-literature’’ written as polemic parallel to the Sonata (Møller 1988:
163–80) or the historical and cultural anachronism of later feminist attacks on
Tolstoy (reviewed by Mandelker 1993: 15–57).5
2 More frequently, readers acknowledge the text’s ideological and didactic center of
interest, and take the monologist’s reliability as normative. Yet they find inconsistencies within the proposed ideology (say, women changing roles between victim
and pursuer; the spirit/body split; the blindness to scientific facts). Since they find
no distance between author and spokesman, they blame the author for the tensions.
Some have indeed banished the Sonata from their Tolstoy canon, as ‘‘a grossly
imperfect work’’ (Davie 1971: 326), which ‘‘shows the worst side of Tolstoy’’
(Spence 1961: 227).6 Others have ingeniously differentiated the artist from the
ideologist as two conflicting Tolstoyan personalities. One ‘‘strong tradition’’ would
have Tolstoy ‘‘during the creative process . . . split into’’ these ‘‘personalities, each
attempting to dupe the other’’ and applauds ‘‘when the artist, as the more
sympathetic of the two, comes out the winner’’ (Møller 1988: 17).
Elsewhere the split migrates to the text’s readings, differentiated by appeal to its
addressees. Thus, after enumerating Pozdnyshev’s many failures of fact, interpretation, ideology, and self-scrutiny, Isenberg suggests that readers of the Sonata can
follow Tolstoy’s directive and echo the ‘‘three kinds of responses’’ voiced by his
‘‘primary listener’’: either absolving the protagonist, or taking his lesson to heart,
or, at least, being ‘‘seduced’’ by his tale without analyzing ‘‘its disruptions of
meaning.’’ But this, for Isenberg, is the response of the ‘‘submissive’’ reader who
‘‘cave[s] in’’ on the model of the equally submissive frame narrator. Instead,
the resisting reader who accepts the work’s surface intentions as the only possible
guide . . . may well end up pronouncing it a tendentious failure – or the reader may
recognize his or her stake in making it work as a story, not as received wisdom, even
if this means rejecting its homiletic aims and making Tolstoy of the devil’s party
without his knowing it.
(Isenberg 1993: 107)
In fact, then, Isenberg offers here three principled lines of reading, roughly
corresponding to the three groups of actual readers, and divided among the
relevant integration mechanisms. The reliable alternative (itself threefold) is possible, he maintains, but only for those who accept Tolstoy’s homily and dictate,
closing their eyes to the many confusions in the text. Those alive and resistant to
the text’s incongruities, vis-à-vis the author’s extratextual intentions, end up with a
genetic reading, and pronounce the Sonata ‘‘a tendentious failure.’’ At best, they
will save the text by ‘‘rejecting its homiletic aims’’ in favor of a storied, ironic
reading, counter to the writer’s and presumably the narrator’s viewpoint.
120
Tamar Yacobi
As this example illustrates, quite a few of the tensions adduced by a genetic
approach coincide with those discovered or imagined or magnified by advocates of
the perspectival hypothesis – except that, on the genetic reading, they point to the
failure of the author instead of his mediator. It’s the ideology and generalizations
(on women, marriage, physicians, etc.) that are most often attacked for their factual
and logical deficiencies, up to self-contradiction. But sometimes the genetic drive
corresponds to the unreliability hypothesis in uncovering a psychopathology: now
again Tolstoy’s own, or his own as well as the character’s, and beyond the artist/
ideologue split.
3 For such readers, Pozdnyshev’s oddities reflect the psychological (even psychotic)
disturbances of his creator. So Daniel Rancour-Laferriere, in Tolstoy on the Couch
(1998), traces Pozdnyshev’s mental disorders (above all, his misogyny) to Tolstoyan
traumas, as far back as the loss of his mother at the age of two. This analyst even
denies that Pozdnyshev’s words on marriage are representative but judges them
‘‘self-oriented, that is, narcissistic’’ (p. 62). Nor is the epigraph, for him, quoted ‘‘to
propagandize Christianity, but to express’’ Tolstoy’s personal beliefs (p. 76). Such a
reading, by the usual genetic logic, locates unity (but not coherence) somewhere
outside the text, namely, in the unbalanced personality of the author.
4 Some genetic judgments are transformed into or linked with textually coherent
(and hence authorized – reliable or unreliable) readings. Thus Emmanuel Velikovsky declares that ‘‘the jealous murderer Pozdnishef is presented as a homosexual
who did not know his own nature; even the author, Tolstoy, failed to realize this’’
(1937: 18). Still, Tolstoy is congratulated throughout on his wonderful, if unconscious, ‘‘intuition’’ in dramatizing repressed homosexuality, and various of Pozdnyshev’s lapses allegedly compare with the behavior of a psychotherapist’s real
patients. So, judged by Freudian insight or (self-)knowledge, the author ultimately
comes out in a positive light, while the diagnosed hero is exposed in his unreliability. The publication of this essay in the Psychoanalytic Review underscores the
external criteria for evaluating the text’s success.7
From a literary viewpoint, Coetzee arrives at another balance between perspectival
and genetic reading. To begin with, Coetzee juxtaposes the various oddities, culminating in the murder, that imply Pozdnyshev’s unreliability with various signals of his
reliability. The notable contribution to the latter hypothesis is the argument from
genre: that Tolstoy did not dramatize Pozdnyshev’s (debatable) conversion, because he
was more interested in advocating the truth than in the psychic experience of its
revelation. Given this downgrading of psychology, we need to follow the lead of the
anonymous narrator who, as listener, supports the protagonist’s version of the truth by
his ‘‘silence’’ (sic) throughout the confession (Coetzee 1985: 204).
So far, Coetzee’s analysis is another variant of the reliability hypothesis. But then he
identifies a genetic oddity: why did the author of Anna Karenina, with his brilliant
psychological insight, write ‘‘so naı̈ve and simple-minded a book, in which the truth
Tolstoy’s Kreutzer Sonata
121
that the truth-teller tells emerges as a bald series of dicta’’ (1985: 231)? Coetzee’s
answer links the author’s peculiar character and history with the peculiarities of our
story. First, he reminds us, even the earlier Tolstoy ranked truth above the dramatic
process of illumination. More importantly, though still able to make this text
psychologically ‘‘richer’’ or ‘‘deeper,’’ ‘‘by making it ambiguous,’’ he has lost interest
in this ‘‘machinery.’’ Coetzee therefore ends with a speculation on the genetic process
of The Kreutzer Sonata: ‘‘Tolstoy’s impatience with the novelistic motions’’ could have
led to his ‘‘(rash?) decision to set down the truth, finally, as though after a lifetime of
exploring one had acquired the credentials, amassed the authority, to do so’’ (p. 232;
emphasis in original). This genetic hypothesis supplies an answer to yet another
problem, namely, that the Sonata is open not only to Tolstoy’s intended (reliablenarrator) reading but to counterreadings, including the unreliability hypothesis.
Tolstoy’s ‘‘lack of armament against other, unauthorized readings’’ (p. 204) would
thus signify his ideological rejection of the well-made tale along with the larger one of
art as a goal.8
I will now leave aside further variants that reinforce the point. Owing to this diversity
of interpretations, and of elements cited for and/or against, The Kreutzer Sonata offers
an effective test case of my theoretical framework. Its exemplarity largely springs from
the fact that no reading is strong enough, nor any worldview consensual enough, to
decide the issue in accounting for all major problems. Less contested texts will be
distributed along fewer, simpler, or more stable lines among the various integration
mechanisms that it exemplifies. However, amid the shift from one perspectival
reading (Pozdnyshev is reliable) to its opposite (he is unreliable) or even to a genetic
failure, both the logic of integration as such and the principled power structure of the
text (author above mediating narrator, Pozdnyshev the creature of Tolstoy) remain in
force. Whatever the final diagnosis, the interpretation necessarily concerns (at least)
two discourse-contexts: one explicit and freely problematic, the other implicit and
determinative for better or worse. Unlike real-life communication, a fictive speaker’s
reliability is determined neither by some objective truth (of fact or idea) or poetic rule
(‘‘artistry’’), nor on the basis of equality (of subjects, evidence, attitudes), but in
relation, concordant or conflictual, to the hypothesized norms and goals of the author.
NOTES
1 Andrea Dworkin is so much in agreement
with Tolstoy on the problem’s universality,
that she opens her feminist book on Intercourse
with a discussion of this tale (1987: 3–20).
Felman (1997) suggests the ongoing relevance
of the Tolstoy/Pozdnyshev analysis of gender
relations, opposing Pozdnyshev’s laudable con-
fession of guilt with O. J. Simpson’s plea of
not guilty.
2 Contrast Isenberg’s (1993: 107) threefold division of those inside responses: the confessor’s
absolving function; the sermon audience’s taking the lesson to heart and disseminating it;
the model reader seduced by the tale without
122
Tamar Yacobi
noticing its inner tensions. For more on this
reading, see below.
3 Other critics who endorse the reliable reading
differ mainly concerning the relevant normative scale. For example, Spence affirms that
nowhere is ‘‘Pozdnyshev’s theory. . . contradicted’’ by the later Tolstoy, being ‘‘the logical
outcome of the despairing passages of the Confession’’ (Spence 1963: 161–2). For Herman,
the mystery that torments both author and
narrator is ‘‘the problem of art’s likeness to
adultery’’ (Herman 1997: 17); etc.
4 Likewise, Christian (1969: 231), Bayley
(1966: 283–4), Herman (1997: 34) or Coetzee
(1985: 204) either ignore or belittle the function of the inside addressee.
5 Of particular ‘‘counter-literary’’ interest is the
story by Countess Tolstoy, which retells the
murder plot from the wife’s viewpoint, in a
double sense. Among modern feminists,
Dworkin is surprisingly ambivalent about
this ‘‘androcentric’’ tale, ‘‘crazed with misogyny and insight.’’ Her solution is in effect
antigenetic, differentiating Tolstoy’s art (‘‘he
articulates with almost prophetic brilliance
the elements that combine to make and keep
women inferior’’) from his own revolting be-
REFERENCES
AND
Bayley, J. (1966). Tolstoy and the Novel. London:
Chatto & Windus.
Booth, W. C. (1961). The Rhetoric of Fiction. Chicago: University of Chicago Press.
Chekhov, A. (1971). From Selected Letters. In Henry
Gifford (ed.), Leo Tolstoy: A Critical Anthology,
trans. S. Lederer (pp. 97–8). Harmondsworth,
UK: Penguin.
Christian, R. F. (1969). Tolstoy: A Critical Introduction. Cambridge, UK: Cambridge University
Press.
Coetzee, J. M. (1985). ‘‘Confession and Double
Thoughts: Tolstoy, Rousseau, Dostoevsky,’’
Comparative Literature 37, 193–232.
Davie, D. (1971). ‘‘The Kreutzer Sonata.’’ In
H. Gifford (ed.), Leo Tolstoy: A Critical Anthology (pp. 326–34). Harmondsworth, UK:
Penguin.
Dworkin, A. (1987). Intercourse. New York: Free
Press.
havior to his wife (Dworkin 1988: 4, 19). She
thus accepts the implied author, concedes reliability to his homicidal speaker – no-sex
program included – and repudiates the sexual
beast who wrote the text.
6 Anton Chekhov first applauded the Sonata for
its art, but particularly because ‘‘it is extremely thought-provoking,’’ even when it
fluctuates between ‘‘true’’ and ‘‘ridiculous’’
points (Chekhov 1971: 97). But he changed
his mind after reading the ‘‘Sequel,’’ pronouncing Tolstoy’s philosophy stupid, arrogant,
and ignorant (Chekhov 1971: 98); see Møller’s
fine chapter on his critical and literary response (Møller 1998: 208–58).
7 Whereas Velikovsky compliments Tolstoy’s
intuition, Isenberg, with his very similar diagnosis of the hero’s psychopathology, emphasizes the tension between the writer’s
intended reliable mediator and the homilyresisting or unreliable alternative.
8 Recall Tolstoy, as quoted in Møller, on the fuss
made about the tale’s artistry, to which he has
actually given ‘‘just enough room for the terrible truth to become visible’’ (Møller 1988:
10). See also Herman (1997: 32, 34–6).
FURTHER READING
Ellis, K. (1971). ‘‘Ambiguity and Point of View in
Some Novelistic Representations of Jealousy.’’
Modern Language Notes 86(6), 891–909.
Felman, S. (1997). ‘‘Forms of Judicial Blindness,
or the Evidence of What Cannot Be Seen.’’
Critical Inquiry 23, 738–88.
Herman, D. (1997). ‘‘Stricken by Infection: Art
and Adultery in Anna Karenina and Kreutzer
Sonata.’’ Slavic Review 56(1), 15–36.
Isenberg, C. (1993). Telling Silence: Russian Frame
Narratives of Renunciation. Evanston, IL: Northwestern University Press.
Mandelker, A. (1993). Framing Anna Karenina:
Tolstoy, the Woman Question, and the Victorian Novel. Columbus: Ohio State University
Press.
Møller, P. U. (1988). Postlude to the Kreutzer Sonata:
Tolstoj and the Debate on Sexual Morality in Russian Literature in the 1890s, trans. J. Kendal.
Leiden, Netherlands: Brill.
Tolstoy’s Kreutzer Sonata
Rancour-Laferriere, D. (1998). Tolstoy on the Couch:
Misogyny, Masochism and the Absent Mother. New
York: New York University Press.
Simmons, E. J. (1960). Leo Tolstoy: The Years of
Maturity 1880–1910, vol. II. New York: Vintage Books.
Spence, G. W. (1961). ‘‘Tolstoy’s Dualism.’’ Russian Review 20(3), 217–31.
Spence, G. W. (1963). ‘‘Suicide and Sacrifice
in Tolstoy’s Ethics.’’ Russian Review 22(2),
157–67.
Sternberg, M. (1978). Expositional Modes and Temporal Ordering in Fiction. Baltimore, MD: Johns
Hopkins University Press.
Sternberg, M. (1983). ‘‘Mimesis and Motivation:
The Two Faces of Fictional Coherence.’’ In
J. Strelka (ed.), Literary Criticism and Philosophy
(pp. 145–88). University Park: Pennsylvania
State University Press.
Sternberg, M. (2001). ‘‘Factives and Perspectives:
Making Sense of Presupposition as Exemplary
Inference.’’ Poetics Today 22, 129–244.
123
Tolstoy, L. N. (1929). ‘‘Sequel to the Kreutzer
Sonata,’’ trans. A. Maude. In Master and Man,
The Kreutzer Sonata, Dramas (pp. 155–70). New
York: Charles Scribner’s Sons.
Tolstoy, L. (1960). ‘‘Appendix to The Kreutzer Sonata.’’ In Great Short Works of Leo Tolstoy, trans. L.
and A. Maude (pp. 429–49). New York: Harper
& Row.
Tolstoy, L. ([1891] 1963). ‘‘The Kreutzer Sonata,’’
trans. M. Wettlin. In Six Short Masterpieces by
Tolstoy (pp. 284–369). New York: Dell.
Velikovsky, I. (1937). ‘‘Tolstoy’s Kreutzer Sonata
and Unconscious Homosexuality.’’ Psychoanalytic Review 24, 18–25.
Yacobi, T. (1981). ‘‘Fictional Reliability as a Communicative Problem.’’ Poetics Today 2, 113–26.
Yacobi, T. (1987). ‘‘Narrative and Normative Pattern: On Interpreting Fiction.’’ Journal of Literary Studies 3, 18–41.
Yacobi, T. (2001). ‘‘Package-Deals in Fictional
Narrative: The Case of the Narrator’s (Un)Reliability.’’ Narrative 9, 223–9.
7
Henry James and ‘‘Focalization,’’
or Why James Loves Gyp
J. Hillis Miller
Henry James’s prefaces to the New York Edition of his novels and tales develop many
wonderful narratological concepts, concepts inexhaustible for meditation. One of
these is the proposition that limitation to a single ‘‘center of consciousness,’’ in
works of fiction, is a basic requisite of good form. It is the best way to avoid
producing what James, in the preface to The Tragic Muse, disdainfully called ‘‘large
loose baggy monsters’’ (James 1971–9, vol. 7: x). Examples James gives are War and
Peace, Les Trois Mousquetaires, and The Newcomes. James opposed such monsters to
‘‘a deep-breathing economy and an organic form,’’ in which he said he ‘‘delighted’’
(ibid.). James’s inordinate imagination and his tendency to ‘‘sprawl’’ and to create
misplaced middles (James, 1971–9, vol. 26: 299, vol. 7: xi-xii, xxi) made some
principle of limitation especially necessary for him. The first half of James’s The
Golden Bowl is, for the most part, limited to the Prince’s consciousness of things and
people, while the second half makes Maggie the center of consciousness. The Ambassadors is limited, with magnificent consistency, to Strether’s ‘‘point of view.’’ It can be
argued, of course, that the real ‘‘center of consciousness’’ in these novels is the
encompassing ‘‘omniscient narrator’’ who presents through free indirect discourse
the way things looked to Prince Amerigo, to Maggie, or to Strether. One of James’s
dominant linguistic strategies is not interior monologue but free indirect discourse,
the presentation, in the narrator’s past-tense language, of the present-tense language
of the character, or, sometimes, of the character’s unworded interiority. That narrator
might be defined as a consciousness of the consciousness of others. The narrator is
almost a collective or community consciousness, if indeed anything like a community
is presupposed in James’s fiction.
Other names for ‘‘center of consciousness’’ are the now old-fashioned term ‘‘point of
view’’ and, in subtly elaborated recent narratological theory, ‘‘focalization.’’ None of
these terms is innocent. Each in one way or another begs the question it is meant to
clarify. All three, for example, evade the fact that novels are made of words. They tend
to imply either that a novel is a matter of ‘‘looking’’ or that it is made of ‘‘con-
Henry James and ‘‘Focalization’’
125
sciousnesses.’’ The term ‘‘focalization’’ is drawn from optics. Its figurative base does
not differ from ‘‘point of view,’’ except that it defines ‘‘point of view’’ not as a matter of
looking from a certain position, but rather as a matter of getting things in focus when
looking through some device or other, such as a telescope, binoculars, a microscope, or
a mind/body compound considered as a focalizing apparatus. Such terms elide the way
the essential mode of existence of any literary fictional work is linguistic through and
through. There ain’t nothing there but words.
Another way to put this is to say that though such terms as ‘‘center of consciousness,’’ ‘‘point of view,’’ or ‘‘focalization’’ may be essential to present-day narrative
theory, they are figures of speech. No consciousness as such exists in any novel, only
the representation of consciousness in words. No looking or bringing into focus exists
in any novel, only the virtual phantasm of these as expressed in words. This is not a
trivial distinction. The term ‘‘free indirect discourse,’’ on the contrary, is a genuinely
linguistic term.
Most narratologists of course know all this. In spite of sometimes seeming to take
more delight in subtle refinements of distinction among various forms of ‘‘focalization’’ than in demonstrating how these formal features are related to meaning,
narratologists, for the most part, know that their distinctions are not useful in
themselves. They are useful only if they lead to better readings or to better teachings
of literary works. Narratological distinctions and refinements are not valuable in
themselves, as ‘‘science.’’ At least they are not valuable in quite the same way as
decoding the human genome is valuable. The latter not only finds out new facts –
those facts are also socially useful. They lead to new medicines and to new cures for
diseases. Narratological distinctions, unlike scientific discoveries, are not facts about
the external world. They are disciplinary artifacts concocted for heuristic purpose to
allow talk about certain features of human language. Narratological distinctions are
useful primarily as aids to better reading, which is socially useful. As Henry James
understood, form is meaning. Sticking to the formal principles he has chosen for The
Awkward Age, James says in the preface to that novel, ‘‘helps us ever so happily to see
the grave distinction between substance and form in a really wrought work of art
signally break down’’ (James 1971–9, vol. 9: xxi).1 All good reading is therefore
formalist reading.
In this essay I want to demonstrate this ‘‘breaking down’’ of the distinction between
form and substance by reading, at least partially, a quite anomalous work by James,
The Awkward Age (1899). This work is anomalous, that is, if we assume that for James
the principle that there must be a ‘‘center of consciousness’’ in order to achieve
economy of form was a universal law. The oddness, or lawlessness, of The Awkward
Age is that it eschews both the omniscient narrator and, almost entirely, what he calls
in the preface ‘‘going behind’’ (p. xvii) the objective presentation of the characters.
This novel has neither an omniscient narrator in the usual sense nor, for the most part,
any centers of consciousness or points of view or focalizations in the minds and
feelings of the characters. The novel describes objectively what the characters said
and did, how they looked, their clothes, faces, gestures, and other behavior, the things
126
J. Hillis Miller
with which they have surrounded themselves. The reader must from this ‘‘superficial’’
or manifest evidence figure out what is going on behind the objective signs. Additional resources are cross-references, repetitions, allusions, echoes, and the like. These
link one scene or utterance in the novel to others. Gérard Genette ([1972] 1980)
identifies this technique as ‘‘external focalization,’’ reading it as evidence that the
narrator knows less than the characters know. Genette’s example is Dashiell Hammett
(p. 190). In James’s case, however, the reader senses that the narrator knows all that is
going on in the characters’ minds and could reveal it if he would. The evidence is
those few places where the narrator does enter the mind of a character, especially
Vanderbank’s mind in scenes early in the novel.
James, in the preface, is quite specific about his procedures. What he says comes
in the context of defending his scrupulous obedience, in The Awkward Age, to the laws
of a stage play. In such a work, he says, he can ‘‘escape poverty’’ and attain ‘‘beauty,’’
‘‘even though the references in one’s action can only be, with intensity, to things exactly
on the same plane of exhibition with themselves’’ (p. xx). In this novel, he goes on
to say, there are no obnoxious ‘‘loose ends,’’ that obtrude like ‘‘the dangle of a
snippet of silk or wool on the right side of a tapestry. We are shut up wholly to
cross-relations, relations all within the action itself, no part of which is related
to anything but some other part – save of course by the relation of the total to life’’
(ibid.). James remains true here to the general principles, endorsed throughout all the
prefaces, of a mimetic realism that values literature for its representational value,
its imitation of real life. That relation, however, is not to be measured by bits and
parts, but by ‘‘the relation of the total to life.’’ The total in itself, however, draws
its intrinsic meaning from cross-relations, not from the referential value of the
parts. In Saussure’s theory of language, in an analogous way, words or phonemes
draw their meanings not from reference, but from their differential relation to one
another. In The Awkward Age these elements are all ‘‘exactly on the same plane of
exhibition.’’ They all obey the law that only what a sharp onlooker would have
seen and heard can be ‘‘exhibited,’’ that is, brought into the open. The rest remains
hidden, secret.
Why in the world does James in this case avoid just those narrative techniques that
are usually his most powerful tools, for example in the three great masterpieces that
followed just a few years after The Awkward Age, that is, The Ambassadors, The Wings of
the Dove, and The Golden Bowl? If I can answer that question I may be able to
exemplify what it means to say that form is meaning, or that the distinction between
substance and form signally breaks down in a ‘‘really wrought work of art.’’
First I must identify the models James claims, in the preface, that he followed in
The Awkward Age. One is the stage play. In the preface James tells the reader,
somewhat ruefully and ironically, how he presented to the editors of Harper’s Weekly,
where the novel first appeared ‘‘during the Autumn of 1898 and the first weeks of
winter’’ (p. xv), a diagram on a sheet of paper of his plan for the novel. The comedy is
that he now suspects the editors did not have the slightest idea what he was talking
about:
Henry James and ‘‘Focalization’’
127
I remember that in sketching my project for the conductors of the periodical I have
named I drew on a sheet of paper – and possibly with an effect of the cabalistic, it now
comes over me, that even anxious amplification may have but vainly attenuated – the
neat figure of a circle consisting of a number of small rounds disposed at an equal
distance about a central object. The central object was my situation, my subject in itself,
to which the thing would owe its title, and the small rounds represented so many
distinct lamps, as I liked to call them, the function of each of which would be to light
with due intensity one of its aspects. I had divided it, didn’t they see? into aspects –
uncanny as that little term might sound (though not for a moment did I suggest we
should use it for the public), and by that sign we would conquer. [ . . . ]
Each of my ‘‘lamps’’ would be the light of a single ‘‘social occasion’’ in the history and
intercourse of the characters concerned, and would bring out to the full the latent color
of the scene in question and cause it to illustrate, to the last drop, its bearing on my
theme. I reveled in this notion of the Occasion as a thing by itself, really and completely
a scenic thing, and could scarce name it, while crouching amid the thick arcana of my
plan, with a large enough O.
(pp. xvi-xvii)
The reader will note that the ‘‘lamps’’ here are not centers of consciousness, but
objectively presented social scenes, the give and take of conversation in a drawing room.
It might appear at first that James is presenting the reader with another exclusively
visual figure for the procedures of the novelist, another figure like ‘‘focalization’’ or
‘‘point of view.’’ A careful reading, however, shows that matters are not quite so
simple. James’s figure is presented overtly, up front, as a figure, a trope, and an
‘‘arcane,’’ ‘‘cabalistic,’’ or ‘‘uncanny’’ one at that. What is cabalistic, arcane, or uncanny
about it? These terms refer to a code for something secret and to the experience of
something spooky that nevertheless seems somehow familiar, something in short, in
Freud’s term for uncanny, ‘‘unheimlich.’’ The mystery lies in the impossibility of
naming, in so many words, the ‘‘central object,’’ his ‘‘situation.’’ That central object
is uncanny because it appears only in ghosts or revenants, doubles of itself, never in
itself or as itself. James nowhere says what he put down, as a picture of this central
object, on his mystic sheet of paper, over which he ‘‘crouched’’ like some medieval
mage. Was it a point, or another circle, or the drawing of some object or other? He
does not say. He does not say because he cannot say. He can only shine lamps on the
object and reveal its ‘‘aspects’’ one by one through that method. The rhetorical label
for this procedure, that is, for the tropological naming of something that does not
have a literal name, is ‘‘catachresis.’’ Each of James’s ‘‘Occasions’’ are catachreses for an
unnamed central object, the situation, which can be named in no other way.
To call each ‘‘Occasion’’ a ‘‘lamp’’ to ‘‘light with due intensity one of its aspects’’
differs fundamentally, moreover, from the figure of ‘‘focalization.’’ The latter implies
that the object of the narration is there to be seen. It is just a matter of getting it into
focus. James’s figure of ‘‘Occasions’’ that illuminate a central object, on the contrary,
implies that the object is in the dark, to some degree permanently in the dark. It, or
128
J. Hillis Miller
rather just one of its ‘‘aspects,’’ can only be brought into the light by way of a given
‘‘Occasion.’’ The latter functions performatively to expose it, or, rather, to expose one
of its aspects. ‘‘Occasion’’: the word names something that happens or befalls, a
fortuitous event. The word comes from Latin occidere (past participle occasus), to fall
down. The difference between a lamp and a focalizer is analogous to Meyer Abrams’s
([1953] 1971) famous distinction between the mirror and the lamp. Abrams set the
mirror of classical theories of mimesis against the quasi-constitutive force of poetic
language. The latter is figured in romantic theories as a lamp.
What justifies my implicit claim, when I speak of catachreses, that these Occasions
are acts of language figured by the optical images James uses, not literally matters of
seeing at all? The answer is that these Occasions are, according to James’s strict formal
or generic rule in this novel, made of conversational exchanges in some social scene.
Each Occasion brings two or more of the chief characters together in talk, with a
minimum of stage directions. A given ‘‘lamp’’ illuminating some aspect of the central
situation is made up almost exclusively of what the characters say, plus their gestures
and bodily behavior as reported by the narrator. The Occasion is made, that is, of what
a hovering, invisible spectator might have seen or heard, especially heard.
A leitmotif of The Awkward Age occurs in sentences like the following: ‘‘As the
reflection of her tone might have been caught by an observer in Vanderbank’s face it
was in all probability caught by his interlocutress [Nanda], who superficially, however, need have recognized there – what was all she showed – but the right manner of
waiting for dinner’’ (p. 389; see also, among other examples, pp. 148, 238, 269, 400,
424, 449, 514, 535–6). David Herman, in terms that are perhaps a little unnecessarily barbarous or even misleading, calls this ‘‘direct hypothetical focalization’’ or the
use of a ‘‘counterfactual focalizer’’ (Herman 1994: 237). He means by this that no real
witness existed. This is true enough if we imagine the events as having really taken
place, but of course the whole story is ‘‘counterfactual’’ in the sense that it precisely
did not take place. It exists only in the words for it. The hypothetical witness is the
narrator as surrogate for the reader. The latter is imagined to be in the situation of an
audience-member carefully watching a stage play. Such a spectator is limited, in the
evidence for his or her interpretation, to what the actors say and how they behave and
look.
That this was James’s deliberate strategy in this novel, and that he was aware that it
differed markedly from his normal procedure, is confirmed by what follows the
passage in the preface about the uncanny little circles arranged in a cabalistic circle
around a mysterious central object. Here James says explicitly that his model was the
stage play:
The beauty of the conception was in this approximation of the respective divisions of
my form to the successive Acts of a Play – as to which it was more than ever a case for
charmed capitals.2 The divine distinction of the act of a play – and a greater than any
other it easily succeeds in arriving at – was, I reasoned, its special, its guarded
objectivity. This objectivity, in turn, when achieving its ideal, came from the imposed
Henry James and ‘‘Focalization’’
129
absence of that ‘‘going behind,’’ to compass explanations and amplifications, to drag out
odds and ends from the ‘‘mere’’ storyteller’s great property-shop of aids to illusion: a
resource under denial of which it was equally perplexing and delightful, for a change, to
proceed. Everything, for that matter, becomes interesting from the moment it has
closely to consider, for full effect positively to bestride, the law of its kind. ‘‘Kinds’’
are the very life of literature, and truth and strength come from the complete recognition of them, from abounding to the utmost in their respective senses and sinking deep
into their consistency. I myself have scarcely to plead the cause of ‘‘going behind,’’ which
is right and beautiful and fruitful in its place and order; but as the confusion of kinds is
the inelegance of letters and the stultification of values, so to renounce that line utterly
and do something quite different instead may become in another connection the true
course and the vehicle of effect.
(p. xvii)
This seems clear enough, but just why does James renounce ‘‘going behind’’? Why
does he give up his special line, give up what he is famous for doing? Why does he
attempt to write a novel that is as much as possible like a stage play, or, one might say,
is like the report of an imaginary play as if seen by a clever spectator? Why does this
‘‘form’’ seem to James especially appropriate for this particular ‘‘substance,’’ this
‘‘central object,’’ this ‘‘situation,’’ this ‘‘subject in itself,’’ this ‘‘theme’’? The answer,
though one hopes this is not the case, may be that no particular consonance between
form and substance, in this instance, exists. James just wanted to try something
different for a change. The novel could have been written as The Portrait of a Lady or
The Golden Bowl or The Ambassadors is written. James just decided to experiment with
dramatic or scenic form. Once he had decided to do that, an aesthetico-moral
obligation not to confuse kinds required that he be strictly consistent. Moreover,
one might argue, James was still feeling angry and humiliated by the failure in 1895
of his stage play, Guy Domville. The audience hissed the play and hissed James off the
stage when he appeared there after the curtain. He was especially humiliated because,
at the moment James’s play failed, Oscar Wilde’s An Ideal Husband was a resounding
success in another nearby London theater. The paragraphs in the preface that follow
the passage I have just quoted are, by way of a discussion of Dumas and Ibsen (Wilde
of course is not mentioned), an angry and defensive attack on the British play-going
public as having no more than ‘‘infantine intelligence’’ (p. xviii). I’ll show the world,
James may secretly have thought, that I can do something of the stage play ‘‘kind,’’
though in my own métier, the novel. I’ll go Wilde – who is a scandalous person in any
case – one better after all.
The other model James claims to have followed suggests, however, that some other
deeper reason may have dictated adoption of dramatic form in The Awkward Age.
Though no one seems to have noticed, says James, he was copying the procedures of
‘‘the ingenious and inexhaustible, the charming philosophic ‘Gyp’ ’’ (p. xii). Who in
the world was Gyp? Many specialists in modern French literature have never heard of
her, as I have found out by asking around. Nor have critics of The Awkward Age often
followed up this precious clue. ‘‘Gyp’’ was the pseudonym of a nineteenth- and
130
J. Hillis Miller
twentieth-century French writer with the almost unbelievable name, so resonant of
the French Revolution, of Sibylle Gabrielle Marie Antoinette Riqueti de Mirabeau,
Comtesse de Martel de Janville (1849–1932). I kid you not! Gyp was the author of
many novels, plays, and other works. Friedrich Nietzsche, of all people, also greatly
admired her. He mentions her in Ecce Homo as a model writer. What Nietzsche
admired especially was the wonderful economy and presto tempo of Gyp’s novels,
along with the limpidity of her idiomatic French. I too admire these. A shared
admiration for Gyp is about the only overt connection between James and Nietzsche
that I know of.
That ‘‘happiest of forms,’’ of which Gyp was mistress, was the use of dialogue in
novels that dispensed almost altogether with narrative comment or ‘‘going behind.’’
Gyp often indicates the speaker simply by giving the character’s name, accompanied
by ‘‘said,’’ for example ‘‘Chiffon said.’’
This ‘‘Gyp taint’’ of The Awkward Age, as James calls it (p. xiv) may suggest that the
dramatic form James chose for it is appropriate because the novel is somehow in
substance as well as in form like Gyp’s novels, for example Le mariage de Chiffon
(1894). This novel has parallels with The Awkward Age, and James may possibly have
read it in its 1895 edition.
Just what was that ‘‘substance,’’ the ‘‘situation,’’ the ‘‘subject’’ of The Awkward Age?
James explains in the preface that the focus is on what happens when a young girl of
marriageable age is, at that particular ‘‘awkward’’ moment in English social history,
first brought downstairs to hear the talk of adults. The term ‘‘the awkward age’’ refers
not only to a moment of adolescence, but also to the historical moment of transition
that is the novel’s setting. ‘‘The awkward age’’ is another name for ‘‘le fin de siècle.’’
What is awkward about the awkward age seems almost ludicrous to present-day
readers. An act of the historical imagination is required to take it seriously. Girls were
still at that time supposed (in what must always have been something of a fiction) to
go to their marriage beds totally ignorant of the facts of sex. At the same time, ‘‘talk,’’
as James calls it, in middle- and upper-class drawing rooms was becoming freer and
freer. Young girls were more and more allowed to hear that talk. The talk among Mrs
Brookenham’s ‘‘set’’ is almost entirely focused on speculation about the sexual misdeeds of themselves or of people they know. Impasse! Aporia! Social behavior contradicted old-fashioned assumptions about necessary virginity of mind as well as of body
in unmarried girls of marriageable age who can hope to get married.
James distinguishes carefully the English situation from the continental or American one. In the United States no such talk occurs, or indeed, in James’s view, any talk
whatsoever that is at all interesting. In France or Italy such talk occurs, but unmarried
young women are carefully sequestered from it. Nanda Brookenham, the heroine of
The Awkward Age, is brought down too soon and hears all the talk in her mother’s
drawing room and elsewhere. She spends time, for example, with a newly married
friend, Tishy Grendon, who fears her husband is betraying her. Nanda learns all about
sex, especially illicit sex. Therefore, though she is wonderfully attractive and intelligent, a little like Gyp’s Chiffon, she becomes, in her particular community or ‘‘set,’’
Henry James and ‘‘Focalization’’
131
unmarriageable. Chiffon, by the way, also knows all about sex and about the sexual
misdeeds of her elders, but this does not prevent a happy ending. Chiffon engages
herself to her rich ‘‘Uncle Marc’’ at the end of the novel and will presumably live
happily ever after.
The central event of The Awkward Age is Vanderbank’s tacit refusal to marry Nanda
Brookenham. This happens even though Mr Longdon, a rich older man from the
country, who loved Nanda’s grandmother and was refused by her, has promised a
fortune to Vanderbank, by way of a dowry to Nanda, if he will marry her. Why does
Vanderbank renounce this opportunity? How is his renunciation related to James’s
choice of what is, for him, such an anomalous form?
I suggest several answers to these questions, all of which may be supported
by citations from the text. The text is, after all, all the evidence we have, plus the
preface, an entry or two in the Notebooks, and comments in the letters. The latter
three may or may not be misleading, deliberately or otherwise. In a letter to Mrs
Humphry Ward, James tells her, ‘‘I ‘go behind’ left and right in ‘The Princess
Casamassima,’ ‘The Bostonians’ . . . just as I do it consistently never at all (save for a
false and limited appearance, here and there, of doing it a little, which I haven’t time to
explain) in ‘The Awkward Age’ ’’ (James 1974–84, vol. 4: 110). Well, what does the
evidence show?
One may begin by saying that if Vanderbank had married Nanda, as Chiffon is to
marry Uncle Marc, in Gyp’s Le mariage de Chiffon, it would have gone against the
almost universal law of James’s fiction. This law says that renunciation is the final act
in almost all his fictions. His novels and stories almost all end with a giving up. This
giving up most often, though not always, is of ordinary marriage or heterosexual
relations. Strether refuses Maria Gostrey’s offer of herself, in The Ambassadors. The
unnamed narrator of ‘‘The Aspern Papers’’ refuses to marry Miss Tina, though that is
the only way he can get the papers. Isabel, in The Portrait of a Lady, refuses Caspar
Goodwood’s offer to free her from her bad marriage to the perfidious Gilbert Osmond,
if she will run away from him. Kate Croy refuses to marry Merton Densher, in The
Wings of the Dove, though she has earlier sworn that she gives herself to him forever.
Densher is so much ‘‘in love with her [Milly’s] memory,’’ as Kate says (James 1976–9,
vol. 20: 404), that he will not take Milly’s money and then marry Kate, which is the
condition Kate sets. The adept reader of James’s fiction expects that somehow or other
Nanda’s passion for Vanderbank will remain unsatisfied. The astute reader will have
noted that ‘‘renounce’’ and ‘‘sacrifice’’ are the words James uses in the preface (p. xvii)
to define his decision, in this case, not to ‘‘go behind.’’ Novelistic form, for James, is
always a matter of sacrifice, a matter of cutting out, a matter of renunciation or giving
up, in order to avoid sprawl, but he is in The Awkward Age sacrificing a procedure
especially dear to him and essential to most of his work. This sacrifice is necessary in
order to conform to the law that form must match theme.
Why, in this particular case, does the main protagonist decide to renounce? One
answer, hinted at by various details in the novel, is that Vanderbank, like James
himself, is not the marrying sort. He may be, however covertly and circumspectly this
132
J. Hillis Miller
is presented, gay. I note that many of the papers presented at a conference on James in
Montreal in May 2004 were within the discipline of ‘‘queer theory.’’ They explored the
ramifications of James’s supposed queerness and of its inscription, however covertly, in
his writing. Mrs Brookenham says firmly at one point in The Awkward Age that she
knows her daughter’s love for Vanderbank is doomed to be disappointed: ‘‘Poor little
darling dear! . . . he’ll never come to the scratch. And to feel that as I do . . . can only
be, don’t you also see? to want to save her’’ (p. 91). When the Duchess asks Mrs
Brookenham, ‘‘What’s Mr. Vanderbank looking for?’’ meaning looking for as a wife,
Mrs Brook answers, ‘‘Oh, he, I’m afraid, poor dear – for nothing at all!’’ (p. 62). In
another place she tells Vanderbank she must protect Nanda from him because,
‘‘ ‘ . . . so far as they count on you, they count, my dear Van, on a blank.’ Holding
him a minute as with the soft, low voice of his fate, she sadly but firmly shook her
head. ‘You won’t do it.’ ’’ (p. 295). It may be that Vanderbank is a ‘‘blank’’ because he
is incapable of feeling heterosexual desire.
Vanderbank is said repeatedly to inspire a ‘‘sacred terror’’ (e.g., p. 308) in the other
characters. This is perhaps because he is outside and above them, in sovereign
independence. He has, it may be, no wish or ability to participate in the exchanges
of heterosexual love that motivate most of the characters. The warmest and most
natural friendships and conversational interchanges in the novel are, it must be said,
those between men. They often involve Vanderbank, for example in the opening scene
between Mr Longdon and Vanderbank, or in scenes between Vanderbank and Mitchy.
A good bit of spontaneous mutual affection, of relaxed warmth, or of what is called
‘‘homosociality,’’ appears in these scenes. Vanderbank, by contrast, is always uneasy
and embarrassed with both Mrs Brookenham (who wants him for herself) and with
Nanda (who also wants him). A more or less explicit reference to ‘‘the love that dares
not speak its name’’ appears in one place. When Mr Longdon and Vanderbank are
discussing the former’s offer of a fortune to Nanda if Vanderbank marries her,
Vanderbank at one point says, ‘‘Well then, we’re worthy of each other. When Greek
meets Greek –!’’ (p. 265). Well, when Greek meets Greek, then comes the Tug of War,
as the old adage avers, but ‘‘Greek love’’ was also a code name in the fin de siècle for
homosexual love.
This is an exceedingly satisfying answer. It is satisfying because it conforms to a
currently fashionable tendency to explain everything in James by way of his covert
homosexuality. It is satisfying because, since James had not by any means come out of
the closet, he had every reason to keep this explanation for Vanderbank’s refusal
implicit, hinting at it in ways that ‘‘Greeks’’ would recognize. This would be a happy
explanation of why James chose the ‘‘Gyp form’’ for this novel, that is, a form that
prohibited him from ‘‘going behind’’ his characters’ speech and overt behavior, or from
revealing directly their secret desires and distastes. Choosing this form kept Vanderbank (and James himself) safely in the closet, while allowing him to write covertly or
secretly about the effects of homosexuality in a set like Mrs Brookenham’s. Going for
this explanation also satisfies a somewhat prurient desire to ‘‘out’’ James. ‘‘See, here is
another one!’’ the triumphant critic exclaims.
Henry James and ‘‘Focalization’’
133
Only one problem impedes accepting this tempting proffered satisfaction: so much
evidence can be cited from the novel and from outside the novel that goes against this
conclusion. Just as strong a case, based on citations, can be made for arguing that
Vanderbank is straightforwardly heterosexual, but that he is either too proud, too
ethically upright, too unwilling to accept a lifelong indebtedness, to allow himself to
be bribed by Mr Longdon, or that he accepts the social code of his day and finds
Nanda unacceptable, nice though she is, because she knows too much about sex,
knows more and more, until she is, as she says of herself, ‘‘a sort of a little drain-pipe
with everything flowing through’’ (p. 358). James’s initial notes for the novel, in the
Notebooks, supports this reading:
A young man who likes her – wants to take her out of it – feeling how she’s exposed,
etc. . . . The young man hesitates, because he thinks she already knows too much; but all
the while he hesitates she knows, she learns, more and more. He finds out somehow how
much she does know, and, terrified at it, drops her: all her ignorance, to his sense, is
gone.
( James 1987: 118)
The ‘‘sacred terror’’ in this note is on Vanderbank’s side. He is ‘‘terrified’’ of a young
unmarried woman who ‘‘knows,’’ though Nanda, in a touching scene, tells Vanderbank that he inspires in her an indefinable fear. ‘‘I don’t know. Fear is fear,’’ she says
when he asks her to specify what she is afraid of in him (p. 212).
Of course James may have changed his mind when he came to write the novel, or he
may not have known what he was doing, that is, he may have written a queer novel
without intending to do so. The ‘‘sacred terror’’ Vanderbank inspires is, however,
when the phrase first appears in the novel, unmistakably defined as a sexual radiance
that attracts women’s desire. Mitchy turns to Vanderbank, when they are discussing
Nanda’s infatuation with Vanderbank and refusal of Mitchy, and tells him, ‘‘The great
thing’s the sacred terror. It’s you who give that out’’ (p. 308). This, in its context,
seems to be saying clearly enough that ‘‘sacred terror’’ refers to Vanderbank’s heterosexual charm, not to his covert homosexuality. In the climactic scene between Nanda
and Mr Longdon, in which she agrees to join him in his country house, she defends
Vanderbank’s refusal by making clear, at least in her understanding, what his reasons
are, and to make absolutely sure that Mr Longdon understands that she is as
Vanderbank thinks she is. Vanderbank’s refusal, she argues, is justified: ‘‘ . . . it’s I
who am the horrible impossible and who have covered everything else with my own
impossibility’’ (p. 541). ‘‘I am like that,’’ she says. And in answer to Mr Longdon’s
‘‘Like what?’’ she says, ‘‘Like what he thinks’’ (p. 543). What that is, the novel has
made clear, is an unmarried young woman who knows too much and therefore cannot
utter traditional marriage vows. Her knowledge disables her performative power,
except for those anomalous performative speech acts that are uttered within a now
‘‘unworked community.’’ Such vows cannot depend on community sanctions for their
efficacy. An example of such a speech act is Nanda’s promise to live permanently with
Mr Longdon.
134
J. Hillis Miller
I conclude that The Awkward Age is what is called ‘‘undecidable’’ in meaning. A set
of incompatible and contradictory answers to the basic question the narrative raises, in
this case the question of why Vanderbank refuses to marry Nanda, can be adduced.
Each can be supported by citations, but no decisive evidence is given endorsing a
choice among them. According to one reading, Vanderbank’s refusal arises from his
covert homosexuality. According to the other, it arises from his being too proud to be
bribed and from his distaste for Nanda because she knows too much about sex. I claim
that a right reading of The Awkward Age reaches this undecidability as its conclusion.
My phrase ‘‘unworked community’’ is an allusion to Jean-Luc Nancy’s book of that
name (Nancy 1991). An unworked (or inoperative) community is made up of a
congregation of singularities who cannot ever fully understand one another. This
collection cannot be gathered together in a traditional community of individuals who
are like one another and who understand one another because they share most
assumptions. One evidence for the breakdown of traditional community in The
Awkward Age, the kind of community assumed in most Victorian novels, is the way
speech acts, guaranteed in J. L. Austin’s (1962) speech act theory by the existence of a
community with firm institutions, all tend to misfire or to be ‘‘infelicitous’’ (Austin’s
word) in this novel. Many Victorian novels end with a community-sanctioned and
‘‘felicitous’’ ‘‘I do’’ uttered in a marriage ceremony. This happy marriage joins the hero
and the heroine. It passes on an inheritance or a title to the next generation, that is, to
the children who will be born of the marriage. Nothing of this sort happens in The
Awkward Age. The novel is full of people who have betrayed their marriage vows or
are about to do so. It could be described as a novel about the breakdown of marriage in
fin de siècle English culture. Vanderbank will not sign, so to speak, the contract or
compact Mr Longdon offers him. Nanda will die unmarried.
This breakdown of community in The Awkward Age corresponds to a breakdown of
the usual form of Victorian fiction. A so-called omniscient narrator tells most of those
stories. That narrator, as in George Eliot’s novels or in Anthony Trollope’s, is a
spokesperson for the collective consciousness of the community. This narrator can
interpret the characters for the reader. He (or she, or it) can utter wise commentary on
the meaning of the story. This commentary guides the reader’s understanding and
evaluation. The narrator can penetrate the deepest recesses of consciousness in the
characters and report to the reader, in free indirect discourse, just what the characters
at a given moment were thinking and feeling. This happens by way of what Nicolas
Royle correctly identifies as an uncanny species of telepathy (Royle 2003: 256–76).
All that has almost completely vanished in The Awkward Age. No going behind.
No wise omniscient narrator who explains everything and speaks for a collective
wisdom. Whatever may be the proprieties of Gyp’s use of a stichomythic dialogue
form with a minimum of intervention by a narrator, in its relation to the stories told
in her novels, the form of The Awkward Age, as James adopted and altered Ibsen’s
form, or Gyp’s, corresponds to its substance. That substance, or central object, or
situation, can be defined in the broadest terms as the ‘‘unworking’’ of community in
late nineteenth-century middle- and upper-class England.
Henry James and ‘‘Focalization’’
135
What about the readers of The Awkward Age? Do they not form a secret and
dispersed community of those who have had a shared experience? It is true that a fairly
large number of recent essays about the novel exist. If they did not tend to disagree
with one another quite sharply, they might be taken as evidence that the novel has
generated a community. In James’s own time, in any case, no such community existed.
James ruefully reports, in a sad confession, as sad, almost, as Nanda’s renunciation of
any hopes of marriage, that the publisher of the book version of the novel, after James
waited vainly for news of how his work was selling, reported to him when asked: ‘‘I’m
sorry to say the book has done nothing to speak of; I’ve never in all my experience seen
one treated with more general and complete disrespect’’ (p. xv).
NOTES
1 Citations from this novel will henceforth be
indicated by page numbers only.
REFERENCES
AND
Abrams, M. ([1953] 1971). The Mirror and the
Lamp: Romantic Theory and the Critical Tradition.
New York: Oxford University Press.
Austin, J. L. (1962). How To Do Things With Words.
Oxford: Oxford University Press.
Genette, G. ([1972] 1980). Narrative Discourse: An
Essay in Method, trans. J. E. Lewin. Ithaca, NY:
Cornell University Press.
Herman, D. (1994). ‘‘Hypothetical Focalization.’’
Narrative 2(3), 230–53.
James, H. (1971–9). The Novels and Tales, 26 vols,
reprint of the New York Edition. Fairfield, NJ:
Augustus M. Kelley.
2 ‘‘Charmed’’ is another word in the vocabulary
range of ‘‘uncanny,’’ ‘‘arcane,’’ and ‘‘cabalistic.’’
FURTHER READING
James, H. (1974–84). Letters, 4 vols, ed. Leon
Edel. Cambridge, MA: Harvard University
Press.
James, H. (1987). The Complete Notebooks, ed. Leon
Edel and Lyall H. Powers. New York: Oxford
University Press.
Nancy, J.-L. (1991). The Inoperative Community,
ed. P. Connor, trans P. Connor, L. Garbus,
M. Holland, and S. Sawhney. Minneapolis: University of Minnesota Press.
Royle, N. (2003). The Uncanny. Manchester, UK:
Manchester University Press.
8
What Narratology and Stylistics
Can Do for Each Other
Dan Shen
On the surface the narratological distinction between story and discourse seems to
match stylistics’ distinction between content and style. ‘‘Discourse’’ refers to ‘‘how the
story is told’’ and ‘‘style’’ to ‘‘how the content is presented.’’ But in this essay I will
argue that (1) the surface similarity conceals an essential difference and (2) recognizing that difference reveals the necessity and the value of synthesizing narratological
and stylistic approaches to ‘‘how narrative is presented.’’ I will also demonstrate the
payoff of this interdisciplinary approach by analyzing a short story by Ernest Hemingway and offer suggestions for future studies.
Differences Between ‘‘Discourse’’ and ‘‘Style’’
The relation between narratology’s ‘‘discourse’’ and stylistics’ ‘‘style’’ is one of superficial similarity and essential difference because discourse is primarily concerned with
modes of presentation that go beyond strictly linguistic matters, and style is in
general concerned more narrowly with choices of language. The narratological distinction between story and discourse is one between ‘‘what’’ is told and ‘‘how’’ to
transmit the story (Chatman 1978: 9, see also Shen 2001, 2002); similarly, the
traditional stylistic distinction between content and style is one between ‘‘what one
has to say’’ and ‘‘how one says it’’ (Leech and Short 1981: 38). In his Linguistics and the
Novel, Roger Fowler writes:
The French distinguish two levels of literary structure, which they call histoire [story]
and discours [discourse], story and language. Story (or plot) and the other abstract
elements of novel structure may be discussed in terms of categories given by the analogy
of linguistic theory, but the direct concern of linguistics is surely with the study of
discours.
(Fowler [1977] 1983: xi, original italics)
Narratology and Stylistics
137
But, in effect, the ‘‘discours’’ in French narratology is to a large extent different from
what Fowler calls ‘‘the language proper of fiction’’ (ibid.), that is, from ‘‘style’’ in
stylistics. There is an implicit boundary separating the two, with a limited amount of
overlap.
In narratology, the most influential work on discourse is Gérard Genette’s Narrative
Discourse ([1972] 1980), which classifies ‘‘discourse’’ into three basic categories: tense
(the relation between story time and discourse time), mood (forms and degrees of
narrative representation), and voice (the way in which the narrating itself is implicated in the narrative). The first category ‘‘tense’’ comprises three aspects: order,
duration, and frequency. In terms of ‘‘order’’ (the relation between the chronological
sequence of story events and the rearranged textual sequence of the events), the
analysis is conducted both on the microstructural and the macrostructural levels. At
the micro level, the object of analysis is a short episode, which is classified into
temporal sections according to the change of position in story time. Genette’s main
concern, however, is the macro level, at which Proust’s Recherche is divided into a
dozen temporal sections, some lasting for more than 200 pages. When analysis is
conducted on such an abstract level, linguistic features simply become irrelevant.
In Genette’s discussion of narrative order, he focuses on various kinds of ‘‘anachrony,’’ that is, discordance between the two orderings of story and discourse, such as
analepsis (flashback) and prolepsis (flash-forward). Such anachronies fall outside the
concern of stylistics with only one exception, namely beginning in medias res (beginning the story from the middle), an anachrony that is often related to the use of
definite expressions. A case in point is the first sentence of Hemingway’s ‘‘The Short
Happy Life of Francis Macomber’’: ‘‘It was now lunchtime and they were all sitting
under the double green fly of the dining tent pretending that nothing had happened.’’
The use of ‘‘they’’ without any antecedent, the use of the definite article ‘‘the’’ and the
reference to something that had happened prior to this ‘‘now’’ all indicate that
Hemingway begins in medias res.
In Style in Fiction, a seminal work in the stylistics of fiction, Geoffrey Leech and
Michael Short devote quite some attention to sequencing on the microstructural level.
They offer a distinction between three kinds of sequencing: presentational, chronological, and psychological (1981: 176–80, 233–9). An example of chronological
sequencing is ‘‘The lone ranger saddled his horse, mounted, and rode off into the
sunset’’ (versus ‘‘The lone ranger rode off into the sunset, mounted, and saddled his
horse’’), which is apparently a matter of syntactic ordering. As for psychological
sequencing, a case in point goes as follows:
Gabriel had not gone to the door with the others. . . . A woman was standing near the
top of the first flight, in the shadow also. He could not see her face but he could see the
terracotta and salmon-pink panels of her skirt which the shadow made appear black and
white. It was his wife.
( James Joyce, ‘‘The Dead,’’ from Leech and Short 1981: 177)
138
Dan Shen
Here the readers ‘‘seem to be with Gabriel, looking up the stairs towards a vague
figure in the shadow, face hidden. . . . The effect [of the psychological sequencing]
would have been nullified if Joyce had begun his third sentence: ‘His wife was
standing . . . ’ ’’ (Leech and Short 1981: 177–8). This is essentially a matter of the
author’s choice of words in reflecting a particular point of view. It is worth noting that
on the micro plane, the examples Leech and Short have chosen are usually in the mode
of scenic presentation, with only one temporal position ‘‘now.’’ These examples, that is
to say, do not involve the reordering of the different temporal positions of past,
present, and future, but involve rather different ways of using language to create
different effects.
The second temporal aspect of ‘‘discourse’’ is ‘‘duration’’ (narrative speed), which is
defined by the relationship between the actual duration of the events and textual
length (Genette 1980: 87–8). In Proust’s Recherche, the ‘‘range of variations’’ goes from
‘‘150 pages for three hours to three lines for twelve years, viz. (very roughly), from a
page for one minute to a page for one century’’ (Genette 1980: 92), a kind of variation
that surely goes beyond linguistic features. A narrative, in Genette’s view, ‘‘can do
without anachronies, but not without anisochronies [accelerations or slowdowns], or,
if one prefers (as one probably does), effects of rhythm’’ (pp. 87–8, original emphasis).
Such narrative ‘‘rhythm’’ (normal speed, acceleration, slowing-down, ellipsis, pause) as
investigated by narratologists is essentially different from the verbal rhythm that
stylisticians are concerned with, the latter being a matter of the features of words and
their combination (e.g., the alternation between stressed and unstressed syllables, the
use of punctuation, the length of words, phrases, sentences). Indeed, to a narratologist,
no matter what words describe an event, the narrative speed will remain unchanged as
long as those words take up the same textual space. A stylistician, on the other hand,
will concentrate on what words are used to describe an event, while hardly paying
attention to the narrative ‘‘rhythm’’ involved.
The last temporal aspect of narratologists’ ‘‘discourse’’ is ‘‘narrative frequency.’’
A narrative ‘‘may tell once what happened once, n times what happened n times,
n times what happened once, once what happened n times’’ (Genette 1980: 114).
Now, whether to tell an event once or more than once is not a choice of language
itself, hence also beyond the concern of stylistics proper.
Of the above-mentioned tripartite distinction of discourse – tense, mood, and voice
– the latter two categories have more to do with the linguistic medium, especially
focalization (point of view)1 and modes of speech presentation, and have hence
attracted attention from both narratologists and stylisticians. But even here, some
elements are nonlinguistic in essence. As for different types of narration, the narratological ‘‘analysis of narrators emphasizes the structural position of the narrator vis à vis
the story he narrates rather than a linguistic characteristic like grammatical person’’
(Rimmon-Kenan 1989: 159), the latter being the concern of stylisticians. Indeed, two
narrators with the same structural position may speak in radically different ways due
to different language choices. But only stylistics takes into account the different
language choices.
Narratology and Stylistics
139
Narratology’s discourse is sometimes also concerned with characterization, especially the different modes of characterization such as direct definition, indirect
presentation, and reinforcement by analogy (Rimmon-Kenan 2002: 59–71). While
narratology is concerned with what counts as direct definition and what structural
function it serves, stylistics focuses on what specific words are used in describing a
character and what effects those words convey as opposed to other potential choices.
As for indirect presentation, Rimmon-Kenan makes a structural distinction between
(different categories of) action, speech, external appearance, and environment. Stylistics usually takes for granted such narrated action, appearance, and environment, and
proceeds to investigate what words the author has chosen to represent those ‘‘fictional
facts’’ (for a purposeful exception, see Mills 1995: 159–63).
Like narratology, stylistics has developed with the times and become increasingly
interested in reader and context, though its interests reflect developments more in
linguistics than in critical theory. By now, we have numerous branches of stylistics,
such as literary (practical) stylistics, functional stylistics, discourse stylistics, critical
discourse analysis, feminist stylistics, literary pragmatics, and cognitive stylistics,
among others (see Wales 2001). Whatever the linguistic model or critical frame
adopted, whatever the definition of style, and whatever the relation to reader and
context, stylistics is in general marked by a concentration on the functions and effects
of language features.
Reasons for the Boundary
The boundary between ‘‘discourse’’ and ‘‘style’’ is in part a result of the different ways
in which narratology and stylistics relate to poetic analysis. The stylistic analysis of
prose fiction is not much different from the stylistic analysis of poetry. Both focus on
the use of language, a use manifested in different forms. By contrast, narratological
analysis of prose fiction has departed from the poetic analytical tradition, focusing
attention on the relation between story events and their rearrangement. In investigating prose fiction, stylisticians have adopted the Prague school’s concept of ‘‘foregrounding,’’ a concept initially based on the investigation of poetry. Not surprisingly,
the concept of ‘‘foregrounding,’’ as a matter of psychological prominence due to
‘‘deviations from the expected or ordinary use of language’’ (Stockwell 2002: 14), has
not entered the realm of narratology. What figures prominently in narratology is the
concept of ‘‘anachrony’’ as mentioned above, which takes the form of various kinds of
deviation from the causal chronological sequence of events.
Another fundamental reason for the boundary between ‘‘discourse’’ and ‘‘style’’ is
that narratology and stylistics have established different relations to the discipline of
linguistics. Stylistics uses the findings of linguistics in the analysis of verbal texts. It
both gains analytical strength from linguistics and, at the same time, is subject to the
confinement of linguistics – to verbal texts and to the use of language itself. By
contrast, narratology uses the findings of linguistics metaphorically. As mentioned
140
Dan Shen
above, Genette has adopted the linguistic term ‘‘tense’’ to cover ‘‘order,’’ ‘‘duration,’’
and ‘‘frequency.’’ In terms of ‘‘order,’’ anachronies often do not have to do with tense
(e.g., ‘‘When she first went to school . . . ’’ or ‘‘Five years later, I saw him again’’). It is
important to note that verbal tense normally goes with the natural temporal facts
(e.g., past tense is used to describe past happenings), but Genette’s ‘‘anachrony’’
(‘‘flashback’’ or ‘‘flash-forward’’) concerns how the discourse deviates from the natural
sequence of story events. In this sense, the relation between Genette’s ‘‘anachrony’’ and
verbal tense change is essentially one of opposition rather than similarity. And
absolutely no real similarity can be perceived between verbal tense and ‘‘duration’’
or ‘‘frequency’’ as such. The relation between the grammatical term ‘‘mood’’ and
‘‘mood’’ in narrative discourse (narrative distance and focalization) is no less metaphorical. Similarly, the area of ‘‘voice’’ – mainly in the form of levels and types of
narration – is miles apart from the grammatical category of ‘‘voice’’ (active voice
versus passive voice).
Across the Boundary
It is worth noting that the boundary between ‘‘style’’ and ‘‘discourse’’ has been crossed
by some especially broad-minded scholars. Of the two camps of narratologists and
stylisticians respectively, the interdisciplinary attempts have come mostly from the
stylistic side. On the narratological side, Rimmon-Kenan in 1989 published a
provocative and insightful paper ‘‘How the Model Neglects the Medium,’’ which
puts forward the ‘‘counterintuitive’’ claim that ‘‘the exclusion of language’’ constitutes
a fundamental reason for the crisis of narratology. But Rimmon-Kenan treats language in a rather different way:
But what exactly do I mean by ‘‘language’’? I have in mind two senses of the term: 1)
language as medium, or the fact that the story in question is conveyed by words (rather
than by cinematic shots, mimed gestures, and the like); 2) language as act, or the fact
that the story in question is told by someone to someone, and that such telling is not
only constative but also performative.
(Rimmon-Kenan 1989: 160)
A note is added to explain the first sense of language: ‘‘This paper takes a different
direction from the one chosen by Phelan [in Worlds from Words (1981)], who says:
‘Nevertheless, my questions about the medium of fiction shall not lead out toward
questions about the similarities and differences among language and other representational media, but shall lead in toward questions of style’ (6–7)’’ (1989: 164 n. 4).
Due to the ‘‘different direction,’’ attention is shifted from ‘‘style’’ to certain properties
of language itself: its linearity, its digital nature, its differential character, its
arbitrariness, its indeterminacy, its iterability, and its abstract nature. With the
emphasis set on such ‘‘technical and semiotic properties of the media themselves,’’
Narratology and Stylistics
141
Rimmon-Kenan (p. 162) purposefully overlooks the problem of style. As for the
second sense of language – language as act – what comes under focus is the general
function and motivation of narration, rather than style. Not surprisingly, RimmonKenan does not draw on stylistics.
The few interdisciplinary publications on the narratological side have typically
come from literary linguists such as Monika Fludernik (1996, 2003) and David
Herman (2002). The two established narratologists started their careers doing both
stylistics and narratology, but gradually moved to the narratological side, bringing
with them their remarkable expertise in stylistics or linguistics, an expertise that is
still evident in their narratological books and essays. The fact that stylistics has not
exerted much influence on other narratologists may be accounted for mainly by the
following: (1) narratologists’ conscious or unconscious ‘‘exclusion of language’’; (2) the
linguistic technicalities in stylistics; (3) the fact that although stylistics has flourished
in Britain (also on the Continent and in Australia), its development has been limited
in America, where, however, the majority of narratologists of the English-speaking
world are based.
As for the stylistic camp, numerous interdisciplinary attempts have appeared,
including Paul Simpson (1993), Sara Mills (1995), Jonathan Culpeper (2001), and
Peter Stockwell (2002).2 Those works are typically marked by multidisciplinarity,
incorporating various approaches, but still retaining a distinctive stylistic identity
because of the focus on language and the dependence on linguistics. Not surprisingly,
since the case is one of stylistics drawing on narratology, narratological models or
concepts are typically used as frameworks for investigating the functioning of language.
It is worth mentioning that recent interdisciplinary attempts tend to draw on
cognitive linguistics or cognitive science. Cognitive stylistics (cognitive rhetoric,
cognitive poetics), which did not come into being until the 1990s, is developing
fast, parallel to the cognitive turn on the narratological side (see Herman 2002,
Bortolussi and Dixon 2003). The complementary relation between stylistics and
narratology may be inferred from the following observation made by Stockwell in
his Cognitive Poetics:
This view of schema theory in a literary context points to three different fields in which
schemas operate: world schemas, text schemas, and language schemas. World
schemas cover those schemas considered so far that are to do with content; text schemas
represent our expectations of the way that world schemas appear to us in terms of their
sequencing and structural organisation; language schemas contain our idea of the
appropriate forms of linguistic patterning and style in which we expect a subject to
appear. Taking the last two together, disruptions in our expectations of textual structure
or stylistic structure constitute discourse deviation, which offers the possibility for
schema refreshment. . . . However, the headers and slots within schemas and the tracks
through schemas can also be discussed in terms of their stylistic and narratological
features.
(Stockwell 2002: 80–2, original italics and boldface)
142
Dan Shen
Here we can see clearly the distinction between story (the content area) and discourse
(the two presentational areas), although, of course, the schemas in the three areas will
be working simultaneously and interacting with each other in the interpretive
process. Significantly, the ‘‘discourse’’ consists of two areas: the textual/organizational
(narratology’s discourse)3 and language/linguistic (stylistics’ style). This is directly
related to the point I’m driving at: since the level of presentation contains both
organizational (narratological) and language (stylistic) choices, focusing only on one
aspect will result in a partial picture of ‘‘how the story is presented.’’ In order to gain a
fuller picture of narrative presentation, it is both desirable and necessary to combine
the concerns of narratology and stylistics.
To illustrate the point, I would like to offer an interdisciplinary analysis, within the
limited scope of the essay, of a mininarrative by Ernest Hemingway:
They shot the six cabinet ministers at half past six in the morning against the wall of a
hospital. There were pools of water in the courtyard. There were wet dead leaves on the
paving of the courtyard. It rained hard. All the shutters of the hospital were nailed shut.
One of the ministers was sick with typhoid. Two soldiers carried him downstairs and
out into the rain. They tried to hold him up against the wall but he sat down in a
puddle of water. The other five stood very quietly against the wall. Finally the officer
told the soldiers it was no good trying to make him stand up. When they fired the first
volley he was sitting in the water with his head on his knees.
(Hemingway [1925] 1986: 51)
This is one of the vignettes that constitute the Paris edition of in our time (1924), and
that appear as interchapters in In Our Time (first published 1925). This vignette is
based on an actual historical event: the atrocious executions of six Greek cabinet
ministers, including the ex-Premier, in Athens in 1922 after Greece’s unsuccessful
campaign against Turkey (see Simpson 1996: 120–2).
Narratologically, a notable feature is what Genette calls ‘‘repeating’’ narration:
narrating twice what happened only once. The narrative begins with a summary of
the event, then, after a description of the setting, moves to a scenic re-presentation of
the same event. The more condensed presentation interacts with the more detailed
presentation so that they reinforce each other.
While reading the vignette, readers feel a strong tension between the terrifying
nature of the executions and the naturalness as conveyed by the presentation. This
tension results in part from a narratological feature: the detached and merely observing position of the external narrator, and in part from a stylistic feature: the neutral
reporting style chosen by Hemingway. The initial summary ‘‘They shot the six
cabinet ministers at half past six in the morning against the wall of a hospital’’ sounds
matter-of-fact, as if the executions were an ordinary event, one that can be reported on
without the need of evaluating or characterizing it. Compare Hemingway’s vignette
with a newspaper account of the same event Hemingway had read: ‘‘ATROCITIES
MARKED GREEK EXECUTIONS OF FORMER LEADERS,’’ which contains
phrases like ‘‘To begin the horrors of that morning . . . a ghastly line . . . an appalling
Narratology and Stylistics
143
instance . . . ’’ (quoted from Simpson 1996: 121, my emphasis). When Hemingway
wrote this short story, he had already experienced the atrocities of World War I, which,
coupled with his experience as a reporter covering crime stories, led him to focus
on war, bull-fighting, and murder in the vignettes. Indeed, Hemingway’s world
is a world of violence and death, where atrocities are treated as commonplace. As
in this vignette, Hemingway often uses the discrepancy between the horror of
what is represented and the matter-of-fact quality of how it is presented to call
attention to that horror. The effect will be even more striking in relation to readers
who are not accustomed to such inhuman atrocities and who are more easily
shocked by such a ‘‘commonplace treatment’’ of the illegal executions of cabinet
ministers.
While the narration is detached, the readers are immediately drawn into the story
by a notable stylistic feature at the beginning: the definite expressions ‘‘They’’ and ‘‘the
six cabinet ministers.’’ Suspense is created leading to a series of questions: who were
‘‘they’’? Who were ‘‘the six cabinet ministers’’? In which part of the (fictitious) world
did this happen? In which year (month, date) did this (fictitious) event take place?
Although Hemingway wrote the story only three months after the historical event,
readers could not get access to the published story until two or more years later, and
most readers, especially later ones, may not associate this vignette with the historical
event. Since the answers cannot be found in the text, the event takes on a sense of
universality due to the lack of specification of time, place, and even nationality (the
nationalities of the ministers and the soldiers remain unspecified). Hemingway, by
suppressing such information, seems to suggest that such murders can take place
anywhere and any time in the world.
Interestingly, in the first sentence, there appears an indefinite article ‘‘a,’’ which
conflicts with the preceding definite expressions. Compare the more natural:
They shot the six cabinet ministers at half past six in the morning against the wall of the
hospital.
A group of soldiers shot six cabinet ministers . . . against the wall of a hospital.
In these more natural contexts, ‘‘hospital’’ appears psychologically less prominent. In
Hemingway’s version, the indefinite article ‘‘a’’ in the ‘‘definite context’’ signals that
‘‘a hospital’’ is a piece of new information that is not to be taken for granted. The
deviant choice of the indefinite article ‘‘a’’ seems to emphasize the point that the
executions do not take place on an ordinary execution ground but in a hospital that is
supposed to save lives. The clash of the nature of the event and that of the setting
heightens the cruelty of the executions in an unobtrusive way. In fact, the historical
executions were carried out at a place ‘‘about one and a half miles outside of the city’’
at about noon time (quoted from Simpson 1996: 121). The ex-Premier, who was sick,
was taken there from a hospital, while the other five ministers were taken from a
prison. In the Hemingway text, all six are shot dead ‘‘against the wall of a hospital’’ at
dawn when life just begins to wake up, resulting in a strong ironic effect.
144
Dan Shen
In the scenic re-presentation of the event, most of the textual space (five out of
six sentences) is devoted to the minister sick with typhoid. In the historical event,
one minister died on the way to the execution ground but his dead body was still
‘‘propped up’’ and executed together with the live ministers. Why does Hemingway
omit this highly inhuman fact and only give the sick minister such structural
prominence? The reasons seem to be both personal and artistic. Hemingway was
severely wounded while fighting on the western front in Italy, and this traumatic
experience may have rendered him particularly attentive to and compassionate towards the sick. In this light, we may appreciate better why Hemingway has chosen
‘‘a hospital’’ as the setting for the murdering. Artistically, Hemingway’s structural
highlighting of the sick minister conveys in a very effective way the inhuman
nature of the executions. In the historical event, the sick minister was made to
stand up in front of the firing party through injections of strychnine. In the Hemingway version, we have instead the climactic progression from ‘‘Two soldiers carried
him . . . ’’ (the sick man could not walk) to ‘‘They tried to hold him up against the
wall . . . ’’ (the sick man could not stand up) and, finally, to ‘‘he was sitting in the water
with his head on his knees’’ (the sick man could not even hold his head while sitting).
The last clause occupies the end-focus position of the text and is psychologically very
prominent. It may arouse strong indignation and compassion on the part of readers:
how can one bear to shoot a man who cannot even sit up? As for the only sentence
devoted to the rest of the victims – ‘‘The other five stood very quietly against the
wall’’ – it seems to convey, in a highly economic manner, Hemingway’s heroic code of
behavior: face death and destruction calmly and undauntedly. This, on the one hand,
counterpoints the miserable sight of the sick man and, on the other hand, may
produce a startled and admiring reaction through conflicting with the expectations
of readers.
The most notable narratological feature of the vignette, however, is what Genette
calls ‘‘descriptive pause’’: scenic description from the perspective of the external
narrator, a description that takes up textual space but does not take up story time,
hence the ‘‘pause’’ of story time. This descriptive pause, which is sandwiched between
the summary and the detailed re-presentation of the event, takes up one fourth of the
whole textual space:
There were pools of water in the courtyard. There were wet dead leaves on the paving of the
courtyard. It rained hard. All the shutters of the hospital were nailed shut. (my emphasis)
Of the other vignettes functioning as interchapters in In Our Time, action and dialogue
predominate and pure scenic description is kept to the minimum. Indeed, of the other
eight vignettes narrated in third person, six do not contain any pure scenic description
and the remaining two contain much less in textual proportion. Stylistic analysis can
help explain why Hemingway devotes so much textual space to a pure scenic
description in this vignette. In this structurally prominent descriptive pause, three
stylistic features are foregrounded: (1) the redundant repetition of ‘‘there were’’ and
Narratology and Stylistics
145
‘‘the courtyard,’’ (2) the abnormal syntactic boundary between the first and the second
sentence, and (3) the reversal of the cause-and-effect relationship. Compare:
It rained hard. There were pools of water on the ground and wet dead leaves on the
paving of the courtyard.
This paraphrase is just a straightforward description of the setting, but Hemingway’s
original takes on additional symbolic meaning. As will soon become clear, the
foregrounded repetition and redundancy function to draw attention to the symbolic
association of ‘‘pools of water’’ with ‘‘pools of blood,’’ and ‘‘dead leaves’’ with ‘‘dead
bodies.’’ In English, one may say ‘‘All my old buddies were gone. I was like the last
leaf on the tree’’ or ‘‘as insignificant and slight as an autumn leaf.’’ The six ministers
were important figures in human history, but they are now treated as useless leaves to
be got rid of. Their dead bodies are no more significant than dead leaves. This
apparently has to do with Hemingway’s nihilistic view of the world. Significantly,
the clumsy verbal repetition and the syntactic boundary add to the semantic weight of
the two symbolic objects, an effect reinforced by placing ‘‘It rained hard’’ after rather
than before the two sentences involved. In the paraphrase offered above, the order is
reversed, a change that increases the psychological prominence of ‘‘It rained hard’’ and
reduces the symbolic weight of ‘‘pools of water’’ and ‘‘dead leaves.’’ Moreover, in the
Hemingway version, the repetition of ‘‘the courtyard’’ signals that what we have are
not ordinary ‘‘pools of water’’ or ‘‘dead leaves’’ like those elsewhere, but ‘‘pools of
blood’’ and ‘‘dead bodies’’ in the courtyard where the bloody executions take place. Furthermore, the deviant syntactic ordering placing the effect first and the cause next
functions to loosen the natural cause and effect relationship (it is not just a matter
of rain leading to . . . ), signaling and highlighting in a very subtle manner the
symbolic significance of ‘‘pools of water’’ and ‘‘dead leaves.’’ In this light, ‘‘wet dead
leaves’’ may be viewed as symbolizing dead bodies soaked in blood. And in this light,
we may understand better the related repetition: ‘‘he sat down in a puddle of water’’ and
‘‘he was sitting in the water with his head on his knees’’ – a pathetic view of a sick man
huddled up in blood. While the courtyard symbolizes a bloody killing ground, the
hospital itself seems to symbolize a big tomb: ‘‘All the shutters of the hospital were
nailed shut.’’ These symbols of setting unobtrusively counterpoint the repeated
presentation of the executions, greatly intensifying the effect.
All in all, in order to appreciate ‘‘how the story is presented’’ in Hemingway’s
vignette, we need to carry out both narratological and stylistic analyses. The former
focuses on the structural techniques, such as the mutual reinforcement between the
curt summary and the more detailed scenic re-presentation, which together interact
with the structurally prominent descriptive pause sandwiched in between; the detached observing position of an external narrator; the structural highlighting and
progressive description of what happens to the sick man, which counterpoints the
brief description of the other victims. The stylistic analysis, on the other hand, focuses
on the initial definite expressions, which form a contrast with the following indefinite
146
Dan Shen
article ‘‘a’’; the lack of verbal specification of time, place, and nationality of the
characters; the neutral style and matter-of-fact tone; the foregrounded verbal
repetition, sentence boundary, and deviant syntactic sequencing in the descriptive
pause. The narratological features and stylistic features interact and reinforce each
other, and it is necessary to see their interaction in order to understand the ‘‘how’’ of
Hemingway’s art.
Current Practices and Future Studies
In view of the implicit boundary between narratology’s ‘‘discourse’’ and stylistics’
‘‘style,’’ and the fact that despite the appearance of numerous interdisciplinary
attempts, the majority of existing publications and courses are purely stylistic or
narratological, I would like to offer the following suggestions for future studies.
In terms of theoretical discussions, it is necessary to give more specific definitions
of stylistics’ ‘‘style’’ and narratology’s ‘‘discourse.’’ As regards narratives in the verbal
medium, it needs to be pointed out in the first place that ‘‘narrative presentation’’ or
‘‘how the story is told’’ consists of two aspects: one organizational and the other verbal,
with a certain amount of overlap in between. Thus, ‘‘style’’ may be defined as ‘‘the
language aspect of how the story is presented.’’ Accordingly, ‘‘stylistic features’’ will
be understood as choices of verbal form or verbal techniques. On the narratological
side, while ‘‘discourse’’ can still be defined as ‘‘how the story is presented’’ or ‘‘the
signifier, statement, discourse or narrative text itself’’ (Genette 1980: 27), it is
necessary to point out that in narratological investigations of ‘‘discourse,’’ attention
is focused on the structural organization of story events, leaving aside style or
language choices. Accordingly, ‘‘narratological features’’ will be understood as narrative strategies or organizational techniques.
Moreover, in a stylistic book, it would be productive to refer to narratology’s
concern with narrative strategies. Since many stylisticians are not yet narratologically
informed, it is necessary to promote the introduction of narratology to stylisticians.
Interestingly, in Britain where stylistics has been thriving, the development
of narratology has been much slower than in America, despite the fact that the
British-based Poetics and Linguistics Association and its official journal Language
and Literature have played an important role in promoting the interface of stylistics
and narratology.
In a narratological book, it would be helpful to delineate the scope of inquiry for
readers, and to point out that in order to gain a fuller view of ‘‘how the story is
presented’’, more attention needs to be paid to the writer’s ‘‘style,’’ which is well
defined by James Phelan in Worlds from Words: ‘‘ ‘Style’ is the specific term, used to
refer to particular uses of [the language] system. ‘Style’ also has another sense, which
I shall develop more fully later: those elements of a sentence or passage that would be lost in a
paraphrase’’ (Phelan 1981: 6, my emphasis). Narratological investigations often do
not bear on ‘‘those elements of a sentence or passage that would be lost in a
Narratology and Stylistics
147
paraphrase.’’ But as shown by Phelan’s illuminating analyses, those are also very
important elements of ‘‘how the story is presented’’ in many literary narratives.
Indeed, a comparison of Phelan’s Worlds from Words (and works in stylistics) with
his 1996 Narrative as Rhetoric (and other works in narratology) will surely lead to the
double realization that paying close attention to the functioning of language can lead
to fruitful results in narrative analysis, and that in order to gain a fuller picture,
attention should be devoted not only to verbal techniques but also to organizational
techniques. Many of the interdisciplinary attempts as referred to above, such
as Monika Fludernik’s 2003 paper ‘‘Chronology, Time, Tense and Experientiality
in Narrative,’’ well demonstrate the gains of treating narratological concerns
(e.g., chronology) and stylistic concerns (e.g., tense) together.
Now, although the works by Fludernik and Herman exemplify how narratology
can benefit from stylistic or linguistic analysis, narratologists and students majoring
in literature may be put off by the linguistic technicalities in many stylistic publications. However, many other stylistic works are intended for readers and students of
both language and literature with linguistic technicalities kept to the minimum, such
as those in ‘‘The Interface Series: Language in Literary Studies’’ published by Routledge. Or as an alternative, attention may be directed to ‘‘close reading.’’ As we all
know, close reading was prevalent for a time but has been discredited for the past two
decades or so because of its ideological limitations. It has to a great extent been
replaced by stylistics in Britain, but in America, where stylistics itself has been very
much excluded, ‘‘style’’ has been to a great extent overlooked since the ‘‘death’’ of close
reading. Interestingly, on May 15, 2003, the America-based NARRATIVE listserv
posted a call for proposals: ‘‘Henry James and new formalisms’’ which directs attention to close reading: ‘‘Long discredited as a conservative, parochial, and even
‘oppressive’ critical practice, ‘close reading’ is showing signs of return. Yet this return
is marked by considerable anxiety. . . In what sense are New Formalisms new? In what
ways might they return to the New Criticism yet discern its limitations? . . . ’’4
While doing close reading, we certainly should try to eliminate the earlier
limitations. Now, no matter whether it is stylistics or close reading, a languageoriented approach may reveal many things about ‘‘how the story is presented’’ that go
beyond the concerns of the structure-oriented narratology. In view of this fact,
students should be encouraged to take courses and read books in both areas of
study. It is hoped that, with a clear awareness of the complementarity between
‘‘style’’ and ‘‘discourse’’ (as investigated by narratology), more conscious efforts will
be made to combine the concerns of narratology and stylistics/close reading in
narrative criticism.
Moreover, as an extension of Rimmon-Kenan’s earlier call, both narratological and
stylistic investigations of narrative presentation can be enriched by looking out at the
relation between the linguistic medium and other media, as can be seen in the
increasing fruits of media studies especially on the narratological side. As regards
the Hemingway vignette, we can look out at how the structural and verbal techniques
can or cannot be conveyed either directly or indirectly on screen. Since film is scenic
Dan Shen
148
in nature, the initial summary can only be replaced by a scene showing, say, soldiers
firing at the six ministers or the six dead bodies after the killing, where the contrast
between the definite expressions and the indefinite ‘‘a’’ can hardly be conveyed,
and where the concealment of nationalities may be difficult to achieve because of
the appearance and clothes/uniform of the persons involved. With this initial scene,
the following re-presentation will function as a flashback. But as distinct from the
natural re-presentation in the Hemingway text, this re-presentation on screen will
appear much less natural, since on screen we can only have a double scenic presentation rather than a summary substantiated by a scene. On screen, while there is no
difficulty in conveying the structural highlighting of the sick minister in the representation, it may be less easy to bring out the symbolic significance of the setting,
as signaled by the joint function of unusual structural prominence, verbal repetition,
deviant syntactic boundary and ordering in the Hemingway text. It is true that a
close-up of ‘‘wet dead leaves’’ may succeed in conveying the symbolic meaning, but
a close-up of ‘‘pools of water’’ may not help and the way out is perhaps to show pools
of water mixed with blood or present analogous scenes of pools of water and pools of
blood.
And finally, in terms of the investigation on ‘‘style,’’ we can further look out, where
applicable, for an interlingual comparison, which may shed interesting light on how
the author’s stylistic choices are related to the properties of the specific language
involved.
NOTES
1 The choice of different modes of focalization is
essentially a structural choice, but different
modes of focalization have different linguistic
indicators, hence attracting attention from
stylisticians.
2 Katie Wales’s A Dictionary of Stylistics also
contains numerous narratological concepts,
implicitly calling for an incorporation of narratological concerns.
REFERENCES
AND
Bortolussi, M. and Dixon, P. (2003). Psychonarratology. Cambridge, UK: Cambridge University
Press.
Chatman, S. (1978). Story and Discourse. Ithaca,
NY: Cornell University Press.
Culpeper, J. (2001). Language and Characterization
in Plays and Texts. London: Longman.
Fludernik, M. (1996). Towards a ‘‘Natural’’ Narratology. London and New York: Routledge.
3 And of course we also have other elements of
textual organization, such as chapter titles,
chapter division, paragraphing, etc.
4 The email was sent by Beartooth <[email protected]> on May 14, 2003, and resent
from [email protected]
Edu on May 15, 2003.
FURTHER READING
Fludernik, M. (2003). ‘‘Chronology, Time, Tense
and Experientiality in Narrative.’’ Language and
Literature 12, 117–34.
Fowler, R. ([1977] 1983). Linguistics and the Novel.
London: Methuen.
Hemingway, E. ([1925] 1986). In Our Time, Collier Books Edition. New York: Macmillan.
Herman, D. (2002). Story Logic: Lincoln and London: University of Nebraska Press.
Narratology and Stylistics
Genette, G. ([1972] 1980). Narrative Discourse,
trans. J. E. Lewin. Ithaca, NY: Cornell University Press.
Leech, G. N. and Short, M. H. (1981). Style in
Fiction. London: Longman.
Mills, S. (1995). Feminist Stylistics. London and
New York: Routledge.
Phelan, J. (1981). Worlds from Words. Chicago: The
University of Chicago Press.
Phelan, J. (1996). Narrative as Rhetoric. Columbus:
Ohio State University Press.
Rimmon-Kenan, S. (1989). ‘‘How the Model Neglects the Medium: Linguistics, Language, and
the Crisis of Narratology.’’ The Journal of Narrative Technique 19, 157–66.
Rimmon-Kenan, S. ([1983] 2002). Narrative Fiction, 2nd edn. London and New York: Routledge.
149
Semino, E. and J. Culpeper (eds.) (2002). Cognitive
Stylistics. Amsterdam and Philadelphia: John
Benjamins.
Shen, D. (2001). ‘‘Narrative, Reality and Narrator
as Construct: Reflections on Genette’s Narration.’’ Narrative 9, 123–9.
Shen, D. (2002). ‘‘Defence and Challenge: Reflections on the Relation Between Story and Discourse.’’ Narrative 10, 422–43.
Simpson, P. (1993). Language, Ideology, and Point of
View. London and New York: Routledge.
Simpson, P. (1996). Language Through Literature.
London and New York: Routledge.
Stockwell, P. (2002). Cognitive Poetics. London and
New York: Routledge.
Wales, K. ([1990] 2001). A Dictionary of Stylistics,
2nd edn. Harlow, UK: Pearson Education.
9
The Pragmatics of Narrative
Fictionality
Richard Walsh
Thoughts about his case never left him now. Several times he had considered whether it
would not be advisable to prepare a written document in his defence and lodge it with
the court. His idea was to present a short account of his life and, in the case of each
relatively important event, explain the reasons for his action, say whether he now
thought that course of action should be condemned or approved, and give his reasons
for this judgement.
(Kafka [1925] 1994: 89)
Josef K., the protagonist of Kafka’s The Trial, finds himself in a situation in which
his efforts to establish his innocence, to explain himself, have no focus and no
boundaries: he can only envisage a plea in the form of an exhaustive narrative of his
life. Our own efforts to make sense of the extraordinary circumstances surrounding
K.’s envisaged narrative act are crucially related to a distinctive fact about Kafka’s:
while K. contemplates autobiography, Kafka engages in fiction. Fiction is usually
understood to have a second-order relation to the real world, via the mimetic logic of
fictional representation: it represents events, or imitates discourses, that we assimilate
through nonfictional modes of narrative understanding. So even where (as here) the
fiction is in some respects unrealistic, it is comprehensible in terms of its relation to
familiar types of narrative: not only the accused person’s effort of self-justification, and
the discourse of moral autobiography, but also psychological narratives of guilt, and
the several kinds of legal narrative that inform the global frame of reference of The
Trial.
On the other hand, the place of narrative in nonfictional contexts such as legal
studies has itself attracted a lot of attention in recent years. H. Porter Abbott’s
Cambridge Introduction to Narrative devotes a whole chapter to ‘‘narrative contestation,’’
for which his paradigm case is the competing narrative efforts of the prosecution and
defense in the notorious 1893 trial of Lizzie Borden, accused (and acquitted) of
murdering her father and stepmother (Abbott 2002: 138–55). Abbott emphasizes
The Pragmatics of Narrative Fiction
151
the way in which the prosecution and defense, in this as in all trials, strive to establish
narrative credibility by aligning their representation of events with rhetorically
advantageous ‘‘masterplots,’’ by which he means familiar skeletal narratives with an
established cultural authority. So Lizzie Borden’s apparent ‘‘lack of affect’’ in the face
of the murders is narrativized by one side as the shock of a virtuous daughter, by the
other as the cold-blooded viciousness of a Lady Macbeth (pp. 147–9). In a legal
context, the issue of narrative truth is especially pointed (Lizzie Borden, after all,
stood in jeopardy of the death penalty): yet the explanatory power of narrative here
depends less upon its relation to fact than its relation to other narratives, and it is in
these terms that both sides make their case. Indeed they can do no other, even in those
self-reflexive moments when they accuse each other of doing so.
The general point here is that all narrative, fictional and nonfictional, is artifice:
narratives are constructs, and their meanings are internal to the system of narrative.
For some theorists this general quality of narrativity subsumes the concept of
fictionality entirely: if all narratives derive their meaning from their relation to
other narratives, rather than any direct purchase upon reality, then it no longer
makes sense to use this second-order kind of relation specifically to characterize
fiction. Yet the awkward fact remains that the narratives elaborated by prosecution
and defense in the Lizzie Borden trial made truth claims that Kafka’s novel does not
make, and accordingly these two cultural modes of narrative invoke quite different
interpretative assumptions. In this respect the rise of the general concept of narrativity, far from superseding the issue of fictionality, has actually exposed it as a
theoretical problem. If the logic of narrative representation does not provide for a
defensible distinction between fiction and nonfiction, then the focus of theoretical
attention is necessarily displaced from the substance of fictional narrative to the act of
fictive narration; from the product to the production of fiction. How are we to
understand fictive narration as a referential act, or as an act of communication? It is
in this context that I want to advocate a pragmatic approach to the issue of
fictionality: one that draws upon philosophical and linguistic fields of inquiry into
the communicative use of language, and in particular invokes the conceptual framework of relevance theory.
Modern accounts of fictionality generally turn upon one or more of a small
repertoire of theoretical gambits, which can be collectively understood as gestures
of disavowal, achieved through several kinds of displacement. That is to say, these
accounts variously respond to the problem of fictionality as a problem of truthfulness,
and resolve it by detaching the fictive act from the domain of truth, that is, from any
codified system of representation or language. The kinds of theoretical move I have in
mind are: the institution of a narrator as the source of the fictive discourse, the
redescription of fictional artefacts as props in a game of make-believe, the notion of
pretended speech acts, and the recuperation of fictional reference as actual reference to
fictional worlds. The first two moves place the language of fiction itself within the
fictional frame; the third move disqualifies it, as nonserious language, from communicative accountability; the fourth move allows the language of fiction to be literal
152
Richard Walsh
and serious but not exactly fictive (that is, to the extent that fictive language does not
make referential commitments), since fictionality has been redefined as a matter of
ontological modality. The nub of fictionality always turns out to be elsewhere: it is as
if fictionality were not a problem except in relation to language. This is odd, given that
language, I take it, is what makes fiction possible – and again by a language here
I mean broadly any codified system of representation, such that fictions in any
medium are equally dependent upon a language, a representational code, and not
merely upon cognitive illusion.
Fictionality, I want to suggest, functions within a communicative framework: it
resides in a way of using a language, and its distinctiveness consists in the recognizably distinct rhetorical set invoked by that use. I assume that narrative fictionality is
worth distinguishing from narrativity in general: that is to say, I want to grant full
force to the claim that all narrative is artifice, and in that very restricted sense fictive,
but maintain nonetheless that fictional narrative has a coherently distinct cultural
role, and that a distinct concept of fictionality is required to account for this role. It is
best explained in functional and rhetorical terms, rather than in formal terms: true,
there are formal qualities strongly associated with fiction, but they do not supply
necessary or sufficient conditions of fictionality. To say instead that fictionality is a
functional attribute is to say that it is a use of language; to say that it is rhetorical is to
say that this use is distinguished by the kind of appeal it makes to the reader’s or
audience’s interpretative attention. No model that treats fictive discourse as framed by
formal, intentional, or ontological disavowal can meet these criteria for a concept of
fictionality. If fictionality consists in a distinct way of using language, it is not
explained by attaching its distinctiveness to some quarantine mechanism conceived
precisely to maintain its conformity with nonfictional usage, at the cost of detaching
it, in one way or another, from its actual communicative context. The rhetorical
distinctiveness of fiction, then, is consistent with a communicative continuity between fictional and nonfictional uses of language. Fictionality is a rhetorical resource
integral to the direct and serious use of language within a real-world communicative
framework.
I want to reformulate the age-old problem of fiction’s claim upon our attention, the
challenge that has prompted various defenses of poesy, or expulsions from republics,
down through the centuries, as the problem of reconciling fictionality with relevance.
The concept of relevance appears, in several quite specific senses, within two distinct
well-demarcated theoretical domains that have been important to recent discussions
of fictionality: one is fictional worlds theory, which focuses upon the referential act;
the other is speech act theory, especially those accounts that engage with Gricean
‘‘conversational implicature’’ (Grice 1989), where the focus is upon the communicative act. Relevance theory itself is a related approach to communication from the
perspective of pragmatics and cognitive linguistics, and although this field of research
has not included any detailed consideration of fictionality, it does suggest that a
pragmatic theory of fiction has much to gain from pursuing the relation between
fictionality and relevance.
The Pragmatics of Narrative Fiction
153
The issue of relevance arises in fictional worlds theories in two respects: the first
and narrowest is internal to a given fictional world, and relates to the problem of
incompleteness; the second is external, and concerns the global pertinence of fictional
worlds to the reader. Incompleteness is a problem for fictional worlds theory because
the text of a fiction cannot be expected to fully specify a world, nor even provide a
sufficient basis for a comprehensive inferential process. There are always going to be
gaps and indeterminacies in the interpretative construction of fictional worlds, which
is a significant divergence from the philosophical model of possible worlds upon
which fictional worlds theory is based, since it is axiomatic that possible worlds are
logically complete. The theoretical response has been to invoke two complementary
recuperative strategies. The first, proposed by Marie-Laure Ryan, is to bring fictional
worlds into line with the logical framework of possible worlds theory by assuming a
‘‘principle of minimal departure’’: the principle of minimal departure dictates that the
world of the text is to be understood as complete, and identical to the actual world
except for the respects in which it deviates from that model, either explicitly or
implicitly, both in its own right and by virtue of any genre conventions it invokes
(Ryan 1991: 51). In this case, then, the fictional world itself is complete after all,
even if the reader’s actualization of it in interpretation of the text is not. But if the
problem really is logical incompleteness, the principle of minimal departure cannot
help: there are indeterminacies in relation to the narrative particulars of any fiction
for which the model of the actual world offers no decisive guidance (exactly how
many times did K. consider the idea of preparing a short account of his life?). On
the other hand, if the principle of minimal departure is assumed just to have a
supplementary role in providing for a ‘‘reasonably comprehensive’’ world (Ryan
1991: 52), it raises a question: how far does the reader pursue the gap-filling process
it licenses? What criterion limits that interpretative pursuit? However we may choose
to define the goals of interpretation, the criterion required is one of relevance to
those goals.
Ryan’s answer to the problem of incompleteness in fictional worlds is cited by
Thomas Pavel, whose alternative solution is to conceive them as worlds of various
sizes, determined by the texts from which they are constructed and open to extratextual information provided for by the principle of minimal departure, without
being ‘‘maximal’’ (Pavel 1986: 107–8). So we might just conceivably infer that K. has
an appendix, in the absence of any textual information to the contrary, but neglect to
infer that his maternal great-grandfather had an appendix – although the latter
inference would be better founded, on the basis of minimal departure, since an
appendectomy would not then have been available. Pavel’s explanation introduces
the crucial notion of relevance: ‘‘Some form of gradual opacity to inference, some
increasing resistance to maximal structures, must be at work in most fictional
worlds, keeping them from expanding indefinitely along irrelevant lines’’ (Pavel
1986: 95). Once the idea of relevance is admitted, however, it entirely supersedes
that of completeness. It makes no practical difference whether the ‘‘facts’’ of a
fictional world are understood maximally, independent of textual interpretation, or
154
Richard Walsh
contingently, on the basis of what it is possible to infer from the text; because in either
case the scope of inference is, in principle, nonfinite. The horizon of the reader’s
encounter with a fiction is determined, not by what it is possible to infer, but by what
is worth inferring: the reader will not pursue inferential reasoning beyond the point at
which it ceases to seem relevant to the particulars of the narrative, in a specific context
of interpretation. This is a pragmatic limit, but only such a limit can provide for the
fact that fictional representations do not merely exist (in whatever qualified sense),
but are communicated.
Such considerations invoke the other sense in which the notion of relevance arises in
fictional worlds theory. Pavel declares a ‘‘principle of relevance’’ as one of only two
fundamental principles of fictional reference (1986: 145); but the kind of relevance he
has in mind here is quite distinct from the internal relevance presupposed by the
discussion so far. Instead, it articulates an external global relation between worlds, and
is therefore contingent upon the reader’s realization of the fictional world. Yet as we
have already seen, that realization itself must be contingent upon relevance criteria of
a quite different order, if it is not to be an endless project. This relation between
worlds, then, is a strangely cumbersome reprise of the reader’s supposed original effort
of world construction, which (under the rubric of ‘‘minimal departure’’) was to be
pursued precisely in terms of difference. Are these two kinds of relevance, internal and
external, ultimately distinct? Or do they, under closer scrutiny, collapse into each
other, in the process extinguishing the concept that intervenes between them, which
is the concept of fictional worlds itself ? In fictional worlds theory, the concept of
relevance is bounded by two assumptions that I want to resist: one is that the ‘‘facts’’
of fiction are meaningfully independent of considerations of relevance; the other is the
idea that relevance can be internal to the fictional world; that it can ever mean
something independently of the communicative act. My counterclaim is that the
reader’s interpretative agenda cannot be understood within the bounds of a fictional
world, or indeed in relation to its fictional existence rather than its actual communication; and that relevance, even when it is described internally as relevance to story,
is always, reciprocally, relevance to the reader.
The issue of communication, of course, is central to the relation between fictionality and speech act theory. The standard speech act account of literary discourse, as
first elaborated by Richard Ohmann (1971) and John Searle (1975), is the imitation
speech act model, in which the authorial speech act is not seriously made, but
pretended, which effectively suspends the appropriateness conditions (or felicity
conditions) normally attaching to the performance of that speech act. But the
imitative model is undermined by the fact that third-person novels routinely deviate
from the norms of any nonfictional, real-world speech act, for instance in such
ordinary narrative strategies as omniscient narration. The pretended speech act
frame does not account for fictionality because the rhetoric of fictionality often
inhabits the narration itself: the first sentence of my quote from The Trial, for
example, resists recuperation as a pretended nonfictional speech act in both content
The Pragmatics of Narrative Fiction
155
(access to the thoughts of another) and form (the dual temporal perspective of internal
focalization, manifested in the otherwise anomalous ‘‘now’’).
A more promising alternative, advanced by Mary Louise Pratt (1977), looks instead
to H. P. Grice’s model of conversational implicature: Grice argued that the appropriateness conditions applicable to speech acts were best understood, not as attaching to
the semantics of specific sentences, but in relation to a few general maxims. These
maxims together constitute a Cooperative Principle, which is the foundation of
successful communication: the crucial ones for my purposes are the Maxim of
Relation, ‘‘be relevant,’’ and the first Maxim of Quality, ‘‘do not say what you believe
to be false’’ (Grice 1989: 27). Grice’s approach allows a great deal of flexibility in the
interpretation of speech acts, because what is actually said may be supplemented by
inferences, or implicatures, in order to maintain the shared assumption that the
Cooperative Principle is in place. Pratt proposes that one way in which speech acts
may be relevant is by being ‘‘tellable,’’ by which she means of intrinsic interest, or
worthy of display: a tellable speech act constitutes an invitation to contemplate, to
interpret, to evaluate. This allows her to propose a distinct category of speech act
called ‘‘narrative display text,’’ which embraces both fictional narratives and the
nonfictional ‘‘natural narratives’’ she cites from the sociolinguistic studies of William
Labov (Pratt 1977: 132–6). These texts adhere to Grice’s Maxim of Relation, and
fulfill the appropriateness conditions of relevance, not by being informative, but by
being exhibitive – by being tellable.
Pratt, however, stops short of addressing the issue of fictionality itself and does not
inquire into the hierarchical relation between Grice’s maxims of Quality and Relation,
truthfulness and relevance. To accommodate fiction, Pratt ultimately falls back upon
the standard speech-act account she had originally rejected, and concedes that fictive
discourse is, on the author’s part, an imitation display text, attributed to a narrator
within the frame of fiction (1977: 173, 207–8). Pratt might have offered an account
of fictive discourse independent of the pretence model: she has a plausible basis for the
relevance of the authorial speech act in the notion of tellability, and comes very close
to seeing tellability as sufficient for a felicitous authorial speech act irrespective of
truth criteria. In short, she makes progress with the aspects of the Gricean model that
emphasize relevance, but is blocked by her assumption that the issue of relevance is
ultimately secondary to that of truthfulness, which precludes any direct authorial
model of fictive discourse.
Pratt’s case leads me on to relevance theory, which I want to introduce in relation to
Grice. Grice’s model of ‘‘conversational implicature’’ was developed in recognition of
the fact that the code model of language was often not sufficient to account for
communication, and needed to be supplemented by an inferential model. The
innovation of relevance theory, as expounded by Dan Sperber and Deirdre Wilson,
is to argue that inference is not a supplementary component of communication, but
its core. For Grice, an utterance is either literally relevant in a given context, or
literally irrelevant, prompting an inferential search for implicatures that might
156
Richard Walsh
redeem the Cooperative Principle. Sperber and Wilson argue that the order of events
in this process of comprehension should be reversed: ‘‘It is not that first the context is
determined, and then relevance is assessed. On the contrary, people hope that the
assumption being processed is relevant (or else they would not bother to process it at
all), and they try to select a context which will justify that hope: a context which will
maximise relevance’’ (Sperber and Wilson 1995: 142). To clarify some of this slightly
technical terminology, a context, here, is a set of assumptions adopted by an individual,
a subset of the individual’s cognitive environment (1995: 15). A cognitive environment is
the total set of facts or assumptions manifest to an individual at a given time (p. 39).
Something is manifest if it is available to perception or inference (p. 39). An assumption
is a thought (a conceptual representation) treated by the individual as true of the
actual world (p. 2). The relevance of a new assumption to an individual is maximized
when its processing achieves an optimal balance between the effort involved and the
contextual effects (or more strictly, the positive cognitive effects) derived (pp. 144,
265). A contextual effect is a modification of the context arising from the interaction
between old assumptions and new assumptions; it is a positive cognitive effect if it
benefits an individual’s cognitive functions or goals (pp. 109, 265). To illustrate,
albeit simplistically, consider a situation in which Dan tells Deirdre that the kettle is
on. Deirdre’s cognitive environment at the time includes all the perceptible physical
phenomena of the office in which she is writing an article, the knowledge of relevance
theory that she is bringing to bear upon the article, the understanding she has gleaned
from past experience about the uses of kettles, her evaluation of Dan’s friendly
disposition and his knowledge of her beverage preferences, and so forth, as well as
the new assumptions made manifest by the words he has just spoken (that he is
talking to her, that he is informing her that he has put the kettle on, etc.). She draws
upon a subset of her old assumptions (kettles are used to make tea; Dan knows she
likes tea, milk but no sugar, etc.) as a context within which to process the implications of these new assumptions (Dan is making tea for her, he expects her to join him
in the kitchen, etc.). In order to maximize the relevance of these new assumptions she
strikes a balance between the effort required to draw those implications (and retrieve
old assumptions) and the cognitive benefits of doing so. She may not bother to draw
some inferences (Dan is within earshot, Dan speaks English) even though they are
strongly manifest, and so easily available, because they have little or no cognitive
effect (she already knew this). On the other hand, she may find it a worthwhile effort
to draw some weakly manifest inferences (Dan is belittling her intellectual staying
power), because the cognitive effects are large (he is not so friendly after all). She
remains tapping away at her keyboard.
The most important consequence of relevance theory, for my purposes, is the new
relation it proposes between the functions of relevance and truthfulness in communication. Relevance theory advances the idea that, for the purposes of communication,
the propositional criterion of truth is a subordinate consideration to the contextual
pragmatic criterion of relevance. This is not to say that the truth or falsehood of
assumptions is a matter of indifference, or even that there are circumstances where it is
The Pragmatics of Narrative Fiction
157
a matter of indifference (as one might be tempted to say, precipitously, is the case
with fiction): an assumption, to be an assumption at all, must be taken as true. But
all assumptions are, to a greater or lesser extent, the products of inference, which is a
pragmatic, relevance-driven process, and the truth of an assumption need not
depend upon the truth of the encoded form of an utterance, or its literal meaning.
Sperber and Wilson offer an extended account of metaphor and irony, which shares
with speech act accounts the assumption that successful communication in such
instances is dependent upon an inferential search for contextual relevance. It does
not, however, present this search as a process resulting in a dichotomy between the
literal sense of ‘‘what is said,’’ which is false, and a recovered implicit meaning, which
is true (1995: 242).
From a relevance theory perspective, the comprehension of figurative language (as
all language) is understood as an inferential process of filling out the linguistic code
until maximal relevance is achieved (that is, up to the point at which the cost in
processing effort exceeds the benefit in contextual effect, for the reader concerned).
Criteria of truth only enter into this process to the extent that truthfulness is a
condition of the particular contextual effects involved, and only apply in relation to
the assumptions producing those effects, which need not include the literal utterance,
or any translation of it, as a proposition: truth criteria are applicable to successful
communication only in the sense that the output of the inferential process, its
cognitive effects, must qualify as information. Relevance to an individual is, definitionally, a measure of cognitive benefit, which Sperber and Wilson generally interpret as an ‘‘improvement in knowledge’’ (Wilson and Sperber 2002: 601), although
they do expressly want to leave open the possibility of taking into account other
kinds of benefit to cognitive functioning. The notion of ‘‘improvement in knowledge’’
itself, however, embraces a wide range of cognitive effects. For instance, Sperber and
Wilson describe a category of cognitive changes they call ‘‘poetic effects’’: an utterance
has poetic effects, in their sense, if it achieves most of its relevance through a wide
array of weak implicatures (1995: 222). That is, the improvement in knowledge
required to achieve relevance may be of an impressionistic or affective nature, yet also
be the cumulative product of many minute cognitive effects, many weakly manifest
assumptions, all of which are outcomes of the process of comprehension, and none
of which is necessarily dependent upon the propositional truth of the input to
that process.
On this basis I want to argue that the problem of fictionality is not, after all, a
problem of truthfulness but a problem of relevance. It is the presumption of relevance,
not any expectation of literal truthfulness, that drives the reader’s search for an
appropriate interpretative context. Relevance theory allows for inference, and the
generation of implicatures, to proceed from an utterance that is clearly false in the
same direct way as for one that is taken as true: evaluations of truth only come into
play in consequence of that process. So the fictionality of a narrative only compromises
the relevance of those assumptions that are contingent upon its literal truth. The
relevance theory model allows for a view of fiction in which fictionality is not a frame
158
Richard Walsh
separating fictive discourse from ordinary or ‘‘serious’’ communication, but a contextual assumption: that is to say, in the comprehension of a fictive utterance, the
assumption that it is fictive is itself manifest. The main contextual effect of this
assumption is to relatively subordinate implicatures that depend upon literal truthfulness in favor of those that achieve relevance in more diffuse and cumulative ways.
Fiction does not achieve relevance globally, at one remove, through some form of
analogical thinking, but incrementally, through the implication of various cognitive
interests or values that are not contingent upon accepting the propositional truth of
the utterance itself; and upon the deployment, investment, and working through of
those interests in narrative form.
There is, certainly, a global, retrospective sense in which narrative can be understood as the suspension of relevance along the line of action, and narrative closure
figures less as the resolution of plot in itself (though it is an effect usually achieved in
terms of plot), than as the resolution of suspended evaluations of relevance. In this
straightforward sense, irrespective of questions of fictionality, narrative form in itself
responds to certain expectations of relevance. K.’s death, at the end of The Trial, is not
just a very emphatic terminal plot event, but also the ‘‘answer’’ to several kinds of
questions raised by the narrative, and in that respect it occasions a range of possible
overall assessments of relevance from the reader (relating, for example, to K.’s moral
desserts and models of justice, whether legal, cosmic, or poetic; to the balance of
power between state and individual, or between structure and agency; to the psychological mechanisms of guilt, and the authority of the superego, and so on). Such
global, thematic relevance is by no means the only kind offered by narratives, nor is it
necessarily the most important; though in fiction, such interpretative logic is likely to
dominate over the kind of factual enrichment of the reader’s cognitive environment
for which nonfictional narratives are better suited. Still, the investment of interpretative effort in the process of reading a fiction requires an ongoing sense of relevance:
there are limits to everyone’s tolerance of delayed gratification, and no ultimate
resolution alone could plausibly justify the effort of reading Proust’s Recherche, or
War and Peace, or Clarissa. The narrative force of fiction depends upon assumptions
carried forward, enriched, modified, reappraised, overturned in the process of reading:
even in fiction, narrative development is only possible on the basis of an established
sense of relevance.
Relevant information, in fiction, is supplied by assumptions with the capacity to
inform a cognitive environment that includes the assumption of fictionality itself, as
well as a set of general assumptions that might be collectively labeled ‘‘narrative
understanding’’ (which would include logical, evaluative, and affective subsets), and
more specific assumptions relating to, for instance, generic expectations of the text in
hand, and the particulars of its subject matter. In this cognitive environment, the
contextual effects that constitute relevance may be produced by new assumptions
informing the project of narrative understanding in general (and the further kinds of
understanding this may facilitate), or by assumptions enabling further inferences from
the narrative particulars, which will themselves contribute to an ongoing cumulative
The Pragmatics of Narrative Fiction
159
experience of relevance (such cumulative effects being analogous to those that Sperber
and Wilson term ‘‘poetic effects’’ in their discussion of how impressions may be
communicated). So a reader of The Trial may find relevance in constructing some of
the subtle hypotheses about psychological motivation needed to comprehend K.’s
behavior; or such comprehension may contribute to an emotional investment of
interest, for which K. becomes the vehicle. In either case, the narrative coherence
that provides for these effects rests upon more manifest assumptions, of a sort that
relates to the familiar idea of what is ‘‘true in the fiction.’’ Such assumptions have the
status of information irrespective of the literal truth value of the utterance, because
their validity, their ‘‘aboutness,’’ is contextual, not referential (though this is not to
exclude the possibility that some of the assumptions made available by fictive
discourse may indeed be referential: as, for instance, in the case of a roman à clef, or
a historical novel, or the many modern forms of documentary fiction).
The notion of truth ‘‘in the fiction’’ does not imply an ontological frame, but a
contextual qualification: assumptions of this kind provide information relative to a
context of prior assumptions. We do not generally attempt to resolve the reference of
fictive utterances because we know in advance that, in the absence of any evidence to
the contrary, their literal truth value will probably be of too little relevance to be
worth determining. But this does not compromise the narrative coherence of fictions,
because successful reference resolution is not necessary for coreference to occur (think
of algebra: we do not need to know the value of x to know that, in x2 ¼ 2xy, each x
refers to the same value, which is also the value of 2y). The communicative efficacy of
multiple references to fictional characters, places, and events is a pragmatic matter,
not a semantic one. As a fictional narrative progresses, further assumptions become
manifest not because earlier assumptions have projected a fictional world within
which the fictional truth of new assumptions can be established, but because the
achieved relevance of the earlier assumptions itself becomes a contextual basis for
maximizing the relevance of subsequent related assumptions.
A relevance-driven pragmatic account of inference in fiction does not need to
proceed by way of a referential world beyond the discourse, or a denotative ‘‘de re’’
semantics beyond the attributive ‘‘de dicto’’ relations between referring expressions.
Everything we can explain by conceiving of fictions as referential constructs projecting fictional worlds, we can explain as well, without cumbersome detour, restrictive
norms, paradox, or redundancy, by understanding fiction as the serious use of a
language’s representational capacity for fictive (imaginary, not literally assertive)
purposes: by thinking in terms of the pragmatics rather than the semantics of
fictionality. If we do so, the communicative criterion of relevance is primary rather
than deferred or indirect, and unitary rather than internal or external to a fictional
frame.
A more developed example from Kafka’s novel might help to clarify the view of
fictionality I am proposing. In what follows, however, I am not advancing a critical
methodology, only illustrating a theoretical model. Relevance theory can help explain
the principles underlying the experience of fictive communication, but it doesn’t lend
Richard Walsh
160
itself to the eloquent articulation of sophisticated instances of such experience, and
still less to the production of striking new interpretations. In my Penguin Classics
edition, the first sentence of The Trial is translated as follows: ‘‘Someone must have
made a false accusation against Josef K., for he was arrested one morning without
having done anything wrong’’ (Kafka 1994: 1). In elaborating the inferential processes
invited by this sentence, I shall deal only with possible assumptions relating to K.,
who is clearly its focus, if not its subject. Without any pretension to analytic precision
or completeness, then, these four possible assumptions immediately present themselves:
1
2
3
4
Josef K. existed.
Josef K. was arrested.
Josef K. had not done anything wrong.
Someone had made a false accusation against him.
The contextual assumption of fictionality also informs the processing of this sentence
(because we found the book in the fiction section of the bookstore, or we are reading it
for a course on the modern novel, or we have a prior general knowledge of Kafka).
How does that affect our processing of these assumptions?
First, it diminishes the relevance of assumption 1. This is an existential assumption, contingent upon the possibility of resolving the reference of ‘‘Josef K.’’ in the
real world. The assumption of fictionality doesn’t rule this out, but it does create a
presumption that it is of negligible relevance, and therefore, within the economy of
effort and effect that drives the process of comprehension, it is not worth processing.
Note that this is a quite different matter from that of K.’s ‘‘existence’’ in his world,
which is either a fictional worlds concept, or (in its more general looser usage) a form
of participatory collaboration between critical discourse and fictive discourse. An
existential assumption adds nothing to the latter perspective, because (following
Kant) existence is not a predicate: it is not in itself a quality of any concept (here,
the concept ‘‘K.’’). There are other kinds of characters, of course, for whom the
existential assumption would indeed be relevant: it would be an impoverished reading
of War and Peace that failed to recognize any reference to a historical figure in the
character of Napoleon. Assumptions 2 to 4 fare differently, because they provide
information about K. presupposing rather than asserting the proposition in assumption 1; their coherence is provided for by coreference, not reference resolution, and
so they are not in direct conflict with the assumption of fictionality. Clearly, though,
they cannot achieve relevance merely as information about K., but they can help
to flesh out several possible narrative schemata, each of which is a potential explanatory framework for information of this kind. Processing these assumptions in accordance with our assessments of their relative strength, and in relation to such
schemata, involves the testing and development of our narrative understanding: in
those terms alone it offers a degree of relevance that we may well find worth the effort
involved.
The Pragmatics of Narrative Fiction
161
This prospect is enhanced by another effect of the assumption of fictionality, which
is to license imaginary extensions of the scope of knowledge – specifically, here, in the
form of internal focalization. The representation of another’s mental perspective is not
in itself a categorical indicator of fictionality, but it is certainly a possibility that the
assumption of fictionality makes much more readily available to interpretation. So
here, ‘‘somebody must have . . . ’’ is, in a fiction, unnecessarily conjectural unless it
reflects K.’s perspective (or it is the voice of a represented narrator – but no other
evidence emerges to support that inference). This is enough to make manifest the
further assumptions
5 Josef K. thought that someone had made a false accusation against him.
And, on the same basis,
6 Josef K. did not think he had done anything wrong.
The interpretative basis for assumption 6 is more explicit in the original German than
in this translation, but it is available here nonetheless (and there is further confirmation of internal focalization in the following sentence, in which the cook who
normally brought K. his breakfast did not come ‘‘this time’’ – the deictic focus
conveying K.’s experiential perspective rather than that of the narration).
These assumptions qualify the manifestness of assumptions 3 and 4, and introduce
a fundamental ambiguity. Internal focalization belongs within the class of utterances
that Sperber and Wilson term ‘‘echoic’’: utterances that interpret another person’s
thought or speech (this class also includes direct speech and first person narration).
They achieve relevance not only by providing information about what that thought
was, as in assumptions 5 and 6, but also by taking an attitude towards it (Sperber and
Wilson 1995: 238). But what attitude? Internal focalization embraces many shades of
irony and sympathy, and under different interpretations, this particular echoic utterance may allow either of the following assumptions to be inferred:
7 Josef K. was the victim of an injustice.
8 Josef K. was ignorant of the law to which he was subject.
And of course further assumptions become manifest in the light of the reader’s
evaluation of these two: that K. had been framed, or was a paranoid victim (7); that
he had been justly reported, or was a paranoid and ignorant offender (8). The
evaluative nature of these inferences involves an affective investment of some degree,
ranging from judgmental detachment to sympathetic involvement, which will be
informed by the reader’s emotional and ideological predispositions toward the relation
between the individual, self or subject and the law, with all its connotations. Given
that the ambiguity of the case inhibits a decisive preference for any one subset of the
available competing assumptions, the affective investment (in any nonreductive
162
Richard Walsh
interpretation) will be complex. This, of course, is fundamental to the effect of the
narrative to come, not least because K.’s own attitude toward his predicament is
complex: in a wonderful passage a few pages later, Kafka has him ‘‘trying to see it
from his own point of view’’ (1994: 7). There is a nice equilibrium between evaluative
detachment and imaginative involvement here: the reader’s uncertainty in relation to
the evaluative import of the internal focalization has an effect of detachment, yet its
congruence with K.’s own anxiety about his standing before the law also invests it
with the quality of affective involvement.
As the inaugural sentence of a fiction, this one achieves its effects within a
relatively simple context. Nonetheless the inferences available here already (necessarily) tend to extrapolate narratively from the utterance itself; and the subsequent
narrative development will carry forward the investment of interpretative effort
already made, along with the effects that secured a sense of relevance from the process,
so that the context for subsequent utterances will furnish many more possible
inferences of all kinds. The inferences actually drawn will vary from reading to
reading, according to cognitive environment and interpretative agenda, because
these contextual factors will qualify the specific expectations of relevance in each
case. But in all cases, the satisfaction of those expectations will require some prioritizing of lines of inference: the pragmatic nature of the process of comprehension
dictates that it is hardly ever exhaustively logical. The goal of relevance does not
require it, which is as well, not because it allows for the assumption of fictionality
(that, indeed, secures the logic of fictive utterance), but because it renders inconsequential the many possible inferences from most fictional narratives that would throw
their representational logic into disarray.
In contrast with extant accounts, a pragmatic theory of fictionality does not require
any detachment of fictive discourse from its real-world context. There is no need for a
principle of minimal departure to supply a background for the narrative particulars,
because this role is filled by contextual assumptions. These are not part of a fictional
world, but of the communicative situation. In this respect as in others, the point is
that fictionality is best understood as a communicative resource, rather than as an
ontological category. Fictionality is neither a boundary between worlds, nor a frame
dissociating the author from the discourse, but a contextual assumption by the reader,
prompted by the manifest information that the authorial discourse is offered as
fiction. This contextual assumption is a preliminary move in the reader’s effort to
maximize relevance. It amounts to a rhetorical orientation: an expectation that the
relevance of the discourse will be most profitably pursued, not by deriving strongly
informative implicatures that depend upon successful reference resolution, but by
deriving a large array of weaker implicatures which cumulatively produce affective
and evaluative effects, and which are not vitiated by any degree of literal reference
failure, but do indeed, ultimately, constitute a cognitive benefit, and an improvement
in knowledge. Nothing in this model excludes the possibility of gaining factual
information from fiction: fictionality does not admit of degree as a rhetorical set, but
fictions do as representations. This distinction, between mutually exclusive commu-
The Pragmatics of Narrative Fiction
163
nicative intentions (the fictive and the assertive) and the relativity of informative
intentions, can accommodate the range of borderline cases that vex definitions of
fiction: historical novel, roman à clef, fictionalized memoir, historiographic metafiction, hoax. The knowledge offered by fiction, however, is not primarily specific
knowledge of what is (or was), but of how human affairs work, or, more strictly, of
how to make sense of them – logically, evaluatively, emotionally. It is knowledge
of the ways in which such matters may be brought within the compass of the
imagination, and in that sense understood. A pragmatic theory of fictionality does
not confine the value of fiction to an improvement in knowledge, even in the broadest
senses I have suggested; but it claims that fictions do offer directly communicated
cognitive benefits, foregrounded by the contextual assumption of fictionality itself.
There is room for dispute about the scope of the properly cognitive in our
experience of fictions, but I am not seeking to characterize that experience, or its
value, in wholly cognitive terms. Nor do I think that the cognitive view of communication here implies a restrictive view of authorial intention in fiction.1 My argument is specifically directed against the entrenched idea that fictive discourse entails a
formal, intentional, or ontological frame. All current approaches to fictionality invoke
some such frame, and literary criticism negotiates with the fact by a kind of
equivocation, doublethink, or fudge: that is, even while its raison d’être is arguably
to bridge the gulf between fiction and reality, it actually tends to oscillate between
views from either side. It collaborates with, participates in, the fiction; or else it
detaches itself, in the process often opening up a gap between the enlightened critic
and the naı̈ve, deluded reader. The first, inside, view dominates in most close reading
and representationally oriented criticism (by which I mean criticism focused upon the
narrative particulars); the second, outside, view is often apparent in formalist and
reader-response critical orientations (at least insofar as they project stories of reading),
and symptomatic modes of criticism that bring to bear (for example) Marxist,
feminist, psychoanalytical, queer, or postcolonial perspectives upon the text. While
I do not think there is anything fundamentally wrong with either approach, neither
actually explains what is going on in fictive discourse: the first takes it for granted
that we are already familiar with fiction (which of course we are); the second tends to
bracket fictionality in pursuit of other interests for which the fictional text provides
occasion. By refusing this inside/outside dualism, a pragmatic approach to fictionality
identifies the issue it effaces: it does not, and should not, conflict with what we
currently do as readers and critics, but it identifies something we are not doing that
I suggest would be worthwhile. It challenges us to explain the force and effect of
fictionality itself in our experience and understanding of fiction.
NOTE
1 For a perspective upon authorial creativity that
lends itself to redescription in relevance theory
terms, see Walsh (2000).
164
Richard Walsh
REFERENCES
AND
Abbott, H. P. (2002). The Cambridge Introduction to
Narrative. Cambridge, UK: Cambridge University Press.
Genette, G. (1990). ‘‘The Pragmatic Status of
Narrative Fiction.’’ Style 24, 59–72.
Grice, H. P. (1989). Studies in the Way of Words.
Cambridge, MA: Harvard University Press.
Kafka, F. ([1925] 1994). The Trial, trans. I. Parry.
Harmondsworth, UK: Penguin.
Kearns, M. (1999). Rhetorical Narratology. Lincoln:
University of Nebraska Press.
Kearns, M. (2001). ‘‘Relevance, Rhetoric, Narrative.’’ Rhetoric Society Quarterly 31, 73–92.
Margolin, U. (1991). ‘‘Reference, Coreference,
Referring, and the Dual Structure of Literary
Narrative.’’ Poetics Today 12, 517–42.
Ohmann, R. (1971). ‘‘Speech Acts and the Definition of Literature.’’ Philosophy and Rhetoric 4,
1–19.
FURTHER READING
Pavel, T. G. (1986). Fictional Worlds. Cambridge,
MA: Harvard University Press.
Pratt, M. L. (1977). Toward a Speech Act Theory
of Literary Discourse. Bloomington: Indiana
University Press.
Ryan, M.-L. (1991). Possible Worlds, Artificial
Intelligence and Narrative Theory. Bloomington:
Indiana University Press.
Searle, J. R. (1975). ‘‘The Logical Status of Fictional
Discourse.’’ New Literary History 6, 319–32.
Sperber, D. and Wilson, D. (1995). Relevance:
Communication and Cognition, 2nd edn. Oxford:
Blackwell.
Walsh, R. (1997). ‘‘Who is the Narrator?’’ Poetics
Today 18, 495–513.
Walsh, R. (2000). ‘‘The Novelist as Medium.’’
Neophilologus 84, 329–45.
Wilson, D. and Sperber, D. (2002). ‘‘Truthfulness
and Relevance.’’ Mind 111, 583–632.
PART II
Revisions and Innovations
10
Beyond the Poetics of Plot:
Alternative Forms of Narrative
Progression and the Multiple
Trajectories of Ulysses
Brian Richardson
While a detailed examination of discussions of plot in the twentieth century would
reveal some significant disagreement (e.g., the structuralist emphasis on a grammatical order of events versus the neo-Aristotelian stress on the affective consequences of a
trajectory of action), stepping back from specific divergences reveals substantial areas
of general agreement, even among theorists who otherwise have little in common. The
analysis offered by a proto-structuralist like Vladimir Propp is entirely compatible
with the formulation of Paul Ricoeur, writing from the rival perspective of hermeneutics: for Ricoeur, plot is ‘‘the intelligible whole that governs a succession of events
in a story. . . . A story is made out of events to the extent that plot makes events into a
story’’ (1981: 167); Propp similarly postulates a connected series of events that leads
to a resolution of the problem or conflict with which the story began. These stances
are not only congruent with Peter Brooks’s post-structuralist psychoanalytic
approach, both are actually cited by him in support of his own position (1984:
13–17). Brooks’s use of plot as a term to embrace ‘‘the design and intention of
narrative, a structure for those meanings that develop through succession and time’’
(1984: 12) is in turn consonant with the other major strand of theorizing plot: the
emphasis on unity, design, completion, and effect produced by the neo-Aristotelians
associated with the University of Chicago, beginning with R. S. Crane. Distilling
these different discussions leads to this conception of plot: an essential element of
narrative, plot is a teleological sequence of events linked by some principle of
causation; that is, the events are bound together in a trajectory that typically leads
to some form of resolution or convergence.1
The problem with such a notion, however, is that many narratives resist, elude, or
reject this model of plot and its explicit assumption of narrative unity, cohesion, and
teleology. This is especially true of twentieth-century texts that remain insistently
fragmentary, open-ended, contradictory, or defiantly ‘‘plotless.’’ The definition just set
168
Brian Richardson
forth is of little use in describing the trajectories, or what I will call narrative
progressions, traced by Ulysses or The Waves, to say nothing of the still more
experimental work of Gertrude Stein, Raymond Roussel, Samuel Beckett, or Alain
Robbe-Grillet.2 (I might add that this assertion is not altered by the fact that we need
to draw on this definition of plot in order to comprehend the specific violations made
by such authors.) In what follows, I will identify the most salient varieties of nonplotbased narrative ordering in recent fiction, point out relevant historical antecedents,
and note how these strategies supplement or supersede more traditional modes of
sequencing. This inventory will have its own sequencing principle, moving from the
most familiar to the most outrageous orderings, that is, from those that almost
invisibly accompany the movement of story to those that most spectacularly overthrow it. It is striking how many different nonplot-based methods of producing
events there are, and surprising that these have not been more thoroughly examined
and discussed by narratologists. I will also look for the presence of each strategy in
Ulysses, a novel that employs a rather large number of ordering techniques other than
those of conventional emplotment.
Ulysses is particularly interesting in this regard. On the one hand, the work seems
to lack what might seem to be the minimal necessary plot: event after event appears to
occur more or less randomly; the first linear progression (Chapters 1–3) is interrupted
and the clock is reset as we meet Bloom in Chapter 4; the two main storylines
often approach each other but never fully meet in any significant way; and the final
chapters resist any mechanism that will tie the events together as Stephen and Bloom
part ways like two ships passing in the night, as a common description of the
book’s inconclusive ending avers. Many shorter sequences seem adventitiously conjoined and are certainly unconnected by any large causal chain; ‘‘The Wandering
Rocks’’ episode, which traces the essentially noninteractive movements of several
spatially adjacent Dublin citizens, can even be seen as a kind of quintessence of the
work’s refusal of traditional plot.3 The book’s frequent use of interior monologue and
free indirect speech makes the sequencing of still shorter passages seem even more
adventitious; to say that sentence B follows sentence A because that thought just
popped into the character’s mind is not much of an explanation at all. Todorov goes
so far as to state that ‘‘the most striking submission to the temporal order is Ulysses.
The only, or at least the main, relation among the actions is their pure succession’’
(1981: 42).
One obvious method by which these ostensibly gratuitous events and episodes are
patterned is by their reproduction of the order of an earlier text.4 The dual linear
progressions just noted appear in Ulysses because the same sequence occurs in the
Odyssey: Joyce puts Stephen’s otherwise unmotivated and inconsequential encounter
with the Protean ocean (Chapter 3) after his meeting with the Nestor figure (Chapter 2)
because Homer’s Telemachus speaks with Nestor (Book Three) before doing battle
with Proteus (Book Four). Homer’s largely causal sequence becomes the template for
Joyce’s otherwise random conjunctions. This ordering continues for much but not all
of Joyce’s text: his Lotus Eaters, Aeolus, Lestrygonians, Sirens, and Oxen of the Sun
Beyond the Poetics of Plot
169
episodes follow the same order as that used by Homer; other episodes, however
(Cyclops, Circe), are rearranged to suit different purposes.
Sheldon Sacks has provided another useful way in which we may think about
narratives organized in a manner largely independent of the customary qualities of
plot. Discussing the genre of the apologue, or fictional exemplification of a thesis or
worldview such as Candide or Rasselas, he points out that the episodes are ‘‘related to
each other in a rhetorical order,’’ rather than a probabilistic one. ‘‘There is no fictional
‘probability’ that Rasselas, after he leaves the haunts of gay young men, will meet a
sage committed to controlling the passions’’ (Sacks 1964: 56); such a sequence,
however, does follow from the demands of the novella’s argument. This kind of
rhetorical sequencing is found in the more ideologically charged turns of many novels.
Joyce, being resolutely antididactic, does not normally use rhetorical progressions in
this way. Nevertheless, we can see numerous miniature and oblique rhetorical sequences in the dialectical progression of the events in ‘‘Scylla and Charybdis,’’ in
which the idealistic theses propounded by the figures in the library are followed by
the crassly materialistic positions of Buck Mulligan, whose entry into the room is
synchronized with the model of thesis/antithesis. In a related manner, the catechistic
structure of ‘‘Ithaca’’ is minimally narrative and is sequenced instead by the linked
series of questions and answers; as C. H. Peake observes, this ‘‘is not naturally a
narrative method; it implies a static situation which is being examined and analyzed
rather than the unrolling of a concatenated series of events’’ (1977: 283).5 Aspects of
the progressions common to apologues also appear in the numerous rhetorical
trajectories present in the ‘‘Aeolus’’ chapter that is set in a newsroom and thematizes
the art of rhetoric.
It is easy to move from a sequence of events that exemplifies an argument to more
general motif-based alternation or progression. But while these modes of composition
are similar, their motives may be opposed. Rhetorical sequencing is an intentional
arrangement set forth as advantageously as possible to produce a particular effect: to
bring the mind of the reader into closer conformity with the beliefs of the author. In
this sense, it is every bit as ‘‘functional’’ as traditional emplotment, whose purpose is
to impel the reader from chapter to chapter observing how sympathetic protagonists
attempt to overcome adversity and attain their desires. Motif-based, architectonic,
numerological, or geometrical kinds of sequencing are primarily formal designs that
have little function other than that of satisfying a desire for symmetry. I will refer to
these forms as ‘‘aesthetic’’ orderings. We may start by looking at the growth of the
‘‘little phrase’’ of music by Vinteuil in Proust’s In Search of Lost Time. The phrase
appears in a number of scenes in different incarnations, but, following a general
trajectory, ever more elaborate instantiations. It begins as a musical phrase heard by
Swann, and then is identified as part of a sonata by a man named Vinteuil. As the
work continues, the composer’s identity is fully disclosed, and the theme, once a token
of Swann’s love for Odette, becomes an emblem of the disintegration of Marcel’s love
for Albertine. Finally, the discovery of a full-length septet by Vinteuil is made.6 The
development of this motif has its own independent trajectory; in response to the
170
Brian Richardson
question, ‘‘Why does the discovery of the Vinteuil septet occur late in the work?’’ it is
as plausible to say, ‘‘Because it is the culmination of the expanding theme of the
musical phrase’’ as to aver that it discloses again the achievement of the neglected
composer at a strategic point of the narrative. After a certain point, the theme does
not merely accompany the narrative; instead, the narrative events are produced to
accommodate the development of the motif.
Numerous other such aesthetic progressions may be enumerated, including the
familiar circle pattern that returns important aspects of the narrative to their starting
points; another is that which E. M. Forster described as the hourglass shape of Henry
James’s The Ambassadors (Forster 1927: 153–62). We may also point to Tolstoy’s
insistent alternation of light and dark scenes throughout Anna Karenina, and further
note James’s perhaps unfair castigation of Tolstoy for producing ‘‘fluid puddings’’ and
‘‘large loose baggy monsters’’ from the perspective of formal design ( James 1972: 267,
262). When examining any of the many scenes that do not obviously impel the plot
forward, one may explain its placement in terms of its motif function rather than as
part of any causal chain of events. Joyce is quite adept at these kinds of progressions:
each of the final 15 chapters of Ulysses thematizes a different organ of the human body;
similarly, a different art or science is foregrounded in each. The specifically generative
functions of these themes occur more at the level of individual events than of the
chapter – Joyce did not add the Aeolus episode because he wanted to dramatize a
specific organ (lungs) and discipline (rhetoric) – though several smaller events
(including mental events) take place for the primary purpose of illustrating these
themes (‘‘The door of Ruttledge’s office whispered: ee: cree. They always build one
door opposite another for the wind to. Way in. Way out’’ [Joyce 1986: 97].)
Still other kinds of symmetrical arrangements of chapters and events can be
adduced; these do not merely provide a structure for an otherwise unorganized
conglomeration of events but at times go on to actually produce some of those events.
Viktor Shklovsky (1990) identified a number of formal arrangements of narrative
materials, including repetition, parallelism, antithesis, and triadic patterns, and
pointed out that much of the Chanson de Roland is composed around dual and triple
repetitions of the same set of scenes and events. In fact, many of these actions are
present only because they complete the formal pattern that animates the rest of the
text, in contravention of other compositional principles like causal connection,
verisimilitude, or rhetorical efficacy. William W. Ryding carries this kind of analysis
much further, describing how a number of medieval narratives eschew narrative unity
in favor of ‘‘artistic duality, trinity, or some other form of multiplicity’’ (1971: 116),
including the multiplication of parallel or antithetical story lines exclusively for this
effect. Thus the second part of Beowulf, which takes place 50 years after his victory
over Grendel and the hag, is an entirely new (though fully symmetrical) story of the
aged Beowulf’s battle with the dragon. This work, like the Chanson de Roland, ‘‘has in
fact two beginnings, two middles, and two ends. The central discontinuity that seems
so clumsy to us appears to have served the medieval writer as a means to a particular
esthetic end – it was, we may suppose, a special grace in story-telling’’ (Ryding 1971:
Beyond the Poetics of Plot
171
43).7 Other comparable methods of production and ordering of narrative segments are
common in numerous periods of literary history.8 These various architectonic progressions are no doubt understudied because in many cases they may seem to be less
important than, or a mere appendage to, the unfolding of the progression of the
story’s main events; so powerful is the pull of the plot in the perception of narrative
that plot may need to be abandoned or suppressed for alternative ordering systems to
become visible. Nevertheless, these methods of sequencing do help explain why a
given narrative has the events and arrangement it does. Even if Dante’s story were
largely completed halfway through the Paradiso, he would have had to stretch his
material out until he had reached the structurally requisite 33 cantos.
As the example from Proust suggests, the arrangement of a cluster of literary motifs
may be modeled on or borrowed from standard musical progressions. For his entire
novel Proust employed the structure of a Wagnerian opera; others have utilized the
general structure of the sonata (Strindberg) or the symphony (Gide, Andrei Biely), the
framework of jazz (Ralph Ellison, Toni Morrison), or the prescriptions of the classical
Indian musical form, the raga (Amit Chaudhuri’s 1993 Afternoon Raag). A trajectory
provided by the fugue has at times proven irresistible, as evidenced by Thomas
Mann’s ‘‘Death in Venice’’ and the fuga per canonem that orders the ‘‘Sirens’’ episode
of Ulysses.9 These last two examples point to an important distinction in nonplotbased ordering devices: often, these are unobtrusive, working in tandem with more
conventional modes of story sequencing, producing a trajectory of otherwise largely
unmotivated sequences or (as in the case of Mann) an overdetermined narrative
progression (that is, Aschenbach dies in Venice both because he has chosen to stay in
the city as the plague spreads and because it is the final expression of the bass theme,
death, as it merges with its contrapuntal theme of sexual desire). Many of the
sequences in Joyce’s ‘‘Sirens,’’ however, make little or no sense if approached from
the traditional perspectives of story or plot; indeed, the episode’s first set of words
make virtually no sense from almost any conventional framework: ‘‘Bronze by gold
heard the hoofirons, steelyringing./ Imperthnthn thnthnthn./ Chips, picking chips off
rocky thumbnail, chips./ Horrid! And gold flushed more./ A husky fifenote blew./
Blew. Blue bloom is on the./ Goldpinnacled hair. A jumping rose on satiny breast of
satin, rose of Castille./ Trilling, trilling: Idolores./ Peep! Who’s in the . . . peepofgold?’’ (p. 210, Joyce’s ellipsis). As an overture that previews the primary versions of
the major themes and motifs to follow (and roughly approximating the order of their
appearance), it is an accurate and useful compendium that performs the same function
as a musical overture – especially a very daring one which offers such disparate motifs
that we are unsure whether they can in fact come together in the development of the
music that follows.
One might designate the presentation of the opening phrases of ‘‘Sirens’’ as
foreshadowing or announcing the material to follow, but one may equally effectively
view them as generating the rest of the chapter. This type of oscillating perspective is
frequently relevant to narratives that are or seem to be generated by (fictional)
pictures within the text. Longus’s second-century novella, Daphnis and Chloe, begins
172
Brian Richardson
with the partial description of a narrative painting, which the narrator finds so
wonderful that he decides to narrate in prose the story it depicts. In Goethe’s
‘‘Novelle,’’ we are presented with drawings of an abandoned castle, the story of a
marketplace fire, and the picture of a tiger leaping on a person. As the tale unfolds,
the protagonist visits the castle, observes a fire break out in the marketplace, and
witnesses a real tiger leaping on a person. Such unlikely repetitions may be properly
viewed as uncanny coincidences or an overactive display of irony; however, one might
also read them as a cunning, proto-Borgesian play with narrative sequencing in which
simulacra come to engender the objects and events that they had represented.
For a Joycean example we may turn to the phantasmagoric ‘‘Circe’’ chapter, in
which images Bloom has seen earlier in the day now come alive, such as the
Greek nymph in the painting in his bedroom (pp. 444–51). This kind of ‘‘pictorial
genesis’’ is also found in many nouveaux romans, perhaps most memorably in RobbeGrillet’s In the Labyrinth, in which a number of shapes described at the beginning
of the book go on to generate objects having a similar shape which then become foci of
unfolding narratives. Thus the layer of dust in the room engenders the snow
that appears in the story that grows from it, the image of the cross-shaped object
on the desk is transformed into the bayonet of the soldier in the inner story, and the
painting, ‘‘The Defeat at Reichenfels,’’ after being described in impossible detail,
comes alive and turns into a narrative, as a description becomes through metalepsis a
sequence of events, and the opposition between temporal and spatial art forms
dissolves.10
Other forms of text generation are common in the nouveau roman and its various
antecedents; a particularly seminal type is the way in which a few select words go on
to generate the object or actions they depict. This is an important and fascinating
method of engendering a narrative and deserves to be far better known. Jean Ricardou
(1972) has called it a structural metaphor, and describes it as a trope that is made
literal and takes on life in the text. I will refer to it as a ‘‘verbal generator’’ and use it to
refer to a practice that names an object or event which then appears or occurs in the
narrative. In a traditional work, there may well be an ironic foreshadowing by an
event before it occurs; in the nouveau roman, this becomes an alternative principle of
narrative progression as words or images produce the events of the text. Thus, in
Robbe-Grillet’s Project for a Revolution in New York, the concept ‘‘red’’ in all its
permutations and variations generates many of the events (including murder and
arson); still more primary is the juxtaposition of contraries, as Thomas D. O’Donnell
(1975) has explained. Noting further narrative proliferation, O’Donnell traces the
avatars that produce the rat in the book. ‘‘Very early in the novel, the narrator informs
us that Ben Said is wearing black gloves; when writing in his notebook, Ben Said
tucks the gloves under his armpit. Another glove appears on the cover of Laura’s
detective story,’’ this in turn suggests that
Ben Said may be responsible for the fate of the girl on the story’s torn cover. Upon closer
examination, it is noted that the ‘‘glove’’ is in reality an enormous furry spider. Laura
Beyond the Poetics of Plot
173
found the book on top of the bookcase while trying to escape from a giant spider or a
rat; henceforth, spider and rat form an elementary combination that may not be
dissociated.
(O’Donnell 1975: 192)
These examples, O’Donnell points out, ‘‘illustrate Robbe-Grillet’s thematic generative technique to provide a long range ‘plot’ for his novel’’ (1975: 192). In Joyce’s
‘‘Circe,’’ we find a clear example of a verbal generator producing a substantial stretch
of text. As Bloom denigrates tobacco, Zoe retorts: ‘‘Go on. Make a stump speech out
of it.’’ What immediately follows in the narrative is the figure of Bloom in workingman’s overalls, giving an oration on the evils of tobacco before an adoring populace
(1986: 390–3), as the phrase ‘‘make a stump speech’’ precedes the event it simultaneously names and produces.
In many of the compositions of Jean Ricardou, on the other hand, individual
French words produce slight lexical variants which go on to generate the newly named
objects or relations in the text as it unfolds. Even the name of the press that appears on
the title page (‘‘Les Éditions de minuit’’) can serve as a textual generator: thus, in La
Prise de Constantinople, the word Éditions engenders the characters Ed and Edith, as
well as the idea of the hill of Sion, while Minuit determines the book will open at
night (Ricardou 1972: 384).11 Though it may appear the exclusive demesne of the
avant-garde, such ‘‘lexical generators’’ can actually be traced back as far as Sterne’s
Tristram Shandy if not in fact to Dante.12
Somewhat less dynamic is the comparable ordering principle of alphabetical
patterns. Such a progression, as Roland Barthes once remarked at the beginning of
one of his own such compositions, has all the order and arbitrariness of the alphabet
itself. This kind of progression, which probably traces its descent from Raymond
Roussel, is not inherently a fictive one, and can be readily found in nonfiction and
nonnarrative forms (e.g., Barthes’s A Lover’s Discourse) as well as works that straddle
the line between narrative and nonnarrative, such as Michel Butor’s Mobile, which
Sherzer describes as ‘‘semiotic catalogue’’ and states (certainly debatably), that it ‘‘is
without narrative’’ (1986: 46, 50). An interesting elaboration of this stratagem can be
found in Milorad Pavić’s Dictionary of the Khazars, a novel in the shape of three
dictionaries. A narratologist might object that alphabetical composition is merely a
different way to rearrange the sjuzhet and does not really affect the fabula; however, in
some works, like Walter Abish’s Alphabetical Africa, the ordering principle is clearly
generating that which is depicted (see Orr 1991: 113–16). Other transpositions of
familiar verbal ordering principles onto narrative fiction (critical commentary in
Nabokov’s Pale Fire, the crossword puzzle in Pavić’s Landscape Painted with Tea) may
be included here. Though Joyce delights in the play of individual letters and is
intrigued by the alphabet (‘‘Ahbeesee defeegee kelomen opeecue rustyouvee’’ [p. 48],
I don’t find any strictly alphabetical orderings in Ulysses of any scale (though the
letters of the last word of the book –‘‘yes’’ – are contained, in reverse order, in its first
word, ‘‘Stately’’).13
174
Brian Richardson
Another generating mechanism used by many recent French authors is based on
repetition of events rather than on a progression from one event to another. The classic
instance of this practice is probably Robbe-Grillet’s Jealousy, in which approximately
the same set of events is presented nine consecutive times, with each version containing significant variations. As Robbe-Grillet has explained, in a traditional narrative,
‘‘what follows phenomenon A is a phenomenon B, the consequence of the first,’’ while
in a nouveau roman like Jealousy, ‘‘what happens is entirely different. Instead of having
to deal with a series of scenes which are connected by causal links, one has the
impression that the same scene is constantly repeating itself, but with variations;
that is, scene A is not followed by scene B but by scene A’, a possible variation of scene
A’’ (Robbe-Grillet 1977: 5). This technique is variously designated by its theorists;
the most useful term is probably that employed by Dina Sherzer, who calls this kind
of progression ‘‘serial constructs’’ (1986: 13–36).14 Serial constructs may be further
sequenced according to other patterns of progression: the repeated, contradictory
depictions that constitute Robbe-Grillet’s ‘‘The Secret Room’’ are shaped in the
form of a temporal spiral; the obsessively reenacted scene in Robert Coover’s Spanking
the Maid gradually rises to a peak of physicality before rapidly subsiding at the end of
the book. That this technique continues to thrive in other genres is evidenced by the
success of the recent German film Run, Lola, Run. In each episode of Ulysses, Joyce
includes echoes of other episodes, some of which can feel quite out of place, and
seemingly present only for their echoic function, but other than this cannot be said to
generate the text. Miniature reproductions of the string of events of the entire book
(that is, mises en abyme) are also present at several points in Joyce’s text (e.g., pp. 543,
552).
A related technique is ‘‘collage’’ composition, in which several key elements are
recombined in a number of different arrangements and contexts and which constitute
the nexus that connects the different units.15 This order may be present (and is no
doubt less jarring) in nonnarrative texts; it is also more of a principle of coherence
rather than progression per se, since after a certain amplitude is reached, there is no
inherent reason for the text to continue. Nevertheless, to answer the question ‘‘How
are the third or fourth sections related to the opening units?’’ a plausible response is
that they are recombinations, analogues, or variations of some of the elements present
in the earlier segments; the collage technique, that is, necessitates such a progression.
As Dina Sherzer remarks, such texts ‘‘are open in that no one referential or morphological element brings about the sense of an ending or a feeling of completion; other
variations and repetitions could be added to the existing ones, lengthening the text
but not changing it otherwise’’ (1986: 14). This observation is a fairly good depiction
of Lyn Hejinian’s text, My Life, a partially autobiographical collage that was originally
published in 1978, when the author was 37. At this time, the work consisted of 37
sections, each with 37 sentences. The second edition, published eight years later, has
eight new sections of 45 sentences, and eight new sentences were added to all of the
previously published sections. Here, Ulysses can function as a model for such practices,
as its central figures, motifs, tropes, and elements are recombined in successive
Beyond the Poetics of Plot
175
chapters, sometimes in ways that violate the book’s mimetic stance, as Hazard Adams
(1986) has pointed out in his study of these deviously ‘‘wandering rocks.’’
Before leaving this arena we need to engage with two other kinds of transgressive
orderings: those that may seem to have too much plot and those that have too little.
The ‘‘forking paths’’ principle articulated by (though not really embodied within)
Borges’s story ‘‘The Garden of Forking Paths’’ can lead us directly to a genuinely new
kind of multiple, mutually exclusive, orderings of a text. A good instance is found in
Ana Castillo’s The Mixquiahuala Letters, a text that offers three different sequences for
reading the book: one for conformists, one for cynics, and one for the quixotic reader.
None of these recommended sequences includes the entire group of the letters, or even
begins with letter number one, with which the book opens; this arrangement
necessarily suggests yet another possible progression, the numerical one. In this
work, like Julio Cortázar’s Hopscotch, whose techniques it extends, the suggested
sjuzhet is largely linear; the different possible trajectories do not result in radically
different fabulas, though the interpretation of the basic fabula will alter depending on
which version is followed. More radical in this regard are the popular children’s ‘‘pick
your own adventure’’ books, where different readerly choices result in quite different
sequences.
The ‘‘forking paths’’ kind of composition is not unrelated to hypertext narratives,
and indeed may be usefully thought of as a simple prototype of the latter. Such
examples are now fairly common on the web, where afternoon, a story is achieving a
certain eminence. Intriguingly, Margaret Atwood has written a kind of quasi- (or,
indeed, anti-) hypertext in her story, ‘‘Happy Endings,’’ in which some 10 available
narrative options all lead sooner or later to the same story of the death of the
protagonists. Atwood’s piece may provide a useful metacommentary on hypertext
generation: these may well be entirely containable within traditional concepts of plot;
the primary difference would be that it is the reader rather than the protagonist or the
author who chooses which event will happen next (or, more precisely, the reader
chooses from a set of limited options made available by the author). Ulysses has
regularly been described as a kind of hypertext due to its difficulty of being
comprehended on a first reading (one must know the whole book to understand any
part), as well as the innumerable patterns and correspondences that extend across
chapters and, indeed, across Joyce’s other works.
A final aspect of narrative unfolding needs to be mentioned, and that is the move
toward a relatively pure sequence that has few connectors and still fewer textual
generators, whether conventional or avant-garde. A number of feminist works move
in this direction, starting with Dorothy Richardson’s fiction, the first works of
literature to which the critical term ‘‘stream of consciousness’’ was applied, and
which Virginia Woolf critiqued for lacking ‘‘unity, significance, or design’’ (1980:
190). Woolf would herself experiment with pure linearity in the ‘‘Time Passes’’ section
of To the Lighthouse, a section that may have inspired Eva Figes’s novel Waking, a
chronicle of seven awakenings into consciousness by the central character, each of
them separated from the others by about a decade, and so loosely conjoined that there
176
Brian Richardson
is fairly little to connect them all to the same individual. Such an ‘‘excessive’’ linearity
also appears in other works that fall under the rubric of écriture féminine, such as Clarice
Lispector’s Agua Viva (The Stream of Life) or Molly Bloom’s soliloquy.16 An even
greater challenge to the notion of a single narrative line is the kind of narrative with a
collective subject dispersed over three continents and 250 years in Caryl Philips’s
novel/novellas of the African diaspora in Crossing the River. These radical examples
show how much traditional plot is actually present in episodic or picaresque novels;
after all, Lazarillo de Tormes ‘‘is a character who is modified and molded by his
adventures and his ambience. The innocent child who has his head smashed against
the bull is quite different from the vengeful child who makes his blind master smash
his head against the pillar’’ (Fiore 1984: 84).
This analysis leads now to our last category, or rather anticategory, of narrative
progression: the aleatory. Popularized by Dadaists who would select phrases that had
been cut out of newspapers and thrown into a hat, a number of authors and composers
including William Burroughs and Karlheinz Stockhausen have utilized this technique. Beckett’s ‘‘Lessness’’ is a short text said to be randomly assembled, though the
numerous interconnections among its elements tend to make any order they may
appear in seem purposive. There are no aleatory elements in Ulysses – Joyce once
wondered whether it wasn’t too meticulously structured – and it may be that there is
only a single aleatory phrase in all Joyce’s works. As Richard Ellmann recounts, Joyce
was dictating part of Finnegans Wake to Samuel Beckett.
In the middle of one such session there was a knock at the door which Beckett didn’t
hear. Joyce said ‘‘Come in.’’ and Beckett wrote it down. Afterwards he read back what he
had written and Joyce said, ‘‘What’s that ‘Come in’?’’ ‘‘Yes, you said that,’’ said Beckett.
Joyce thought for a moment, then said, ‘‘Let it stand.’’ He was quite willing to accept
coincidence as his collaborator.
(Ellmann 1982: 649)
What conclusions are we to draw from the preceding analyses, a cluster of examples
that at times no doubt threaten to turn into an episodic collocation of entries devoid
of all the basic elements of a good plot: unity of design, inescapable development, and
definitive conclusion? We may begin by observing the curious fact that the full range
of mechanisms for the development and progression of narrative fiction in all its
varieties has, to my knowledge, never been systematically set forth (though I hasten to
add that theorists like Monika Fludernik [1996: 269–310] and Brian McHale [1987]
have approached many of these issues from somewhat different perspectives).17
Nevertheless, I find this vast theoretical gap quite surprising and hope this essay
will help to begin to rectify this unusual situation.
Looking back over the various sequencing practices discussed above, we may
attempt some general observations on narrative progression. It is clear that a thorough
analysis of nonplot-based forms of narrative progression reveals how prevalent they are
and how significant they can be. Some of these, like rhetorical or aesthetic ordering,
Beyond the Poetics of Plot
177
can be complementary to more conventional kinds of emplotment; they add an
additional though rarely discussed motivation for the exact narrative trajectory that
emerges. Traditional emplotment often or even typically works in a kind of unacknowledged counterpoint with other methods of progression. At times these
diverge or come into collision, a situation most evident when ideological trajectories
displace probability in a realist work.18 I suspect it is where these modes clash most
visibly that we may find ideological work being done in the most undisguised
fashion, as in ideologically imposed closures that do violence to a work’s mimetic
economy. By contrast, radical aesthetic ordering techniques, such as verbal generators,
are quite disruptive in a different, more obviously deliberate, manner and regularly
defy conventions of mimesis and supplant plot altogether.
Nevertheless, I suspect that all of the more vigorous nonplot or antiplot
mechanisms of narrative sequencing are ultimately dependent on a prior notion of
emplotment, and work in a dialectical way to negate the conventional pattern that the
new arrangements nevertheless presuppose. Thus, while the concept of plot alone
cannot describe the various sequencing patterns present in many recent works of
fiction, most of those patterns can only be fully comprehended in relation to plot.
Even chance compositions are interesting not for any intrinsic reason but for the ways
they appear to mimic or contravene the kind of order produced by emplotment. The
symbiotic between plot and other antithetical orderings is especially pronounced in
the arrangement of the collection of nine contradictory versions of essentially the same
set of events in Robbe-Grillet’s Jealousy. The shifting intensity of these descriptions, as
Jean Ricardou has pointed out, nevertheless traces the traditional structure of slowly
rising and rapidly falling action typical of the conventional novel. Similarly, the
sequence of variants that compose the experimental film Run, Lola, Run follow the
general pattern of comedy, with a successful conclusion at the end of the final
sequence.
We may affirm that narrative progression is a protean, dynamic process, with
multiple sources of narrative development operating at different points in the
text, as Ulysses exemplifies so clearly. The concept of plot alone is not adequate to
explain all the sequences even of many plot-driven compositions, let alone the
more experimental textual generators identified above. We would do well to think
of plot as a component of narrative sequencing that is independent of, and working
in varying degrees of complementarity with or opposition to, other kinds of progression, especially rhetorical and aesthetic orderings. Looked at from the vantage
point of literary value, it may well be that the most compelling narrative sequences
are those that seamlessly interweave two or more strategies of progression, making
the independent orderings seem to be coextensive and unobtrusive. In any
event, it is clearly the case that the more simple and streamlined the mode
of progression, whether mere allegory, facile verbal generation, overly predictable
plot, or an unmotivated concatenation of adventures, the easier it is for experienced
readers to lose interest in the work. We may conclude by urging that the
theoretical study of plot should be subsumed within the logically more
178
Brian Richardson
capacious category of narrative progression. Such a more extensive framework is
essential if we are to have a thorough account of the way narrative fiction actually
unfolds in time.19
NOTES
1 As Peter Brooks states, ‘‘The very possibility of
meaning plotted through sequence and
through time depends on the anticipated
structuring force of the ending’’ (1984: 93,
cf. 90–112).
2 I wish to point out that James Phelan’s flexible
and very useful account of narrative progression as a dynamic event that treats its subject
as a developing whole does not suffer from
the limitations of the concepts of plot noted
above and partly for this reason I use his term
(1989: 14–22). Ralph Rader (1973) presented
another important earlier attempt to move
beyond what he terms ‘‘the realism-plotjudgment’’ model of narrative he identified
with the earlier Chicago School theorists; his
account, while disclosing three very different
forms a realistic narrative progression may
take, nevertheless remains grounded in a mimetic framework and thus cannot encompass
antimimetic patterns.
3 Timothy Martin observes that ‘‘many of the
later episodes betray principles of wholeness
independent of Ulysses as a whole’’ (1998:
208); the rest of his essay provides an impressive study of the unity and disunity of the
events and other aspects of the text.
4 Though this strategy is a favorite of modernist
and postmodern authors, its origins stretch
back to antiquity. An early, playfully selfconscious example can be found in Aristophanes’ Thesmophoriazusae. Toward the end of this
drama, Euripides’ cousin finds himself captured by those he fears will harm him. In
order to escape, he tries to imitate analogous
roles of characters similarly stranded in different plays by Euripides. After two false starts,
he enacts the scene from Andromeda that most
closely resembles his situation and is then able
to escape.
5 For an analysis of the nonnarrative nature of
‘‘Ithaca,’’ see Monika Fludernik (1986).
6
7
8
9
10
11
12
13
E. M. Forster (1927: 165–9) has a deft account of the development of this motif.
Such a concern for symmetrical presentation
may explain in part some of the contradictions in biblical narrative noted by David
Richter in this volume.
For a comprehensive overview of symmetrical
and numerological progressions in narrative
literature, see R. G. Peterson (1976). For a
study of symmetries in chapter sequencing in
the traditional novel, see Marshall Brown
(1987).
There is still some disagreement on just how
closely this episode approximates a fugue.
For a recent approach that employs Schoenberg’s 12-tone system to explicate Joyce’s
sequencing, see Herman (1994).
For other examples of this kind of text generation, see Julio Cortazar’s ‘‘Blow Up’’ and
Claude Simon’s Triptych. For the opposite
movement, in which a narrative progresses
only to end up as a painting, see Alejo Carpentier’s Explosion in a Cathedral, in which the
metaphor of spatial form becomes literalized,
as it were. Emma Kafalenos (2003) has compellingly discussed the intriguing narrative status
of ‘‘Blow Up’’ and other ekphrastic works.
For additional discussion of these important
yet insufficiently known modes of generation, see Sherzer (1986: 13–36) and Hayman (1987: 104–46).
Tristram’s father, Walter Shandy, actually
writes a book that attempts to prove that
the name one is born with strongly influences
one’s fortunes in life; or in his words, ‘‘that
magic bias which good or bad names irresistibly impress upon our characters and conducts’’ (Bk 4, Ch 8). This is certainly true
for Tristram’s sad fate. For Dante’s use of this
strategy, see the 30th canto of the Purgatorio.
There is, however, plenty of alphabetical play
in ‘‘Ithaca’’ (anagrams, acrostics, etc).
Beyond the Poetics of Plot
14 Rather less felicitously, David Hayman refers
to this as ‘‘nodality’’ (1987: 73–104).
15 Dina Sherzer discusses this type of work
under the rubric of ‘‘multidimensional montages’’ (1986: 37–76).
16 For a brief discussion of some of these texts,
see my article on linearity (Richardson
2000).
17 For additional discussion of many of the issues
raised in this essay, see my introductions to the
sections on ‘‘Plot and Emplotment,’’ ‘‘Narrative Progressions and Sequences’’ and ‘‘Narrative Temporality’’ in Narrative Dynamics
(Richardson 2002: 9–14, 64–70, 159–63).
18 An example that comes readily to mind is the
ending of D. H. Lawrence’s ‘‘The Fox,’’ in
REFERENCES
AND
Adams, H. (1986). ‘‘Critical Construction of the
Literary Text: The Example of Ulysses.’’ New
Literary History 17, 595–619.
Brooks, P. (1984). Reading for the Plot: Design and
Intention in Narrative. New York: Random.
Brown, M. (1987). ‘‘Plan vs. Plot: Chapter
Symmetries and the Mission of Form.’’ Stanford
Literature Review 4, 103–36.
Ellmann, R. (1982) James Joyce, revised edn.
Oxford: Oxford University Press.
Fiore, R. L. (1984). Lazarillo de Tormes. Boston:
Twayne.
Fludernik, M. (1986). ‘‘ ‘Ithaca’ – An Essay in NonNarrativity.’’ In G. Gaiser (ed.), International
Perspectives on James Joyce (pp. 88–105). Troy,
NY: Whitsun.
Fludernik, M. (1996). Towards a ‘‘Natural’’ Narratology. London: Routledge.
Forster, E. M. (1927). Aspects of the Novel. London
and New York: Harcourt, Brace and World.
Hayman, D. (1987). Re-Forming the Narrative:
Toward a Mechanics of Modernist Fiction, Ithaca,
NY: Cornell University Press.
Herman, D. (1994). ‘‘ ‘Sirens’ after Schoenberg.’’
James Joyce Quarterly 31, 473–94.
James, H. (1972). Henry James: Theory of Fiction,
ed. James E. Miller. Lincoln: University of Nebraska Press.
Joyce, J. ([1922] 1986). Ulysses: The Corrected Text,
ed. H. W. Gabler. New York: Random.
179
which Lawrence is so determined to show
that two women cannot live happily together
without a man that, in violation of the realism that governs the events of the rest of the
text, his hero chops down a tree that falls on
and kills the more pushy woman. Other
readers can no doubt supply their own favorite such examples.
19 I wish to thank many individuals at the
‘‘Contemporary Narrative Theory’’ conference
for a number of excellent comments, including Wayne Booth, Royal Brown, Melba
Cuddy-Keene, David Richter, and Dan
Shen; special thanks go to James Phelan and
Peter Rabinowitz for numerous helpful observations and suggestions.
FURTHER READING
Kafalenos, E. (2003). ‘‘The Power of Double Coding to Represent New Forms of Representation:
The Truman Show, Dorian Gray, ‘Blow Up,’ and
Whistler’s Caprice in Purple and Gold.’’ Poetics
Today 24(1), 1–33.
Martin, T. (1998). ‘‘Ulysses as a Whole.’’ In
R. Frehner and U. Zeller (eds.), A Collideorscape
of Joyce (pp. 202–14). Dublin: Lilliput.
McHale, B. (1987). Postmodernist Fiction. London:
Methuen.
O’Donnell, T. D. (1975). ‘‘Thematic Generation in
Robbe-Grillet’s Projet pour une révolution à New
York.’’ In G. Stambolian (ed.), Twentieth Century
French Fiction: Essays for Germaine Brée
(pp. 184–97). New Brunswick, NJ: Rutgers
University Press.
Orr, L. (1991). Problems and Poetics of the Nonaristotelian Novel. Lewisburg, PA: Bucknell University Press.
Peake, C. H. (1977). James Joyce: The Citizen and the
Artist. Stanford, CA: Stanford University Press.
Peterson, R. G. (1976). ‘‘Critical Calculations:
Measure and Symmetry in Literature.’’ PMLA
91, 367–75.
Phelan, J. (1989). Reading People, Reading Plots:
Character, Progression, and the Interpretation
of Narrative. Chicago: University of Chicago
Press.
Rader, R. (1973). ‘‘Defoe, Richardson, Joyce,
and the Concept of Form in the Novel.’’ In
180
Brian Richardson
W. Matthews and R. Rader, Autobiography, Biography, and the Novel (pp. 31–72). Los Angeles:
William Andrews Clark Memorial Library,
UCLA.
Ricardou, J. (1972). ‘‘Naissance d’une fiction.’’ In
J. Ricardou and F. van Rossum-Guyon (eds.),
Nouveau Roman: hier, aujourd’hui, vol 2: Practiques (pp. 379–92). Paris: 10/18.
Richardson, B. (2000). ‘‘Linearity and its Discontents: Rethinking Narrative Form and
Ideological Valence.’’ College English 62, 685–95.
Richardson, B. (ed.). (2002). Narrative Dynamics:
Essays on Time, Plot, Closure, and Frames. Columbus: Ohio State University Press.
Ricoeur, P. (1981). ‘‘Narrative Time.’’ In W. J. T.
Mitchell (ed.), On Narrative (pp. 165–86). Chicago: University of Chicago Press.
Robbe-Grillet, A. (1977). ‘‘Order and Disorder in
Film and Fiction.’’ Critical Inquiry 4, 1–20.
Ryding, W. R. (1971). Structure in Medieval Narrative. The Hague: Mouton.
Sacks, S. (1964). Fiction and the Shape of Belief.
Berkeley: University of California Press.
Sherzer, D. (1986). Representation in Contemporary
French Fiction. Lincoln: University of Nebraska
Press.
Shklovsky, V. (1990). ‘‘The Relationship between
Devices of Plot Construction and General
Devices of Style.’’ In Theory of Prose, trans.
B. Sher (pp. 15–51). Elmswood Park, IL: Dalkey
Archive Press.
Todorov, T. (1981). Introduction to Poetics, trans.
R. Howard. Minneapolis: University of Minnesota Press.
Woolf, V. (1980). Women and Writing. New York:
Harcourt Brace Jovanovich.
11
They Shoot Tigers, Don’t They?:
Path and Counterpoint in The
Long Goodbye
Peter J. Rabinowitz
The Long Goodbye is Raymond Chandler’s longest and most serious novel – and, to
many, his darkest: a meditation on failure and despair, ending in what Frank
MacShane has called a ‘‘blank’’ and a ‘‘void,’’ almost as if he were describing the finale
of Flaubert’s l’Éducation Sentimentale (MacShane 1976: 207). I’m going to challenge, or
at least complicate, that standard reading here, but I’m going to do so circuitously. So
let me start out in a different forest of the night altogether, John Tavener’s 1987
choral setting of Blake’s ‘‘The Tiger.’’ It’s not, to my mind, a particularly effective
setting, but there is a moment of heightened attention on the line ‘‘Did he who made
the Lamb make thee?,’’ where the music’s dense texture – a rich polyphony over a long
drone – abruptly resolves into simple harmony, creating a sense of suddenly clarified
vision. I mention it here not because Chandler, in his dreadful early poetry, showed a
partiality for Blake.1 Nor do I bring it up because Tavener’s decision is musically
ingenious or interpretively subtle. Rather, the moment is valuable for my purposes
because it demonstrates an extremely elementary contrapuntal effect that’s readily
available even to a fledgling composer, but that written narrative, even in the hands of
an expert, can’t reproduce. Or can it?
As someone who divides his life between literary narrative and music, I’ve long
been intrigued with the ways in which creators in one medium in fact do manage to
find analogues for techniques and effects that would seem, on the surface, limited
to the other.2 In the past, I’ve worked mostly on musical analogues to literature – in
particular, on the ways in which absolute music, even without referentiality (and
hence with no immediate way of distinguishing the true or the real from the false or
the imaginary), can nonetheless manage to play games with fictionality or irony. But
spurred by my recent experiences teaching a series of seminars on time and narrative,
I’ve been turning to literary analogues to music – in particular, the ways in which the
single line of narrative prose can engage in contrapuntal games.
182
Peter J. Rabinowitz
The problem facing would-be contrapuntal writers, of course, comes from
the difficulty of expressing simultaneity in words, with their unavoidable one-afterthe-other order. One key insight generating Borges’s famous story ‘‘The Garden of
Forking Paths’’ is that the only way for a writer of prose narrative to express temporal
simultaneity (in this case, ‘‘parallel’’ times in which different choices are made by the
characters) is to project it onto spatial contiguity (here, consecutive versions of a
‘‘single’’ chapter), leaving it up to the reader to reconceptualize the relationship
between the events.3 Thus, for instance, the double time frame of Horace McCoy’s
Depression downer They Shoot Horses, Don’t They? – which presumably represents the
protagonist’s memories of events as he is listening to the judge’s pronouncement of
the verdict in his murder trial – is indicated by alternating chapters and interchapters;
Joyce Carol Oates, in her short story ‘‘The Turn of the Screw,’’ tries to convey
simultaneity by using columns; Carol Shields’s Happenstance binds together two
separate novels that cover the same time period from different perspectives.4 Yet as
became increasingly clear to me during my time seminars, narrative fakes simultaneity in more ways than I had realized, many of which have little to do with
simultaneous events in the physical world. There are, for instance, techniques for
representing simultaneous thoughts on a variety of different levels of consciousness
and reality (memory, fantasy, whatever). Narrative writers have also figured out ways
to represent simultaneity (or, perhaps more accurately, ‘‘copresence’’) in visions of a
sedimented present which contains several layers, a present that is ‘‘caused by’’ or is a
‘‘repetition’’ or an ‘‘echo’’ of a past, real or imagined, that continues to linger,
somehow, in the present: whether in Brontë’s Wuthering Heights or Faulkner’s Absalom,
Absalom! or more radically in Borges’s claim that any human proposition implies ‘‘the
entire universe; to say ‘the jaguar’ is to say all the jaguars that engendered it, the deer
and turtles it has devoured, the grass that fed the deer, the earth that was mother to
the grass, the sky that gave light to the earth’’ (Borges 1998: 252).
More generally, we have the contrapuntal intertwining of the authorial and narrative
audiences (Rabinowitz 1998: 93–104), of tensions and instabilities (Phelan 1996: 30),
and of the mimetic, synthetic, and thematic dimensions of characters (Phelan 1989:
2–3); we also have the counterpoint of perspectives in any ironic discourse, and the kind
of polyphony celebrated by Bakhtin (see also Richardson, this volume). Indeed, one can
argue that narrative is, for all its apparent reliance on a single line, linearity, fundamentally contrapuntal because it is founded on the duality of story and discourse (or fabula/
sjuzhet or whatever alternative terms you want to use) (e.g., Chatman 1978). I’d like to
start from that fundamental premise and develop it further, suggesting that the story/
discourse distinction is too simple, and that narratives actually get quite far beyond the
two-part inventions that scheme would suggest.
More specifically, my argument is that we need to supplement the story/discourse
distinction with a third term: path. Seymour Chatman once claimed that ‘‘time passes
for all of us in the same clock direction’’ (Chatman 1978: 98) – and in a certain sense,
that’s true. But the Theory of Relativity makes it clear that the order of events is,
under certain circumstances, dependent on the situation of the observer;5 and that’s
Path and Counterpoint in The Long Goodbye
183
the case even in our everyday Newtonian lives. Different people (or, in literature,
different characters) experience events in different orders. In other words, a character’s
order of experience may conform to neither the story order nor the discourse order.
Let me exemplify the problem with a concrete example from a novel: as happens in
so many narratives, a character takes a trip (Event A), returns from the trip (Event C),
and gathers his friends together to tell them what adventures befell him while he was
away (Event E). In this case, the text we’re reading (a report of the traveler’s narration
by one of those friends) includes the following anecdote about a woman the traveler
met while on his journey, an anecdote that occasions a brief interruption in the
traveler’s story:
[She] ran along by the side of me, occasionally darting off on either hand to pick flowers
to stick in my pockets . . . [Event B] And that reminds me! In changing my jacket
I found . . . [Event D]
[He] paused, put his hand into his pocket, and silently placed two withered flowers, not unlike
very large white mallows, upon the little table [Event F]. Then he resumed his narrative.
At first, this looks like a standard kind of anachronic zigzag (nearly a rondo): although
the discourse order is ACEBDF, experienced readers will know how to reconstruct the
underlying story order, ABCDEF. In fact, though, that’s not the story order at all. For
this passage comes from H. G. Wells’s The Time Machine ([1895] 1934: 49–50, first
ellipsis added, italics in original), and Event B occurs in the distant future.6 The story
order is thus ACDEFB. So what do we call the structure ABCDEF? It clearly has a
significant role in the novel, for it’s the order of the events as experienced by the
protagonist, the Time Traveler. But it’s neither the order of events as they happen(ed)
nor the order in which the narrator presents them – no surprise, since the essence of
time travel is that you experience things ‘‘out of order.’’ I call this order the path of the
protagonist. And the more time travelers there are, the more paths there are.7
Were this kind of multiplicity simply a characteristic of time-travel narratives, it
would be at best a minor narratological blip. In fact, though, it has broad implications
for narrative more generally. Standard narratology encourages us to think of ‘‘discourse’’ from the perspective of the author or narrator. That is, to the extent that we’re
talking about ‘‘order’’ (and it’s order rather than, say, focalization or voice that I’m
concerned with here), discourse is usually considered to involve, in Gerald Prince’s
terms, ‘‘the order of presentation of situations and events’’ (Prince 1987: 21, emphasis
added). But if we think as reader critics, it’s possible to reconceptualize discourse
order as the order in which the situations and events are received or experienced by the
reader. And once we’re talking about the order of reception, why so privilege the
reader’s that we’re unable to recognize other orders that emerge in texts? After all,
characters, too, ‘‘receive’’ the events, either by experiencing them directly or by
hearing about them second hand, sometimes from the narrator as a narratee but
more often in some other way. Rarely do they all do so in the same order or from the
same source. Indeed, we could easily reconceptualize dramatic irony as a clash of
184
Peter J. Rabinowitz
paths: the clash between Oedipus’s path and Tiresias’s, for instance, or the clash
between the paths of Grace Ansley and Alida Slade, the main characters of Edith
Wharton’s ‘‘Roman Fever.’’ As a result, most narratives turn out, on closer examination, to be gardens of fuguing paths.
So what’s at stake here? What happens if we take the existence of the multiple paths
seriously? There are, I believe, both theoretical and interpretive payoffs to the
consideration of path as a third term. To begin with theory: the notion of ‘‘path’’
points out some of the blind spots in our traditional terms. Let me just point to two
biases introduced by the traditional thinking of order through the story/discourse
dyad, as a distinction between ‘‘order of events’’ and the ‘‘order of presentation.’’
First, as I’ve said, the binary distinction tacitly assumes that those two orders cover
the ground – as if it weren’t possible for the order of experience or perception to be
quite different. And that assumption tends, if only subtly, to place more stress on
event than on experience or perception. Now given all the attention paid to focalization, especially in analysis of James and post-Jamesian fiction, it would be hard to
justify the claim that experience has been erased. Still, our terminology encourages us
to think of event as somehow prior to or more significant than or more basic than
experience. Thus, while Gerald Prince’s invaluable Dictionary of Narratology and the
forthcoming Routledge Encyclopedia of Narrative Theory include entries for ‘‘event,’’ there
are no parallel entries for ‘‘experience.’’ Of course, experience isn’t entirely expunged:
David Herman’s excellent discussion of ‘‘event-types’’ in the Routledge Encyclopedia
does talk about ‘‘states of mind’’ and ‘‘processes of reflection.’’ But events, in the sense
of ‘‘time- and place-specific transitions from some source state . . . to a target state,’’ are
surely primary (Herman forthcoming).
Second, to the extent that narrative theory does highlight experience, the story/
discourse distinction tends to privilege either the narrator’s or the protagonists’
experience or else – as, say, in James Phelan’s valuable notion of ‘‘progression’’ (Phelan
1996) – the reader’s experience. Even if you throw in unreliable narration, you end up
with a triad: (implied) author, narrator, (implied) reader. My approach, in contrast,
makes us take the experiences of other characters more seriously on the level of the
narrative audience. Or, more accurately, it encourages us to take them seriously in a
different way than the way promoted by classical narratology: specifically, if we take
the notion of path (and experience) seriously, we might want to consider a narratology
that included not only ‘‘actants’’ but also ‘‘passants’’ – those (especially including those
beyond the narrator and his or her chosen focalizers) on whom impressions are
registered.8
Since our categories influence the way we read, this theoretical shift to include ‘‘path’’
as a third term also has consequences on interpretation. For instance, taking other
characters seriously encourages a kind of refocalization, a rethinking of a narrative in
terms of how it’s experienced from positions other than those focalized by the narrator.
Needless to say, most narratives invite us to follow one path – or one limited set of
Path and Counterpoint in The Long Goodbye
185
paths – that’s presented as more relevant to its purposes than others. At the same
time, however, most narratives of any complexity offer other paths that, when
pursued, offer significant illuminations of the literary landscape.
Sometimes, these paths are simply potentials that the text does not fully develop.
Take Chekhov’s story ‘‘Lady with a Dog.’’ It’s a classic ‘‘experience-saturated’’ text, one
in which there are few real ‘‘events.’’ But even readers ready to focus on inward
experience are likely to consider only Anna’s and Dmitri’s experiences. That is, even
though (unusually for Chekhov) the story begins with a reference to public knowledge
(‘‘People were saying that someone new had shown up on the promenade: a lady with a
dog’’) (Chekhov 1964: 173, my translation) – and even though the first part of the
story takes place largely in public places – few of us are apt to ask ourselves how the
nameless gossips are viewing Anna once her relationship with Dmitri begins. (In fact,
we’re not even apt to ask ourselves how their spouses are viewing the affair, although
Anna’s husband suspects that something is going on.) An analysis that took ‘‘passants’’
seriously might well give us new perspectives from which to reflect on the story.
Likewise, what happens if we take seriously Sonia’s path in Crime and Punishment, a
path that’s quite different from that of the reader, who knows of Raskolnikov’s crime
from the beginning? What can thinking about the book this way teach us about
Dostoevsky’s formal artistry, about Sonia’s psychology, about Dostoevsky’s philosophy
of acceptance? I didn’t choose those three areas – artistry, psychology, philosophy – at
random; they are of course aspects of what Phelan (1989: 2–3) calls the synthetic,
mimetic, and thematic dimensions; and path opens up our understanding of all three.
Sometimes, in contrast, these new paths are more actively discouraged – and
following them results in ‘‘counterreadings’’ that resist or undermine those most
clearly invited by the text. What happens if we try to read Hammet’s The Maltese
Falcon in terms of Brigid O’Shaughnessey’s path – or if we take seriously Albertine’s
path in In Search of Lost Time? There is, in fact, a whole genre of ‘‘rewritings’’ – best
typified, perhaps, by Jean Rhys’s Wide Sargasso Sea – that recenter familiar novels by
following new paths.9
Even beyond this, however, there are novels where the concept of path opens up
fundamental features of the narrative structure in a way that uncovers what the
narrative itself is depending on. In these cases, I’d argue, attending to path doesn’t
simply enrich the intended meaning – much less resist it. Rather, it makes that
meaning available in the first place. Let me demonstrate in more detail by returning
to The Long Goodbye.
Now, in a sense, nearly all detective stories make a gesture toward the importance of
path by engaging an ‘‘If I had but known’’ structure – but The Long Goodbye gives that
generic gesture unusual psychological and thematic weight, since the consequences of
lack of knowledge are finally so detrimental to the well-being of the protagonist,
detective Philip Marlowe. Marlowe sets up this structure toward the beginning,
saying of his friend Terry Lennox, ‘‘He would have told me the story of his life if I
had asked him. But I never even asked him how he got his face smashed. If I had and
186
Peter J. Rabinowitz
he told me, it just possibly might have saved a couple of lives. Just possibly, no more’’
(Chandler 1992: 22). And Marlowe’s on-again off-again friend Bernie Ohls, a lieutenant working out of the LA Sheriff’s office, recapitulates the gesture toward the end:
‘‘If you had connected up Wade and the Lennox frail for me the time Wade got dead I’d
have made out. If you had connected up Mrs. Wade and this Terry Lennox, I’d have had
her in the palm of my hand – alive. If you had come clean from the start Wade might be
still alive. Not to mention Lennox.’’
(Chandler 1992: 337)
This kind of knowledge-curve is heightened by a stylistic tic: Marlowe, as narrator,
often postpones giving us the information we need in order to understand his
position. It’s not simply that he doesn’t tell us what he knows at the discoursemoment of his narration; he often keeps us in the dark about what he knows at the
story-moment, too.10 He does this not only in ways that heighten the mystery; often,
it’s just a brief twitch that has little to do with the question of who-dun-it. Thus,
Marlowe first describes the gangster Mendy Menendez as if he were a complete
stranger (Chandler 1992: 74); but we learn two pages later that Marlowe knows
exactly who he is.11
As I’ve said, Chandler’s novel is traditionally read as a story of failure and frustration,
chronicling how Marlowe’s friendship with Terry Lennox – a friendship for which he
has had made serious sacrifices – unravels under what Marlowe calls Terry’s moral
defeatism. To return to Frank MacShane’s description: ‘‘At the end there is just a blank,
a void that has to be filled with something, even a code of behavior that sounds
sentimental’’ (MacShane 1976: 207). Or in Fredric Jameson’s words,
The form of Chandler’s books reflects an initial American separation of people from each
other, their need to be linked by some external force (in this case the detective) if they
are ever to be fitted together as parts of the same picture puzzle. And this separation is
projected out onto space itself: no matter how crowded the street in question, the
various solitudes never really merge into a collective experience.
(Jameson 1970: 633)
But if we look at the odd showdown between Marlowe and gangster Mendy Menendez
toward the end of the novel in terms of story/discourse/path, we cast some doubt on
these by now canonical interpretive claims.
The scene is itself problematic – indeed, many critics skip over it entirely. Thus,
McCann gives a plot summary that leaves it out, calling the events revealing Eileen’s
guilt the novel’s ‘‘climactic scenes’’ (McCann 2000: 178). R. W. Lid’s summary (Lid
1969: 173–6) similarly ignores this scene, as do MacShane’s (1976: 199–200) and
Hiney’s (1997: 207–9). And even those who take it into account miss much of the
(counter)point: Johanna Smith says simply that Ohls ‘‘sets up Marlowe as a decoy so
that he can capture the gambler Menendez’’ (Smith 1995: 197); Marling also mentions it, but leaves much out: ‘‘He . . . goes home, where Menendez is waiting to beat
Path and Counterpoint in The Long Goodbye
187
him up. But it is Menendez who gets beat up: his hired thugs turn out to be disguised
policemen sent by Lieutenant Ohls’’ (Marling 1986: 134).
I suspect these critics treat the scene cavalierly because they consider it simply
filler. After all, thematic resonances aside, Chandler’s title has a formal meaning as
well: the novel takes longer to say goodbye than most of its generic siblings. In the
Vintage edition, the unmasking of the killer comes on page 313. That’s already long
for a detective novel of this era – but Chandler gives us a Beethovenian coda, taking
another 66 pages to wrap things up. And this scene between Marlowe and Menendez,
which occurs between pages 343 and 352, can seem like simply another delay to keep
the novel from ending.12
What happens during this coda? To summarize – and summarizing a Chandler
novel is never an easy task – the ‘‘solution’’ to the mystery comes when Eileen Wade
commits suicide, and takes responsibility for the murder of her novelist husband
Roger (until then considered a suicide) and for that of Sylvia Potter Lennox (a murder
for which Terry Lennox had taken the rap in a self-sacrificing suicide note of his own).
The powers that be want to suppress Eileen’s confession, since it will only create bad
publicity for the wealthy and influential Potter family and for the District Attorney,
who had botched the case from the beginning. But Lieutenant Ohls lets Marlowe steal
a photostat of Eileen’s confession. Marlowe delivers it to the press because he wants to
clear Terry’s name – even though he knows that publication will anger not only the
powers that be, but also the mobster Menendez. Menendez, along with Las Vegas
casino operator Randy Starr, had been saved by Terry’s heroic action during the war –
he has been indebted to Terry ever since, and has consistently asked Marlowe to stop
stirring things up. Marlowe ignores warnings that he’s acting rashly, even when Linda
Loring (Sylvia’s sister) tells him that he’s being set up: ‘‘ ‘Do you know how they shoot
tigers?,’ ’’ she asks. ‘‘ ‘They tie a goat to a stake and then hide out in a blind. It’s apt to
be rough on the goat’ ’’ (Chandler 1992: 342). And she urges him not to imitate Terry,
who has been a fall-guy himself.
Sure enough, shortly after that conversation, Marlowe returns to his house to find
himself facing Menendez and a trio of ‘‘hard boys.’’ Marlowe is punched and pistolwhipped, and just as he thoughtlessly starts to fight back – an act that threatens to get
him killed – Lieutenant Ohls appears, like a deus ex machina, to confirm the tigershooting scenario. Ohls’s reasons for letting him take the photostat had nothing to do
with the murders. Rather, he is using Marlowe as bait for Menendez, who had
previously crossed the line of acceptable mobster conduct by beating up a crooked
vice-squad cop named Magoon. Menendez’s apparent bruisers are, in fact, Nevada cops
supplied by casino operator Starr, who, we’re told, is just as angry as Ohls that
Menendez has upset the delicate balance of power: ‘‘ ‘Somebody in Vegas don’t like the
way you forget to clear with them’ ’’ (Chandler 1992: 349), Ohls remarks as the hard
boys lead Menendez away, presumably to his death. As Marlowe coolly puts it, ‘‘ ‘The
coyotes out in the desert will get fed tonight’ ’’ (Chandler 1992: 350).
Why is this scene here? One might argue that it is plot-driven: after all, Chandler
needs to clear up the Magoon subplot. But in fact that subplot is very sketchy, and
Peter J. Rabinowitz
188
introduced late in the novel. If anything, it seems more likely that Chandler
introduced the Magoon subplot in order to set up the tiger-trap scene – which only
adds force to the question of why it’s included. One might argue, alternatively, that
the scene is simply the result of Chandler’s famous difficulty with plot: he often had
trouble keeping things from getting away from him. In that regard, it’s roughly
analogous to a climax-postponing episode in The Postman Always Rings Twice (Chandler might cringe at the comparison [Chandler 1981: 26]) – Frank’s brief liaison with
Madge, a woman who, by a bizarre coincidence, raises tigers and other big cats. But
I’d like to propose that the scene burns a lot more brightly than this – especially when
we consider how it ends. For the tiger-trap scene has its own long goodbye, which
occurs 25 pages later, during Marlowe’s last conversation with Terry (whose suicide,
we learn, has been faked), a conversation facilitated by Starr. Almost as an afterthought, we’re told that Menendez is alive and well in Acapulco, leaving us to infer a
key shadow-event that’s never explicitly mentioned in the book: Starr has set up Ohls
just as Ohls has set up Marlowe, planning the whole scene as a way to spirit Menendez
away from the police.
So much for the events. But what about the experiences? Let me outline the story,
the discourse, and the various paths in a chart:
STORY
1
2
3
4
5
6
DISCOURSE
OHLS
MARLOWE
or
MENENDEZ
or
or
STARR
3
1
3
3
3
1
2
1
Ohls sets up trap
(Shadow event: Starr sets up Ohls)
Photostat stolen by Marlowe
Linda warns Marlowe
Scene in the house
Marlowe’s final conversation with Terry
4
3
4
4
5
2
1
2
5
5
1
5
1
3
3
3
1
6 (implies 2)
5
1
2
5
5
5
6 (implies 2)
6 (implies 2)
6
The variety of paths has synthetic, mimetic, and thematic significance. First, and
perhaps least important, looking at the paths gives us an appreciation for the formal
intricacy of Chandler’s plotting. I’d argue that the moment we learn that Menendez is
in Mexico is comparable to the moment of resolution in Tavener’s ‘‘The Tiger’’ – a
moment where the dissonant clashes of the novel’s contrapuntal lines give way to a
piercing clarity in which everyone understands what the tiger-trap scene was really all
about. But even if you don’t accept that particular analogy, it would be hard to deny
that Chandler’s handling of the contrapuntal paths reveals considerable technical
ingenuity.
Path and Counterpoint in The Long Goodbye
189
On the mimetic level, working out the paths clarifies just how much we don’t
know about the characters’ experiences – in particular, Menendez’s and Marlowe’s.
Their heated exchange is based on illusion – whatever the apparently deadly terrors of
this scene, neither is actually in danger. But here we have another of the knowledge
twitches – one, in this case, that’s postponed indefinitely. Does either of them know
that he’s not in danger, and if so, does he know whether the other knows it or not? We
never find out, so the psychological energy of the scene has a curious resonance –
although it only has that resonance if we think carefully in terms of path. Geoffrey
Hartman (1975: 216) has argued that in Chandler (and Ross Macdonald) ‘‘the only
person . . . whose motives remain somewhat mysterious, or exempt from this relentless
reduction to overt and vulnerable gestures, is the detective.’’ This scene suggests
something more complex: it’s impossible to reconstruct either Menendez’s or Marlowe’s path.
But it’s on the thematic level that the analysis in terms of path pays its most
significant rewards. At first glance, Starr is a background figure, someone who barely
shows up in the novel, and even then at the other end of a telephone. But the chart
suggests that, in this climactic scene at least, he’s the cantus firmus or the drone around
which the other figures are swirling – or, to put it less metaphorically, Starr is the one
who knows (and controls) the story that everyone else is trying to reconstruct.
Marlowe, in contrast, is the last of the major characters to understand that story,
suggesting that our notion of center and periphery may be skewed. What happens if
we reconsider the tiger-trap scene from Starr’s perspective? Refocalized in this way, it
emerges as a replay of Terry’s fake suicide, which was an elaborate hoax set up by
Starr and Menendez to fashion Terry with an escape. But although Linda Loring has
worried (and thus warned the reader) that Marlowe will become another Terry,
although the pistol-whipping numbs Marlowe’s face in a way that mimics the
disfigurement of Terry’s own face (the result of Nazi torture during the war), in fact
it’s Menendez, not Marlowe, who takes Terry’s role in this scene.
Most important, the refocalized tiger-trap scene, like Terry’s suicide, gives the lie to
Marlowe’s final observation. In one of Chandler’s more memorable endings, the novel
concludes with the hero in despair, saying, ‘‘I never saw any of them again – except the
cops. No way has yet been invented to say goodbye to them’’ (Chandler 1992: 379).
Chandlerians know that Marlowe is technically wrong in fact here because his next
novel Playback ends with his plans to marry Linda Loring.13 But if we look at the
world from Starr’s perspective, we see that Marlowe is wrong in principle at a more
profound level. No way has yet been invented to say goodbye to the cops? But that’s
precisely what Terry and Menendez, with Starr’s help, have been able to do. And
they’ve been able to do so in part because of their absolute, unquestioning, unbreakable friendship.
The end of the novel, then, doesn’t suggest that friendship is impossible, much less
that without the detective, as Jameson puts it, the various solitudes fail to merge.
Rather, it reveals Marlowe’s inability to break away from his solitude, his inability to
join this trio, this clique that manifests, in its actions, precisely the commitment he so
190
Peter J. Rabinowitz
clearly values. And reading in terms of path allows us to reconceive the long war of
words between Marlowe on the one hand and Menendez and his minions on the other,
not as Marlowe’s distaste for his hoodlum flash, but as Marlowe’s envy for his tight
circle of friendship. The ending is all the more poignant to readers familiar with
Chandler’s earlier work, especially The Big Sleep, where Marlowe’s hard-boiled approach gives him a reliable, even privileged, vision of the world. In The Long Goodbye,
one senses that he has so exaggerated his persona that he’s lost his ability to interpret
the world around him.
Now I’m not so naı̈ve as to believe that my interpretation of the novel would not be
available without my terminological meddling; there is, as they say (for reasons I’ve
never understood), more than one way to skin a cat, and that probably goes for tigers,
too. Still, the terminology illuminates it, makes it easier to see. Certainly, I had never
seen the ending of The Long Goodbye in this light until I decided, more or less
arbitrarily, to see what would happen when I applied this theoretical perspective.
And that seems sufficient justification for adding path to our categories of analysis.14
NOTES
1 How close are the ties between Chandler and
Blake? It’s hard to tell. But by some quirk of
fate, when I opened Chandler’s book of verses
to see if anything relevant caught my eye, the
poem that appeared at random was ‘‘A Lament
for Youth,’’ which centers (as The Long Goodbye
does) on ‘‘sweet and long’’ yearning – and
which begins with the line ‘‘From the forests
of the night’’ (Chandler 1973: 36).
2 For a fuller discussion of the relationships
between absolute music and written narrative,
see Fred Maus’s essay in this volume.
3 The same need for projection haunts Joseph
Frank’s notion of ‘‘reflexive reference’’ in his
influential essay on ‘‘spatial form’’ (Frank
1963).
4 Film has it easier in this regard, although even
so, films like Time Code are the exception,
simultaneity more often being represented
through alternation, as in 24, than through
long-term use of a split screen.
5 Specifically, when two spatially separated
events occur in such a manner that no beam
of light can get from A to B before B occurs
(and vice versa), the order of those events will
vary according to the frame of reference of the
observer.
6
See also the temporal confusion in the following sentence: ‘‘It was here that I was destined,
at a later date, to have a very strange experience – the first intimation of a still stranger
discovery – but of that I will speak in its
proper place’’ (Wells [1895] 1934: 24).
7 For a truly Bachian, well-temporaled experience of multiple paths, I’d recommend Hilary Brougher’s dizzying time-travel film
Sticky Fingers of Time.
8 The ‘‘receiver,’’ whose function is to receive
the object sought by the ‘‘subject,’’ is quite a
different category; so, of course, is ‘‘ficelle.’’
9 Gerry Brenner’s ‘‘performative criticism’’ – in
which the critic writes from inside a literary
text by ‘‘becoming or impersonating a character in or close to a text’’ – can be seen as a
way of using path as a critical tool, too
(Brenner 2004: 2).
10 For that reason, this is not an instance of
what Phelan calls ‘‘paradoxical paralepsis’’
(Phelan 1996: 82–104). For further discussion of paradoxical paralepsis, see Alison
Case’s essay in this volume.
11 Similarly, Marlowe knows the connection between Endicott and Eileen well before we do
(Chandler 1992: 96).
Path and Counterpoint in The Long Goodbye
12 For an excellent appreciation of the form of
this novel – in particular, the way Chandler
uses ‘‘background action’’ to convey its ideology – see Richter (1994).
13 Chandler never finished the novel in which
they are actually married.
REFERENCES
AND
Brenner, G. (2004). Performative Criticism: Experiments in Reader Response. Albany: State University of New York Press.
Borges, J. L. (1998). Collected Fictions, trans. A.
Hurley. New York: Penguin Books.
Chandler, R. (1973). Chandler Before Marlowe: Raymond Chandler’s Early Prose and Poetry, 1908–
1912, ed. M. J. Bruccoli. Columbia: University
of South Carolina Press.
Chandler, R. (1981). Selected Letters of Raymond
Chandler, ed. F. MacShane. New York: Columbia University Press.
Chandler, R. ([1953] 1992). The Long Goodbye.
New York: Vintage.
Chatman, S. (1978). Story and Discourse: Narrative
Structure in Fiction and Film. Ithaca, NY: Cornell
University Press.
Chekhov, A. P. ([1899] 1964). Izbrannye Proizvedeniia: Tom tretii. Moscow: Izdatel’stvo Khudozhestvennaia Literatura.
Frank, J. ([1945] 1963). ‘‘Spatial Form in Modern
Literature.’’ In The Widening Gyre: Crisis and
Mastery in Modern Literature (pp. 3–62). New
Brunswick, NJ: Rutgers University Press.
Hartman, G. (1975). The Fate of Reading and Other
Essays. Chicago: University of Chicago Press.
Herman, D. (forthcoming). ‘‘Events and EventTypes.’’ In D. Herman, M. Jahn, and M.-L.
Ryan (eds.), The Routledge Encyclopedia of Narrative Theory. London: Routledge.
Hiney, T. (1997). Raymond Chandler: A Biography.
New York: Atlantic Monthly Press.
Jameson, F. (1970). ‘‘On Raymond Chandler.’’
Southern Review 6(3), 624–50.
191
14 Thanks to Hilary Dannenberg, Elizabeth
Jensen, Jessica Kent, James Phelan, Haley
Reimbold, and Brian Richardson for their
assistance and their valuable suggestions.
FURTHER READING
Lid, R. W. (1969). ‘‘Philip Marlowe Speaking.’’
Kenyon Review 31, 153–78.
MacShane, F. (1976). The Life of Raymond Chandler.
New York: E. P. Dutton & Co.
Marling, W. (1986). Raymond Chandler. Boston:
Twayne Publishers.
McCann, S. (2000). Gumshoe America: Hard-Boiled
Crime Fiction and the Rise and Fall of New Deal
Liberalism. Durham, NC: Duke University
Press.
Phelan, J. (1989). Reading People, Reading Plots:
Character, Progression, and the Interpretation of
Narrative. Chicago: University of Chicago Press.
Phelan, J. (1996). Narrative as Rhetoric: Technique,
Audiences, Ethics, Ideology. Columbus: Ohio State
University Press.
Prince, G. (1987). Dictionary of Narratology. Lincoln: University of Nebraska Press.
Rabinowitz, P. J. ([1987] 1998). Before Reading:
Narrative Conventions and the Politics of Interpretation. Columbus: Ohio State University Press.
Richter, D. (1994). ‘‘Background Action and
Ideology: Grey Men and Dope Doctors in Raymond Chandler.’’ Narrative 2, 29–40.
Smith, J. M. ([1989] 1995). ‘‘Chandler and the
Business of Literature.’’ In J. K. Van Dover
(ed.), The Critical Response to Raymond Chandler
(pp. 183–201). Westport, CN: Greenwood
Press (originally in Texas Studies in Language
and Literature 31(4), 592–610).
Wells, H. G. ([1895] 1934). The Time Machine. In
Seven Science Fiction Novels of H. G. Wells. New
York: Dover.
12
Spatial Poetics and Arundhati
Roy’s The God of Small Things
Susan Stanford Friedman
Space is not the ‘‘outside’’ of narrative, then, but an internal force, that shapes it from
within.
(Franco Moretti 1998: 70)
Every story is a travel story – a spatial practice.
(Michel de Certeau 1984: 115)
In 1967, Michel Foucault presciently observed: ‘‘The great obsession of the nineteenth
century was, as we know, history. . . . The present epoch will perhaps be above all the
epoch of space’’ (Foucault 1986: 22). In citing Foucault’s ‘‘Of Other Spaces,’’ geographer Edward W. Soja charged that by 1989 ‘‘no hegemonic shift has yet occurred to
allow the critical eye – or the critical I – to see spatiality with the same acute depth of
vision that comes with a focus on durée. The critical hermeneutic is still enveloped in a
temporal master-narrative, in a historical but not yet comparably geographical
imagination’’ (Soja 1989: 11). But since Soja’s call in Postmodern Geographies for a
compensatory emphasis on spatiality to counteract the hegemony of temporal modes
of thought, there’s been a sea change in cultural theory, a veritable flood of spatial
discourses proliferating across the disciplines in the 1990s, as an effect (I believe) of
the intensified form of globalization in the late twentieth century.
Narrative theory, however, has largely continued its privileging of narrative time
over narrative space, with a few notable exceptions. In spite of M. M. Bakhtin’s
insistence in the 1920s and 1930s on topos as coconstituent of narrative along with
chronos, prominent narrative theorists from Paul Ricoeur and Gérard Genette to Peter
Brooks mute or altogether delete considerations of space in their analysis of narrative
discourse and narrative as a mode of human cognition. Space in narrative poetics is
often present as the ‘‘description’’ that interrupts the flow of temporality or as the
‘‘setting’’ that functions as static background for the plot, or as the ‘‘scene’’ in which
Spatial Poetics and The God of Small Things
193
the narrative events unfold in time. Even when theorists acknowledge that particular
settings such as Thomas Hardy’s Wessex or William Faulkner’s Yoknapatawpha
County are charged with signification, these acknowledgements do not disrupt the
prevailing view of the relation between space and time in narrative. I want to
challenge this prevailing view by first reviewing some standard theoretical formulations, then exploring some alternative views, and finally testing a revisionist emphasis
on space in narrative poetics with a brief reading of Arundhati Roy’s prize-winning
and controversial novel, The God of Small Things (1997).
Space, Time, and Narrative Poetics
As a form of telling, narrative exists in time: a narrative takes time to tell and tells
about a sequence of events in time. It is perhaps understandable, therefore, that
temporality has dominated discussions of narrative poetics. Ricoeur, the preeminent
theorist of narrative temporality, writes, for example: ‘‘My first working hypothesis is
that narrativity and temporality are closely related. . . . [I]ndeed, I take temporality to
be that structure of existence that reaches language in narrativity and narrativity to
be the language structure that has temporality as its ultimate reference’’ (Ricoeur
1981: 165). In Reading for the Plot, Brooks’s challenge to structuralist narratology
takes the form of giving priority to temporality in narrative: ‘‘We might think of plot
as the logic . . . that develops its propositions only through temporal sequence and
progression. . . . And plot is the principle ordering force of those meanings that we try
to wrest from human temporality’’ (Brooks 1984: xi). H. Porter Abbott in The
Cambridge Introduction to Narrative similarly focuses on time as the key property of
narrative. ‘‘Narrative,’’ he writes emphatically, ‘‘is the principle way in which our species
organizes its understanding of time’’ (Abbott 2002: 3). ‘‘Narrative,’’ he sums up aphoristically, ‘‘gives us what could be called the shape of time’’ (Abbott 2002: 11). As ‘‘the
representation of an event or series of events,’’ narrative is distinct from ‘‘description’’
(Abbott 2002: 12).
Genette’s magisterial Narrative Discourse identifies three main components of
narrative discourse for analysis: tense (‘‘temporal relations between narrative and
story’’); mood (‘‘modalities [forms and degrees] of narrative ‘representation’ ’’); and
voice (‘‘the narrative situation . . . : the narrator and his audience, real or implied’’)
(Genette 1980: 30–1). Three of the five chapters are devoted to the temporal
dimension of narrative, and space hardly enters into his account of narrative discourse
at all. He mentions in passing only the way in which ‘‘narrative exists in space and as
space,’’ by which he means the pages of a book that require ‘‘the time needed for
‘consuming’ it [as] the time needed for crossing or traversing it, like a road or a field’’
(Genette 1980: 35). He is interested in ‘‘the narrating situation, the narrative matrix
– the entire set of conditions (human, temporal, spatial) out of which a narrative
statement is produced,’’ but his chapter on ‘‘voice’’ – that is, the narrative situation –
does not develop the spatial dimension of ‘‘situation’’ (Genette 1980: 31).1
194
Susan Stanford Friedman
This prevailing privileging of time over space in narrative poetics is evident as well
in a fascinating thread of exchanges on the Narrative listserv in 1997 in response to
Ruth Ronen’s article in Narrative on the canonical binary of ‘‘description’’ and
‘‘narration’’ in narrative studies.2 Monika Fludernik observes, for example, that
‘‘characters (agents) require a setting (description), whereas narrative (plot) is configured most in terms of actions (chronology)’’ (December 1, 1997). This posting elicited
from prominent narratologist Gerald Prince the comment that ‘‘description itself can
be narrative, of course; but it has very low narrativity because it stresses the spatial
rather than temporal, the topological rather than chronological existence of events’’
(December 1, 1997). In reply, Marie-Laure Ryan suggests that in many writers ‘‘the
description is reasonably free from narration’’ and that ‘‘Descriptions . . . can be
skipped by the reader without serious damage for the understanding of the plot’’
(December 4, 1997), a practice that is particularly prevalent in the reading of
description-rich Victorian novels.3 In such formulations, Bakhtin’s concept of narrative chronotope shifts subtly into narrative chronotype, with the time–space axes of
story morphing into a figure–ground binary. What happens to characters in time is
the ‘‘figure’’ we pay attention to; where the plot happens in space is the ‘‘ground’’ we
can ignore at will. In this view, narrative is the function of temporal sequence and
causation that take place against a static background of spatial setting.
It is worth returning, I believe, to Bakhtin’s resonant definition of chronotope in his
essay ‘‘Forms of Time and of the Chronotope in the Novel’’ in The Dialogic Imagination.
He acknowledges his debt to Einstein’s theory of relativity and then writes:
We will give the name chronotope (literally, ‘‘time space’’) to the intrinsic connectedness
of temporal and spatial relationships that are artistically expressed in literature. . . .
What counts for us is the fact that it expresses the inseparability of space and
time. . . . Time, as it were, thickens, takes on flesh, becomes artistically visible; likewise,
space becomes charged and responsive to the movements of time, plot and history. This
intersection of axes and fusion of indicators characterizes the artistic chronotope.
(Bakhtin 1981: 84)
Bakhtin’s sense of the mutually constitutive and interactive nature of space and time
in narrative has largely dropped out of narrative poetics. Following Soja, I suggest
that we need a compensatory emphasis on space in order to bring back into view
Bakhtin’s continual attention to the function of space as an active agent in the
production of narrative. We need a topochronic narrative poetics, one that foregrounds
topos in an effort to restore an interactive analysis of time with space in narrative
discourse.
Walter Benjamin’s essay ‘‘The Storyteller’’ offers an implicitly topochronic account
of narrative, one that anticipates the compelling spatial homonym in James Clifford’s
‘‘Traveling Cultures,’’ in which he identifies ‘‘roots’’ and ‘‘routes’’ as opposing but
interrelated spatialized dimensions of culture (Clifford 1997: 88). In tracing the roots
of modern narrative to archaic forms of storytelling, Benjamin divides early narrative
Spatial Poetics and The God of Small Things
195
into two ‘‘archaic types’’ whose defining nature is based on the storyteller’s relationship to space. The first type is the story told by the seaman, the man ‘‘who has
come from afar’’ and ‘‘has something to tell about’’ (Benjamin 1969: 84). The second
type is told by the tiller of the soil, the ‘‘man who has stayed at home . . . and knows
the local tales and traditions’’ (Benjamin 1969: 84). ‘‘The realm of storytelling in its
full historical breadth,’’ he continues, ‘‘is inconceivable without the most intimate
interpenetration of these two archaic types’’ (Benjamin 1969: 85). Home and elsewhere – both spatial locations – are for Benjamin the coconstituents of story, not
incidental to it, as narrative tradition has evolved through time and across many
societies.
In ‘‘Spatial Stories,’’ Michel de Certeau theorizes narrative as a ‘‘practice of everyday
life’’ for which the building blocks are the spaces that enable various kinds of cultural
practices or movements in time. To repeat the epigraph, ‘‘Every story is a travel story
– a spatial practice’’ (Certeau 1984: 115). For him, ‘‘narrative structures have the
status of spatial syntaxes.’’ Like buses and trains, ‘‘every day, they traverse and organize
places; they select and link them together; they make sentences and itineraries out of
them. They are spatial trajectories’’ (Certeau 1984: 115). For Franco Moretti in Atlas
of the European Novel, the spatial trajectories of narrative not only establish linkages,
but also actively enable narrative. For him, space is not incidental to narrativity, but
rather generative of it.4 ‘‘Geography,’’ he writes, ‘‘is not an inert container, is not a box
where cultural history ‘happens,’ but an active force that pervades the literary field
and shapes it in depth’’ (Moretti 1998: 3). Mapping the locations of novels, he
explains, helps to identify what has been submerged in literary studies, how space
‘‘gives rise to a story, a plot’’ (Moretti 1998: 7).
Space restored to its full partnership with time as a generative force for narrative
allows for reading strategies focused on the dialogic interplay of space and time as
mediating coconstituents of human thought and experience. In this sense, space is not
passive, static, or empty; it is not, as it is in so much narrative theory, the (back)ground upon which events unfold in time. Instead, in tune with current geographical
theories about space as socially constructed sites that are produced in history and
change over time, the concept of narrative as a spatial trajectory posits space as active,
mobile, and ‘‘full.’’5 This is in part what Lawrence Grossberg calls for in positing a
‘‘spatial materialism’’: ‘‘space as the milieu of becoming’’ allows for understanding
reality as a ‘‘question not of history but of orientations, directions, entries and exits. It
is a matter of a geography of becomings . . . ; it refuses, not only to privilege time, but
to separate space and time. It is a matter of the timing of space and the spacing of
time’’ (Grossberg 1996: 179–80). For Grossberg, the separation of space and time is
itself an illusory construct of human thought, one that results from the privileging of
the temporal. Perhaps a debatable assertion, this view nonetheless suggests that a revisioning of space in narrative poetics can lead to a new understanding of how space
and time interact as constitutive components of story, what Grossberg calls ‘‘the
timing of space’’ and the ‘‘spacing of time.’’ It fosters comprehension of the dialogic
interplay of location and action in the topochrone of the narrative.
196
Susan Stanford Friedman
Space within the story told – the space through which characters move and in
which events happen – is often the site of encounter, of border crossings and cultural
mimesis. For Certeau, narrative establishes ‘‘frontiers’’ (that is, borders) that both
mark difference and establish relations across it. ‘‘A narrative activity,’’ he writes,
‘‘continues to develop where frontiers and relations with space abroad are concerned. . . . [Narrative] is continually concerned with marking out boundaries. . . . On the
one hand, the story tirelessly marks out frontiers. It multiplies them’’ (Certeau 1984:
125–6). But on the other hand, the story continually makes ‘‘bridges’’ – it ‘‘welds
together and opposes insularities’’ (Certeau 1984: 128). Narrative is built out of a
contradiction of interactions in space: a ‘‘complex network of differentiation’’ and ‘‘a
combinative system of spaces’’ (Certeau 1984: 126). Certeau likens this spatial
practice to the contact of two bodies: ‘‘Thus, bodies can be distinguished only
where the ‘contacts’ (‘‘touches’’) of amorous or hostile struggles are inscribed on
them. This is a paradox of the frontier: created by contacts, the points of differentiation between two bodies are also their common points. Conjunction and disjunction
are inseparable in them’’ (Certeau 1984: 126–7).
Moretti’s claim about borders is more modest, but nonetheless suggestive. He
theorizes first that the locations of the novel – at least the European novel from 1800–
1900 – are entwined with the national imaginary, with the ‘‘imagined community’’
that was being formed in conjunction with the modern nation-state. ‘‘The novel,’’ he
writes, ‘‘functions as the symbolic form of the national state . . . and it’s a form that
(unlike the anthem, or a monument) not only does not conceal the nation’s internal
divisions, but manages to turn them into a story’’ (Moretti 1998: 20). As the imagined
space of history, the nation is in one sense ‘‘the sum of all its possible stories’’ (Moretti
1998: 20). Borders – both external and internal – play a particularly important role in
the formation of these stories, he argues, particularly in the historical novel. As a
‘‘phenomenology of the border’’ (Moretti 1998: 35), the historical novel narrates the
confluence of space and time by telling stories about ‘‘external frontiers’’ where
opposites or enemies collide, and ‘‘internal borders’’ as sites of treason and rebellion
(Moretti 1998: 35–6).
Although I have learned to be suspicious of claims about ‘‘all narrative,’’ I wish to
hypothesize provocatively that all stories require borders and border crossings, that is,
some form of intercultural contact zones, understanding ‘‘culture’’ in it broadest sense
to incorporate the multiple communal identities to which all individuals belong
(Friedman 1998: 134–40). Isn’t it borders that make movement through space/time
interesting, suspenseful, agonistic, reconciliatory? Borders insist on purity, distinction, difference, but facilitate contamination, mixing, and creolization. Borders of all
kinds are forever being crossed; but the experience of crossing depends upon the
existence of borders in the first place. Borders function symbolically and materially
around the binaries of pure and impure, sameness and difference, inside and outside –
polarities that set in play spatially enacted oscillations, migratory movements back
and forth that I have elsewhere called an intercultural fort/da (Friedman 1998:
151–78). Identity is unthinkable without borders, whether individual or communal.
Spatial Poetics and The God of Small Things
197
Bodies too are border sites, marking the distinction between inside and outside, self
and other. Bodies are a flesh and blood upon which the social order marks its
hierarchies based on boundaried systems of gender, race, ethnicity, class, caste,
religion, sexuality, and so forth. In all these modes and functions, borders are sites
and social locations that generate and shape narrative in conjunction with time.
Spatial Poetics in The God of Small Things
Spaces – particularly border spaces – generate and shape the stories of caste and gender
division as they unfold in the palimpsestic sequence of colonial, postcolonial, and
postmodern time in The God of Small Things. While it ultimately interweaves the
poetics of space and time, the novel’s narrative discourse privileges space over time,
tropes locations as ‘‘figures’’ on the ‘‘ground’’ of time, and thus illustrates more than
many narratives the compensatory emphasis on space that cultural theorists like Soja
and Grossberg have called for. An architect, screenwriter, and political activist,
Arundhati Roy won the Booker Prize for her worldwide best-selling novel. Like
Salmon Rushdie’s Midnight’s Children, Roy’s novel is a political allegory about the
newly formed nation-state of India and the silences it needs to confront in order to
fulfill the promise of freedom. But more than Rushdie, Roy preeminently calls
attention to the borders within the nation-state – power relations of gender and
caste, in particular – as an inseparable element of the postcolonial dilemma. Like a
number of non-Western feminist writers and activists, Roy situates her writing at the
contradictory conjuncture of gender, race/caste/class, and nation. Attacked as a traitor
to the national cause of postcolonial liberation, she nonetheless insists that future
freedoms depend on painful interrogations of social inequities at home as these
mediate and are mediated by national and transnational histories.6
In setting the novel in her home state of Kerala on the southwestern tip of the
subcontinent, Roy features a region of India that is highly anomalous and provides the
novel’s political allegory with a complexly doubled reference, first to Kerala and
then more generally to India. Kerala is the only state ruled for most of the postIndependence period by elected communist officials. Kerala prides itself on its rates of
literacy, network of social welfare programs, strong labor movement, and the status of
its women – all among the highest in India. Less attention has been paid to its
stagnating economy and to its having among the worst records in India for land
reform directed at the Dalits, that is, the unscheduled castes also known as the
‘‘untouchables.’’ Kerala is also distinctive for its relatively large Christian population
– about 20 percent, some of whom descend from the conversions made by British
missionaries in the nineteenth century, but most of whom are upper-caste Syrian
Christians who trace their ancestry back to the conversion of one hundred Brahmins
by St Thomas the Apostle in the first century of the Christian era. As Indians allied
with Eastern Orthodox Christianity, they have traditionally formed the social
and economic elite of Kerala, thus marking the region’s difference from the rest of
198
Susan Stanford Friedman
Hindu-dominated India. Kerala’s status as uncharacteristic anomaly in India’s already
heterogeneous landscape makes it an unlikely setting for political allegory for India as
a whole. Nonetheless, close attention to the novel’s spatial poetics reveals ways in
which Roy uses this locale to address the politics of regional, national, and transnational landscapes through time.
The novel moves back and forth between two brief time periods in the life of three
generations of the Ipe family, members of the Syrian Christian elite: 13 days leading
up to catastrophe in December of 1969 and one day in June of 1992, when the
youngest generation, the twins Rahel and Estha, are reunited for the first time in
23 years. The narration of trauma and its disastrous effects unfolds out of temporal
sequence, in fragments of the story that keep emerging as returns of the repressed.
Spatially oriented memories mark the porous border between the conscious and
unconscious, the remembered and the forgotten, the forbidden and the desired.
At the core of this resisted remembrance of things past lies a series of dystopic and
utopic border crossings that transgress material, social, psychological, sexual, and
spiritual frontiers. It starts when the orangedrink man in the Abhilash Talkies movie
theater forces seven-year-old Estha to rub his sticky penis. Then, the twins’ cousin,
Sophie Mol, just arrived from England and treasured for her white Englishness,
drowns in the dangerous waters of the Meenachel River. Ammu, the twin’s divorced
mother, crosses the river every night to touch and be touched by Velutha, the
charismatic and tender untouchable who had left the area with the new nation’s
abolition of untouchability, returned with an education to be the engineer and
foreman at the Ipe’s Paradise Pickles & Preserves factory, and played the part of the
twins’ surrogate father. Velutha’s father betrays his son to Ammu’s mother and aunt,
Baby Kochamma, who denounce him to the police as a rapist and child kidnapper.
Comrade Pillai, the town’s communist leader, refuses Velutha’s plea for help. The
twins secretly watch the police beat Velutha nearly to death. Baby Kochamma locks
her niece Ammu in her room and manipulates Estha into supporting her false
accusation to the police to save his mother. Velutha lives just long enough to hear
Estha’s betrayal, but not long enough to hear Ammu’s affirmation of her love to the
police. Ammu’s brother Chacko banishes her from the Ipe home, and she later dies,
unable to support herself or reunite her family. Estha is returned to his dissolute father
in Calcutta, and Rahel is sent to boarding school, never to see her twin again until
1992.
The events of 1969 are 13 days of doom that leave the children emotionally frozen
in time. Estha hardly speaks again, obedient and servile, performing a kind of
exaggerated femininity as penance in contrast to the loud, unruly, and rebellious
Rahel, who wanders the globe, finally to return home when her twin is ‘‘re-returned’’
to Kerala with his father’s migration to Australia. The day of remembrance in 1992
ends in the incestuous embrace of the twins, the connection of souls figured in the
anguish of forbidden touch.
The story is convoluted, melodramatic, sensationalist – borrowing from the conventions of ‘‘Bollywood,’’ India’s hugely profitable film industry in Bombay. But the
Spatial Poetics and The God of Small Things
199
events in The God of Small Things are not narrated in the kind of chronological
sequence I have provided. Exhibiting its own form of the narratological distinction
between fabula and sjuzhet (Propp 1968), ‘‘story’’ and ‘‘narrative’’ (Genette 1980), or
‘‘story’’ and ‘‘discourse’’ (Chatman 1978), the novel refuses the foregrounding of
temporality that my summary embodies and demonstrates instead an emphasis on
narrative as a spatial practice. Rather than history containing space, different spaces in
the novel contain history. The novel moves associationally in and out of these spaces,
rather than sequentially in linear time, with each location stimulating different
fragments of events, often lyrically rendered through motif, metonym, or image.
Each of these spaces contains multiple borders of desirous and murderous connection
and separation, borders that are continually erected and transgressed in movement
that constitutes the kinetic drive of the plot.
Reflecting, no doubt, Roy’s profession as an architect, each space is troped in a
building that is a charged site of historical overdetermination. These buildings are
more than settings or backgrounds. Instead, they embody the social locations that
concretize the forces of history. They are places that palimpsestically inscribe the
social order as it has changed over time. Containing history, they set in motion the
identities of the people who move through them. No mere backgrounds for plot, they
embody narrativity. They make things happen. They contain the borders that juxtapose sharp edges of different cultures – the communal differences of nation, gender,
caste, class, religion, sexuality. They trope contact zones where these differences clash
and blend, where borders get crossed.
Foucault would call these spaces of narrativity ‘‘heterotopic.’’ His neologism
appears in ‘‘Of Other Spaces,’’ where he defines heterotopias as ‘‘real places’’ that
bring into focus the interrelationship of other spaces and ‘‘slices of time’’ as structures
of the social order (Foucault 1986: 24–7). Cemeteries, prisons, theaters, brothels,
museums, libraries, fairgrounds – these are some of his examples of heterotopic sites
that relate to larger cultural structures of crisis, deviation, incompatibility, juxtaposition, compensation, or continuity. What I take from his term is the notion that
certain real spaces contain by inference other spaces and the spatial trajectories of the
disciplinary social order. I would add to his notion a psychoanalytic concept of spatial
cathexis, the powerfully magnetic drawing into a given site other geographical and
geopolitical formations. In The God of Small Things, various buildings function
metonymically as heterotopic places that bring into focus the social, cultural, and
political systems that form identities; set in motion the transgression of borders; and,
in effect, generate the story, the unfolding of events which cannot, because of their
anguish, be told in sequence and can only be apprehended in fragments attached to
specific locations.
The Abhilash Talkies movie theatre, for example, is just such a heterotopia that
contains the history of colonialism, postcolonialism, and the growing American
cultural and economic hegemony. The faded opulence of the movie palace built in
the provincial capital of Cochin after the advent of the ‘‘talkies’’ recalls the period just
before the demise of the Raj. The Ipe family journeys from Ayemenem to Cochin in
200
Susan Stanford Friedman
1969 to see The Sound of Music, the Hollywood movie about the escape of the Von
Trapp family from the clutches of the Nazis. The ‘‘clean white children’’ in The Sound
of Music make Estha and Rahel feel irreparably dirty and naughty as they nervously
anticipate meeting their half-English cousin for the first time. It is a moment of
postcolonial angst complicated by the conflation of the British Raj with America’s
growing postwar dominance (Roy 1997: 101). But the (post)colonial plot generated
by Abhilash Talkies turns ‘‘domestic,’’ when the orangedrink man violates Estha’s
body border at least in part out of class and caste resentment of the borders between
himself and the privileged boy who have been brought together in the space of the
Abhilash Talkies theater. To echo Bakhtin (1981: 84), ‘‘Time, as it were, thickens,
takes on flesh, becomes aesthetically visible’’ in the ornate building that both contains
many stories of the past and generates stories of the future, specifically the first
transgression in a chain of events leading to disaster.
Many other buildings function heterotopically as generators of story in The God of
Small Things. Ayemenem House is the ancestral home of the Ipe family, where the
mixture of Anglophilia and Anglophobia passes down through the generations and
ends in decay by 1992. Their nearby factory, Paradise Pickles & Preserves, is another,
as a hybrid masala of Mammachi’s local business acumen and the Western practices
that Chacko introduces and that eventually ruin the factory. The factory’s spatial
trajectory incorporates Velutha’s rise as foreman and the opposition from lower-caste
workers to this once-unthinkable position of authority for an untouchable. As a
contact zone, the factory brings together not only East and West, touchables and
untouchables, but also manager and communist leader; for Comrade Pillai, who wants
both the votes of the lower-caste workers and the money of the factory’s owner, needs
to maintain his contract with Chacko to print the factory’s advertising fliers. The
corruption embedded in the factory as heterotopic space is a critical component of
Comrade Pillai’s betrayal of Velutha.
Velutha’s family home, a squatter hut built long ago on the grounds of the
Ayemenem House, is still another heterotope, highlighting the patronage relationship
that Velutha transgresses in touching Ammu and that his father fulfills in betraying his
son. The Kottyam police station embodies the power relations of the state, guards the
borders of caste and gender, and empowers the ‘‘touchable’’ police to violate the bodies
of the dying Velutha and the desperate Ammu, who has come too late to save him with
her public avowal of her affair and must withstand Inspector Thomas Mathew’s
lecherous gaze and baton as he taps her breasts. In 1992, the nouveau riche house of
Comrade Pillai contains the history of his rise in the new India, a success built on his
alliance with the lower-caste workers against the untouchables. The temple is the
building in front of which Kathakali dancers perform full sections of the Mahabharata
after doing ‘‘truncated swimming-pool performances,’’ packaged for tourists at the
Heritage Hotel in 1992. And so forth.
The buildings of the novel are crucial to the topochronic dimension of the narrative
because they bring into focus the social structures that built them, the history of their
change over time, and the identities of the people whose lives are inseparable from the
Spatial Poetics and The God of Small Things
201
structures into which they are born. The buildings further bring into focus Roy’s
complex politics – on the one hand, unveiling the geopolitical structures of colonialism, postcolonialism, and multinationalism; on the other hand, turning a critical
searchlight onto India’s internal affairs. As a site of theatrical spectacle for the ‘‘true’’
Mahabharata, for example, the temple opposes the spatial trajectories of both the
Abhilash Talkies movie palace and the stage at the Heritage Hotel. But the temple
also enables the stratification of Hindu society. As members of an unscheduled caste,
the Kathakali dancers are not allowed inside the temple, and the epic they perform
represents an India destroyed by internal family feuding. Where the novel’s buildings
contain distinct stories of multiple borders, the Meenachal River meanders in between
them as a figural representation of the dangers and allure of fluidity, of transition and
transgression. It is the river that gets crossed and recrossed on the days leading up to
the day of everyone’s doom, the day that changes life forever.
More than all the other heterotopic buildings, the Akkara house is the place that
serves as ‘‘an active force’’ that ‘‘gives rise to a story, a plot,’’ to cite Moretti again
(Moretti 1998: 7). Troped repeatedly by the twins and the narrator as the ‘‘History
House,’’ Akkara sits at the center of a rubber plantation around the bend and down
river from Ayemenem House. Invoking India’s colonial past, Akkara had belonged to
Kari Saipu: ‘‘The Black Sahib. The Englishman who has ‘gone native.’ Who spoke
Malayalam and wore mundus. Ayemenem’s own Kurtz. Ayemenem his private Heart
of Darkness. He had shot himself through the head ten years ago, when his young
lover’s parents had taken the boy away from him’’ (Roy 1997: 51). Akkara is Roy’s
indigenized answer to Conrad’s Heart of Darkness, one that attacks the colonialism he
reflected in spite of the anticolonial dimensions of his novel, but also one that borrows
his project of critique to point a finger at the politics at home. The History House is a
metonym for what Benjamin regards as the basis of narrative poetics: ‘‘the most
intimate interpenetration of these two archaic types’’ – the intertwined stories of
home and travel (Benjamin 1969: 85). The structure of the house mines the structures
of history, embodying in a spatial figure what Grossberg calls ‘‘the timing of space and
the spacing of time’’ (1996: 179–80).
Historical change over time is marked on the face of the History House. In the
postcolonial India of 1969, Akkara is empty: the Englishman’s Heart of Darkness has
become India’s own. For Chacko – Oxford-educated, married briefly to a white
woman, back in Kerala, and unlike his sister Ammu, invited by his mother to run
the factory and have sex with lower-caste women – India is doomed to postcolonial
angst, loving and hating its former rulers. ‘‘History,’’ he tells the twins, ‘‘was like an
old house at night. With all the lamps lit. And ancestors whispering inside’’ (Roy
1997: 51). The problem with India, he expounded to them in his spatialized allegory
of British colonial rule and the lingering effects of Anglophilia, is that:
we can’t go in . . . because we’ve been locked out. And when we look in through the
windows, all we see are shadows. And when we try and listen, all we hear is a
whispering. And we cannot understand the whispering, because our minds have been
202
Susan Stanford Friedman
invaded by a war. A war that we have won and lost. The very worst sort of war. A war
that captures dreams and re-dreams them. A war that has made us adore our conquerors
and despise ourselves.
(Roy 1997: 52)
What the children, and by extension the readers, learn is that colonialism cannot
explain all of India’s ‘‘darkness.’’ For, in spite of Chacko’s insistence, the children do,
in some sense, enter the History House, at least as far as the verandah, as least far
enough to see unfold before them ‘‘History in live performance’’ (Roy 1997: 293).
Here, they watch as the ‘‘Dark of Heartness tiptoed into the Heart of Darkness’’ (Roy
1997: 290). Here, Velutha and Ammu broke the Love Laws against touch.
Here, Velutha retreats after the Ipe family and Comrade Pillai betray him. Here,
the children watch as the ‘‘posse of Touchable Policemen acted with economy, not
frenzy. Efficiency, not anarchy. Responsibility, not hysteria. They didn’t tear out
his hair or burn him alive. They didn’t hack off his genitals and stuff them in his
mouth. They didn’t rape him. Or behead him. After all they were not battling an
epidemic. They were merely inoculating a community against an outbreak’’ (Roy
1997: 293). Velutha the untouchable has polluted an upper-caste woman through his
touch, and the police, those ‘‘patrolling the Blurry End,’’ must contain his desire by
methodically beating him to the borderlands between life and death, in effect
condemning Ammu to exile and slow death, a modern immolation of the woman
with the ‘‘Unsafe Edge’’ (Roy 1997: 5, 44). The History House is, in Certeau’s terms,
the ‘‘frontier’’ where ‘‘the ‘contacts’ (‘‘touches’’) of amorous or hostile struggles are
inscribed’’ on the walls and bodies that move through it (Certeau 1984: 126). For
Roy, the body is the initial site upon which the spatial meanings of the social order are
written. The body is the primal border, the first space of passionate connection and
violent disconnection.
By 1992, when Rahel returns to the History House where the caste laws had
polluted her life ever after, she finds the Heritage Hotel, reborn through an infusion
of multinational millions into a playground of the past: ‘‘Toy Histories for rich
tourists to play in’’ (Roy 1997: 120). ‘‘So there it was,’’ the adult Rahel realizes,
‘‘History and Literature enlisted by commerce. Kurtz and Karl Marx joining palms to
greet rich guests as they stepped off the boat’’ (Roy 1997: 120). While the Ayemenem
House ‘‘on the Other side’’ decays, the History House morphs into a postmodern
space, burying in its playful facade the sedimented layers of desire and trauma that
characterized the erection and dissolution of colonial, postcolonial, caste, and sexual
boundaries. Postmodernity, the age of representational play and transnational corporations, has supplanted the (post)colonial age of modernity metonymically present in
the novel through the allusions to Conrad’s Heart of Darkness, Forster’s A Passage to
India, and Hollywood’s The Sound of Music.
As much as colonial domination, postcolonial angst, and transnational corporate
tourism are central to the History House of Roy’s border narrative, they constitute
only a part of the Heart of Darkness. And, for someone like Chacko, and the Indians
Spatial Poetics and The God of Small Things
203
whom he represents in contemporary India and its diaspora, the power exerted over the
nation by outside forces becomes all too easy an excuse for suppressing the history of
gender and caste inequities that preexisted the arrival of the Muslims and the British
and have continued to evolve in the complex contact zones of a heterogeneous
subcontinent and through interaction with colonial, postcolonial, and transnational
historical formations. As a political allegory, the novel asks that attention be paid as
well to the ‘‘Dark of Heartness,’’ that is, not only to the legacy of Western imperialism
and late twentieth-century transnationalism, but to the dark borders erected in the
‘‘heart’’ of India – the laws of love and touch that regulate relations between genders
and castes within the nation. In Moretti’s terms, The God of Small Things ‘‘functions as
the symbolic form of the nation state . . . [,] a form that . . . not only does not conceal
the nation’s internal divisions, but manages to turn them into a story’’ (Moretti 1998: 20).7
Roy tells her story through buildings, through spatial entities that heterotopically
draw within their walls the geopolitical and domestic structures that have taken shape
through time. As Bakhtin writes about space in the novel, these buildings become
‘‘charged and responsive to the movements of time, plot and history’’ (Bakhtin 1981:
84). In Roy’s narrative, space and time exist in Einsteinian relativity, each understood
only in relation to the other, its discursive forms fulfilling Grossberg’s call for the
‘‘timing of space’’ and the ‘‘spacing of time.’’ At the same time, the discursive form of
Roy’s narrative foregrounds spatiality. The narrative’s structural reliance on buildings
to move the narrative forward gives a compensatory emphasis to space over time as
constitutive of narrative discourse. A tale of many borders, The God of Small Things
narrativizes story as a spatial practice, one that doesn’t erase time, but rather constitutes space as the container of history and the generator of story. Roy narrativized
what Certeau theorizes when he writes: ‘‘On the one hand, the story tirelessly marks
out frontiers. It multiplies them’’ (Certeau 1984: 126). Such borders, frontiers, are not
the background of narrative, mere description where time unfolds its plot. They are,
instead, the generative energy of narrative, the space that contains time.
NOTES
1 See also Ricoeur’s Narrative Time; Chatman’s
‘‘Story: Events’’ in Story and Discourse for his
discussion of sequence, causality, and time
(1978: 45–95); Scholes and Kellog’s The Nature of Narrative, in which they argue that
‘‘Plot can be defined as the dynamic, sequential element in narrative literature’’ (1966:
207) and ignore the spatial element of narrative entirely; and Martin’s Recent Theories of
Narrative, where he rejects the notion of setting, scene, or description as merely static, but
follows Genette in emphasizing the modes of
narrative temporality (1986: 122–5, 229–30).
Chambers’ Story and Situation and Room
for Maneuver, as their titles suggest, pay considerable attention to the context of narration
– the narrator as the teller to an audience; but
he doesn’t directly address the function of
space within the story told. Joseph Frank’s
([1945] 1963) essay, ‘‘Spatial Form in Modern
Literature,’’ which spawned decades of debate,
deals only with the formal techniques developed to interrupt chronological sequencing
and give the illusion of simultaneity; see also
204
2
3
4
5
Susan Stanford Friedman
Frank’s The Widening Gyre and Frank (1978)
and Smitten and Daghistany (1983). In
Mappings, I discuss the predominance to
time in narrative studies (Friedman 1998:
137–40).
Ronen responds in part to the special issue of
Yale French Studies, ‘‘Towards a Theory of
Description’’ (Kittay 1981), importantly challenges the ‘‘almost canonical description of
description as non-narrative’’ (Ronen 1997:
275), and argues that description is a part of
narrative, not its opposite. But, like the articles in Yale French Studies, she does not discuss
description as spatiality.
A recent thread on the Narrative listserv is
more spatially oriented, answering Christine
Junker’s inquiry about ‘‘narratives of place and
the relationship between narrative and place’’
(October 13, 2003); various answers from Jim
Holm, George Perkins, Alison Booth, Rick
Livingston, and others continue to assume
that place is the backdrop of plot, not a part
of it.
In Mappings (Friedman 1998), published in
the same year as Moretti’s Atlas, I made a
similar point; see chapters 5 and 6 for my
earlier efforts to shift the emphasis in narrative
poetics from time to space. For a discussion of
spatialization as a reading practice, see my
‘‘Spatialization’’ (Friedman 1993) which does
not take up the question of how the space that
characters inhabit and move through works in
the narrative.
In addition to Soja, see for example Lefebvre
(1991), Massey (1994), Keith and Pile, and
Bourdieu’s concept of ‘‘habitus’’ (esp. 1990:
REFERENCES
AND
Abbott, H. P. (2002). The Cambridge Introduction
to Narrative. Cambridge, UK: Cambridge University Press.
Ahmad, A. (1987). ‘‘Jameson’s Rhetoric of Otherness and the ‘National Allegory’.’’ Social Text 17
(Autumn), 3–25.
Bakhtin, M. M. (1981). The Dialogic Imagination:
Four Essays, ed. M. Holquist, trans. C. Emerson
and M. Holquist. Austin: University of Texas
Press.
52–65). In this essay, I make no distinction
between ‘‘space’’ and ‘‘place,’’ primarily because such distinctions (rampant in geography
and social theory) vary considerably and are
often contradictory. For some different formulations of space/place, see Tuan and Hoelscher
(2001), Sack (1997), and Certeau (1984:
117–18). My concern is spatiality as an aspect
of human cognition and as a component part
of narrative discourse, both of which incorporate space as place or location.
6 For fuller discussion of the production and
reception of the novel, its status as political
allegory, its relation to feminism, and its
setting in Kerala and the debates about the
‘‘Kerela model’’ of development, see my
‘‘Feminism, State Fictions, and Violence’’
(Friedman 2001). For other discussions of
Roy’s novel, see Dhawan (1999).
7 In reading The God of Small Things as a political
allegory of the nation state, I do so outside the
framework of Fredric Jameson’s (1986) muchcritiqued ‘‘Third-World Literature in the Era of
Multinational Capitalism,’’ where he hypothesizes preposterously (in my view) that ‘‘all
third-world cultural productions seem to have
in common’’ a quality that ‘‘distinguishes them
radically from analogous cultural forms in the
first world’’ – namely, that ‘‘all third-world
texts are necessarily . . . allegorical, and in a
very specific way’’ – as ‘‘national allegories’’
( Jameson 1986: 60). I would suggest in contrast that Roy’s novel is a national allegory in
much the same way as Heart of Darkness and A
Passage to India. See Ahmad’s stinging critique
of Jameson’s argument (1987: 3–25).
FURTHER READING
Benjamin, W. ([1936] 1969). ‘‘The Storyteller.’’
In Illuminations: Essays and Reflections, ed.
H. Arendt, trans. H. Zohn (pp. 83–110). New
York: Schocken Books.
Brooks, P. (1984). Reading for the Plot:
Design and Intention in Narrative. New York:
Vintage.
Bourdieu, P. ([1980] 1990). The Logic of Practice,
trans. R. Nice. Stanford, CA: Stanford University Press.
Spatial Poetics and The God of Small Things
Certeau. M. de. ([1974] 1984). ‘‘Spatial Stories.’’
In The Practice of Everyday Life, trans. S. Rendell
(pp. 115–30). Berkeley: University of California
Press.
Chambers, R. (1984). Story and Situation: Narrative
Seduction and the Power of Fiction. Minneapolis:
University of Minnesota Press.
Chambers, R. (1991). Room for Maneuver: Reading
Oppositional Narrative. Chicago: University of
Chicago Press.
Chatman, S. (1978). Story and Discourse: Narrative
Structure in Fiction and Film. Ithaca, NY: Cornell
University Press.
Clifford, J. ([1992] 1997). ‘‘Traveling Cultures.’’
In Routes: Travel and Translation in the Late
Twentieth Century (pp. 17–46). Cambridge,
MA: Harvard University Press.
Dhawan. R. K. (ed.) (1999). Arundhati Roy: The
Novelist Extraordinary. New Delhi: Prestige Books.
Foucault, M. ([1984] 1986). ‘‘Of Other Spaces,’’
trans. Jay Miskowiec. Diacritics, 16(1), 22–7.
Frank, J. ([1945] 1963). ‘‘Spatial Form in Modern
Literature.’’ In The Widening Gyre: Crisis and
Mastery in Modern Literature (pp. 3–62). New
Brunswick, NJ: Rutgers University Press.
Frank, J. (1978). ‘‘Spatial Form: Some Further
Reflections.’’ Critical Inquiry 5, 275–90.
Friedman, S. S. (1993). ‘‘Spatialization: A Strategy
For Reading Narrative.’’ Narrative 1 ( January),
12–23.
Friedman, S. S. (1998). Mappings: Feminism and the
Cultural Geographies of Encounter. Princeton, NJ:
Princeton University Press.
Friedman, S. S. (2001). ‘‘Feminism, State Fictions,
and Violence: Gender, Geopolitics, and Transnationalism.’’ Communal/Plural 9(1), 112–29.
Genette, G. ([1972] 1980). Narrative Discourse: An
Essay in Method, trans. J. E. Lewin. Ithaca, NY:
Cornell University Press.
Grossberg, L. (1996). ‘‘The Space of Culture, the
Power of Space.’’ In I. Chambers and L. Curti
(eds.), The Post-Colonial Question: Common
Skies, Divided Horizons (pp. 169–88). London:
Routledge.
205
Jameson, F. (1986). ‘‘Third-World Literature in
the Era of Multinational Capitalism.’’ Social
Text, 15 (Autumn), 65–88.
Keith, M. and Pile, S. (1993). Place and the Politics
of Identity. London: Routledge.
Kittay, J. (ed.) (1981) ‘‘Towards a Theory of
Description.’’ Special Issue, Yale French Studies
61.
Lefebvre, H. ([1974] 1991). The Social Production of
Space, trans. D. Nicholson-Smith. Oxford: Basil
Blackwell.
Martin, W. (1986). Recent Theories of Narrative.
Ithaca, NY: Cornell University Press.
Massey, D. (1994). Space, Place, and Gender. Minneapolis: University of Minnesota Press.
Moretti, F. (1998). Atlas of the European Novel,
1800–1900. London: Verso.
Propp, V. (1968). The Morphology of the Folktale,
2nd edn., trans. L. Scott. Austin: University of
Texas Press.
Ricoeur, P. (1981). Narrative time. In W. J. T.
Mitchell (ed.), On Narrative (pp. 165–86). Chicago: University of Chicago Press.
Ricoeur, P. (1983, 1985). Time and Narrative,
2 vols. Chicago: University of Chicago Press.
Ronen, Ruth. (1997). ‘‘Description, Narrative and
Representation.’’ Narrative 5(3), 274–86.
Roy, Arundhati. (1997). The God of Small Things.
New York: Random House.
Ryan, M.-L. (1997). Posting on Narrative Listserv.
December 4.
Sack, R. (1997). Homogeographicus: A Framework for
Action, Awareness, and Moral Concern. Baltimore,
MD: Johns Hopkins University Press.
Scholes, R. and Kellogg, R. (1966). The Nature of
Narrative. Oxford: Oxford University Press.
Smitten, J. R. and Daghistany, A. (eds.) (1983).
Spatial Form in Narrative. Ithaca, NY: Cornell
University Press.
Soja, E. (1989). Postmodern Geographies: The Reassertion of Space in Critical Theory. London: Verso.
Tuan, Y.-F. and Hoelscher, S. (2001). Space and
Place: The Perspective of Experience, 2nd edn. Minneapolis: University of Minnesota Press.
13
The ‘‘I’’ of the Beholder:
Equivocal Attachments and the
Limits of Structuralist
Narratology
Susan S. Lanser
A piece by Ann Beattie in the November 5, 2001, New Yorker begins like this: ‘‘True
story: my father died in a hospice on Christmas Day’’ (Beattie 2001: 84). The ‘‘I’’ who
tells this story is a writer of fiction named Ann. Is this narrator then Ann Beattie or
some other Ann, and am I reading a short story that’s playing with conventions or a
piece of autobiography? When Ann tells us she is superstitious, is Beattie confessing
her frailties? When Ann gives us her views about writing, are these also Beattie’s
views? Did Beattie’s own father die on Christmas Day, or die at all?
As the history of literary reception has made dramatically evident, there is simply
no way to resolve these questions from the text ‘‘itself.’’ Because autobiography and
homodiegetic fiction deploy the same range of linguistic practices, the ontological
status of most I-narratives cannot be proved by citing any part or even the whole of
the text. Techniques of omniscience and focalization conventionally signal the fictivity of a heterodiegetic (third-person) narrative,1 but homodiegetic texts provide no
such formal index of their status as fiction or fact. Indeed, writers and publishers have
exploited this ontological ambiguity for centuries, passing off fictions for histories
and hiding autobiographies beneath claims of fictitiousness; the ‘‘rise’’ of the novel is
indebted to just such practices. Nor is modern literature exempt: Maxine Hong
Kingston, for example, has said that she wrote her acclaimed The Woman Warrior:
Memoirs of a Girlhood Among Ghosts as a novel but let her editor market it as
autobiography. That the book has continued to be read as the true story of Kingston’s
childhood not only reveals the conventional rather than essential distinction between
fiction and autobiography, but casts both the historicity and the fictivity of The
Woman Warrior into doubt.
Most of the time, the generic identification of I-narrative hinges on the single and
frail signifier of the proper name – that is, upon the narrator’s ‘‘identity or nonidentity with the author in whose name the narrative has been published’’ (Cohn
The ‘‘I’’ of the Beholder
207
1999: 31–2). But if I want to distinguish between the author Ann Beattie and the
narrator-character Ann, I am quite dependent on someone else’s word for it. That
word usually comes in the rather flimsy and accidental form of a narrative’s placement
within categorical space: the ‘‘fiction’’ or ‘‘nonfiction’’ shelves in a bookstore, the blurb
on a book cover, the information in the library catalog, the table of contents of a
magazine. I myself would have taken Beattie’s story for memoir had not the New
Yorker placed the word ‘‘fiction’’ on its first page.
The question should be settled, then: the announcement of ‘‘fiction’’ opens a chasm
between the life and opinions of the author ‘‘Ann Beattie’’ and the life and opinions of
the character ‘‘Ann.’’ For, as all good students of narrative have learned, the cardinal
rule for reading fiction is that the narrating ‘‘I’’ is not the author, and thus the claims
made by the narrating ‘‘I’’ are not to be taken for the author’s claims. Structurally
speaking, the ‘‘I’’ who speaks in a fiction is simply and wholly a fictional ‘‘I’’: that is
indeed, as narrative theorists from Wolfgang Kayser to Dorrit Cohn have argued,
what makes fiction fiction. Fiction defines itself by the pretended quality of its speech
acts: thus Gérard Genette argues that even when names and histories are parallel, as in
a novel such as A la recherche du temps perdu, ‘‘no speech acts belong to Marcel Proust,
for the good reason that Marcel Proust never takes the floor . . . no matter how the
narrative content may happen to relate to the biography, the life and opinions, of its
author.’’ Genette concludes, therefore, that ‘‘we are just as entitled to set aside the
discourse of first-person fictional narrative as to set aside that of fictional characters
themselves’’ (Genette 1993: 34). Under these strictures, Ann is not Ann Beattie, and
that is that.
Or so it seems. I will argue, however, that the impossibility of linguistic distinction opens a seductive bridge across the structural chasm between a first-person
narrator-character and the author of the text. Against the demurrals of authors and
the onto-logic of narratologists, although we are ‘‘entitled’’ to take fiction as nothing
but fiction, I-narrative taunts us with the possibility that the ‘‘I’’ of the fiction has
some relation to the author’s ‘‘I’’ even when the I-character is not also a writer or does
not share the author’s first name. Dorrit Cohn says as much when she recognizes that
the readerly inclination to blur the lines between I-narrator and author ‘‘must be
acknowledged, even if it is nowadays severely frowned upon.’’ But for Cohn, the
essential chasm remains: while in theory some ‘‘indeterminate’’ fiction ‘‘condemns us
to vacillate, or allows us to oscillate, between referential and fictional readings’’ (Cohn
1999: 34), in the end, ambiguous I-narratives require us to read referentially or
fictionally, and not both at once.
I will take a contrary stance: that readers routinely ‘‘vacillate’’ and ‘‘oscillate’’ and
even double the speaking voice against the logic of both structure and stricture. I want
to ask, therefore, not whether unauthorized crossings happen, but under what circumstances they are likely to occur. Since homodiegetic texts cannot effectively demonstrate their own status as fiction or history, exploring this question from within a
structuralist narratology keeps us immured in precisely the rules of separation that
I believe have failed to address the ways in which readers actually apprehend
208
Susan S. Lanser
first-person texts. I want, rather, to proceed pragmatically by mapping a larger field of
I–I relationships that I think are available to texts, and by locating fiction in general,
and homodiegetic fiction in particular, upon that broader field. Such a project reveals
some of the complicated and transgressive ways in which the ‘‘I’’ of the beholder – the
‘‘I’’ a reader constructs according to what we might call his or her intelligence, in
every sense of that term – is not always the singular ‘‘I’’ of a fictional speaker but the
‘‘I’’ of an author as well. Understanding this equivocal I–I relationship may provide
new insights into the project of literature in general and narrative fiction in particular.
Attachments and Detachments
My map divides all discourse into three parts, each inhabited by a cluster of genres
that we usually separate, and each representing a different set of conventions for
apprehending a text’s enunciating ‘‘I.’’ The first encompasses works as diverse as Mary
Karr’s The Liars’ Club, Foucault’s Discipline and Punish, a letter to the editor of the New
York Times, any history of the French Revolution, and the essay you are reading now.
All these texts evoke an ‘‘I’’ who is conventionally assumed to be the author of the
work. Foucault does not have to utter the first person pronoun for Discipline and
Punish to function as his words. The acclaim given to the painful and powerful Liars’
Club is rooted in the belief that the ‘‘I’’ is recounting a very close approximation of
Mary Karr’s own history: Karr’s thanks to her mother for ‘‘bravery’’ and ‘‘support,’’ the
photographs that divide the book’s three sections, and the thematization of truth and
lying that generate the book’s title, all lend authentication to what would become an
outrage if discovered untrue. If my colleague writes a letter to the New York Times
opposing a presidential policy, I will be startled and perhaps resentful if she tells me
the letter does not really represent her views. While histories of the French Revolution will vary in their selection, narration, and interpretation of events, I read them
with the assumption that to the best of her or his knowledge, each author has
attempted a factual account, compromised perhaps by errors but not by lies. By the
same generic compact, you have a right to assume that the person named as the author
of this essay is indeed its author and that at least while I am writing, I believe what
these pages assert.
I thus call works in these genres attached or contingent: they depend for their
significance – that is, for their value and arguably also for their meaning – on the
equation of the work’s presumptive author with the text’s primary ‘‘I.’’ Readers may
deconstruct this ‘‘I,’’ may argue that it utters meanings the author did not ‘‘intend,’’
may challenge its assertions, may charge it with various misprisions – for every ‘‘I’’ is
ultimately the beholder’s ‘‘I,’’ by which I mean the ‘‘I’’ that the reader constructs in the
process of reading. Readers may of course question the attribution of a text to a certain
author, contest the legitimacy of a particular edition of a text, seek out the truth
behind a pseudonym. If the text presents itself as its author’s own story, I may even
allow for the ways in which the narrative fashioning of a life stretches or complicates
The ‘‘I’’ of the Beholder
209
the autobiographical compact whereby, to use Philippe Lejeune’s (1980) formulation,
author ¼ narrator ¼ character. But in practice, I suggest, not even the most
postmodern reader treats a work of history or philosophy or scholarship, a letter to
the editor, or an autobiography as if the author in the text were not also the author of
the text. That may be why we so rarely use the term ‘‘implied author’’ for attached texts
(with the singular exception of ‘‘creative nonfiction’’); for example, I’ve not yet heard
anyone allude to the ‘‘implied author’’ of Discipline and Punish rather than alluding
simply to Foucault. It might of course be quite useful to speak of a Foucauldian
‘‘implied author’’ as a way of understanding the complicated relationship of writing
persons to written words. But readers do not typically make this distinction between
author and implied author of attached texts.
The effects can be explosive when the authorial and textual ‘‘I’’ do not cohere in this
expected way. In 1983, a work called Famous All Over Town by ‘‘Danny Santiago’’ was
hailed as the first ‘‘authentic’’ account of a Mexican-American adolescent’s life in the
Los Angeles barrio. The book was later revealed to be a pseudonymous novel written
by an elderly white social worker and Yale graduate named Daniel James. The author’s
achievement underscores the power of writers to inhabit subject positions far different
from their own. But the readers’ outrage underscores the investment in the compact
of contingency that relies on the connection between presumed author and textual
‘‘I.’’ An even more infamous case is The Education of Little Tree (Carter 1976), touted as
the ‘‘memoir of a Cherokee boyhood’’ and later identified as a novel by a former Klan
leader and white supremacist. Henry Louis Gates’s argument that all writers are
‘‘cultural impersonators’’ (Gates 1991: 26) did not stem the tide of protest or reevaluation in light of the text’s ‘‘new’’ authorship. In a more complicated vein, the
1996 publishing events known as the ‘‘Alan Sokal/Social Text Affair’’ worked similarly
on an outraged academy: when physicist Sokal submitted a bogus essay on the
‘‘hermeneutics of quantum gravity’’ and Social Text published it, some scholars reveled
in the alleged exposure of postmodern theory, but others argued that scholarship must
operate on conventions of trust that preclude a parodic distance between the textual
voice and the author-scientist.2
My second group of texts carries no such suppositions. Even more diverse in genre
than the first, it includes ‘‘The Star Spangled Banner,’’ a stop sign, the Apostle’s
Creed, the Bhagavad Gita, a joke about the President, and an advertisement for
AT&T. All of these texts are conventionally read in detachment from and usually in
indifference to their authors’ – though not, of course, their culture’s – identities.
I doubt that more than a handful of people could name the advertising firm, let alone
the writers, who taught the United States to reach out and touch someone, and the
significance of ‘‘The Star-Spangled Banner’’ is rarely dependent on its connection to
Francis Scott Key. The Bhagavad Gita is read and revered despite its unknown
authorship, and most of the believers who recite the Apostle’s Creed probably neither
know nor care who composed the declaration of their faith. When an ‘‘I’’ matters in
such instances, it is not the author’s but the beholder’s or performer’s ‘‘I,’’ the ‘‘I’’ who
sings the anthem, recites the creed, or tells the joke. What I am calling ‘‘detached’’
210
Susan S. Lanser
texts simply don’t derive their meaning from a link between any textual voice and its
authorship. Often these texts are specimens or signifiers of a particular culture, but
their significations are not anchored by any person (other than the reader) who has a
speaking part within the text.
Although I’ve been speaking of texts as attached and detached, the distinction is at
least as much a matter of context as a matter of form. Käte Hamburger reminds us
that generic distinctions are themselves pragmatic ones: thus Novalis’s ‘‘Sacred Songs’’
would be read as lyric ‘‘when we encounter this work in a collection of poems,’’ but as
hymn, with the ‘‘I’’ referring to the collective congregation, ‘‘when we encounter it in
a Protestant hymnal’’ (Hamburger 1973: 242). Different cultures are differently
invested in contingent and detached writings; it is possible, for example, that Finns
hear in ‘‘Finlandia’’ the ‘‘I’’ of Sibelius, himself a cultural icon, in ways that people do
not hear in ‘‘The Star Spangled Banner’’ the ‘‘I’’ of Key. Conversely, if a Democratic
candidate were known to be the author of a joke about a Republican President, the
joke would become an attached text for which its author would doubtless pay.
A detached ‘‘I,’’ moreover, can be merely a convenient fiction referring to ‘‘no one’’
at all: if I join a songfest in a pub and belt out ‘‘I’m in love with you, honey,’’ my ‘‘I’’ is
quite detached from both myself and any author. Robert Griffin (1999) has noted that
anonymous writings pervade our daily lives; what I call detached genres are functionally
even if not technically anonymous because the ‘‘I’’ who wrote them is not connected to
any specific enunciating entity that the text inscribes.
This mapping of texts according to the attachment or detachment of a textual
speaker to the text’s (putative) author offers a pragmatic axis that cuts across conventional generic groupings. In ‘‘Literature as Equipment for Living,’’ Kenneth Burke
reminds us, however, that textual categories are ‘‘neither more nor less ‘intuitive’ than
any. . . classification of social events. Apples can be grouped with bananas as fruits,
and they can be grouped with tennis balls as round.’’ But if ‘‘the range of possible
academic classifications is endless,’’ Burke continues, the most salient classifications
are those that ‘‘apply both to works of art and to social situations outside of art’’
(Burke 1941: 302–3). Recognizing authorial attachment as a significant aspect of any
text’s ‘‘social situation’’ challenges us to return to literature, and particularly to
homodiegetic fiction, as a place or set of places on a larger discursive map.
The Equi-vocation of Literature
Somewhere between attachment and detachment sit the imaginative genres that we
call ‘‘literature’’ in its constitutive rather than evaluative sense, the sense that encompasses those genres, loosely identified in classrooms as poetry, fiction, and drama,
‘‘that impose [themselves] essentially through the imaginary character of [their]
objects’’ (Genette 1993: 21).3 I call these genres equivocal – literally equi-vocal—for
they not only manifest but, I submit, rely for their meaning on complex and ambiguous
relationships between the ‘‘I’’ of the author and any textual voice. The ‘‘I’’ that
The ‘‘I’’ of the Beholder
211
characterizes literary discourse, in other words, is always potentially severed from and
potentially tethered to the author’s ‘‘I.’’ I would go so far as to propose that the
condition of ‘‘literature’’ (again, in the constitutive rather than evaluative sense)
resides precisely in this equivocality – this vocation, as it were, to evoke voices
ambiguously related to their historical origins, and that from this equivocality derive
many of the liberties and liabilities of literature along with some of the interpretive
dilemmas that underlie our scholarly debates.
I want to emphasize, however, that when I speak about the equivocality of
literature I am not concerned with either implied or biographical authorship. The
sense of an authorial presence ‘‘behind’’ the text that Wayne Booth has famously called
the ‘‘implied author’’ (Booth 1961) may be constructed for any text, attached or
detached, although I have already suggested that the need for the term ‘‘implied
author’’ may be diminished for attached texts because ‘‘the’’ author is already being
associated with the text’s enunciating ‘‘I.’’ Nor do I mean here the commonplace
practice of reading texts in relation to the historical persons who wrote them – a
practice that despite postmodern proclamations of ‘‘the death of the author’’ still
dominates literary scholarship (as a glance at the names of academic societies, journals,
and conference panels makes evident). Both the concept of implied author and
the reading of texts as extensions or embodiments of the historical writer entail
attaching to the author the whole of a text. While such critical methods may intensify
the likelihood that readers will also attach authors to specific textual ‘‘I’’s, practices of
authorial reading are not dependent on aligning any textual speaker with the
author’s ‘‘I.’’
My concern here, by contrast, is specifically with the ways in which specific voices
inscribed in literary texts may – at once, or in oscillation – be both attached to and
detached from the author’s ‘‘I.’’ When I speak about the equivocality of literature,
then, I am speaking pragmatically about whether readers link a voice within the text
to the actual author’s voice, even when they ‘‘know’’ that a speaking ‘‘I’’ is fictional and
‘‘ought’’ to be considered purely as such. That is, against the claims of imaginative
literature to be purely imaginative, the ontological status of literary speech acts as
fictions, and the efforts of narrative theorists to enforce boundaries between fiction
and ‘‘reality,’’ under what circumstances might readers nonetheless attach a speaking
voice within a literary text to the (presumptive) author of that text? What textual
signals might lead us to take a formally fictional voice for the author’s against our
instructions to do otherwise?
Terms of Attachment
The likelihood that readers will form some attachments between the ‘‘I’’ of an author
and an ‘‘I’’ in the text seems to me to be in large measure generically contingent in
ways that map surprisingly well onto the broad distinction of lyric, narrative, and
dramatic modes. Strictly speaking, of course, all three of these genres are constructed
212
Susan S. Lanser
on the premise of imaginative content and thus of putative fictivity. But I will claim
that this fictivity does not extend identically for all three genres to the identities and
the assertions of speakers within texts. Not only are the three equivocal genres not
identically equivocal, but within any given literary work, both attachments and
detachments may occur. As a starting point for what must be a much more extensive
project of parsing out the terms of attachments, I want to speculate that readers will
make attachments between an author and a textual ‘‘I’’ according to five criteria which
I will call singularity, anonymity, identity, reliability, and nonnarrativity. Together, these
criteria help to explain why, in normative cases, lyric is the most attached, drama the
most detached, and narrative the most equivocal of the equivocal genres.
By singularity I mean the presence of one rather than many voices at the highest
diegetic level of the text. It is my hypothesis that readers are significantly more likely
to attach an ‘‘I’’ to the author when there is only one ‘‘I’’ to attach and that the
introduction of a second voice at the same diegetic level disturbs this tendency.
Lyric poetry is nearly always characterized by a singularity of voice, while drama –
except in the case of dramatic monologue or performance art – is, conversely,
characterized by a multiplicity of voices at the same diegetic level. Narrative fiction,
of course, lends itself to a great diversity of diegetic modes. ‘‘Third person’’ (heterodiegetic) fiction is likely to deploy a single narrator at the highest diegetic level, but
Dickens’s Bleak House, which places the omniscient narrator and the I-narrator Esther
Summerson at the same diegetic level, is only one of many complicating cases in
point. And homodiegetic fiction may use any number of first-person voices, just as
fully focalized third-person fiction may limit itself to the consciousness of a single
character or shift among characters. Thus narrative fiction occupies a spectrum of
possibilities between the singular ‘‘I’’ of most lyric poetry and the multiple ‘‘I’’ of most
dramatic works.
Anonymity and identity are somewhat interrelated if technically distinct. By anonymity I mean the absence of a proper name for the textual speaker, as is the case with a
surprising number of homodiegetic novels and nearly all lyric poems. I hypothesize
that it is easier to attach an unnamed than a named ‘‘I’’ to its author except perhaps
when, as in Beattie’s story , the ‘‘I’’ shares the author’s name. The term identity, as I use
it here, encompasses all (perceived) social similarities between a narrator and an
author: of name, gender, race, age, biographical background, beliefs and values, or
occupation as writer, and I suggest that textual voices closely allied in identity to that
assumed voice of the author provide readers with encouragement to attach. I would go
farther, and suggest that when a fictional narrator is unnamed, it may take clearly
marked differences in identity to deter readers from imposing the identity of the author
onto an anonymous textual voice, for historically speaking, and perhaps especially in
the case of women writers, even when a textual ‘‘I’’ has differed from the author in
name, the convergence of the author’s known identity with that of the narrating
character has promoted attachment: nineteenth-century critics read Jane Eyre as
Charlotte Bronte’s autobiography just as twentieth-century critics have read The
Bell Jar as the autobiography of Sylvia Plath.
The ‘‘I’’ of the Beholder
213
Identity is one index of a fourth criterion, reliability, by which I mean a reader’s
(complex) determination that the narrator’s values and perceptions are consistent with
those of the author – that is, with the values and perceptions the reader believes the
author holds. I would claim that a reader is very unlikely to attach an author to a
narrator whom that reader has already deemed unreliable. This judgment of reliability
does not require the reader to agree with the author’s values, but it does mean that the
reader thinks the author shares the narrator’s values. If, for example, I consider
Robinson Crusoe’s presumption of cultural and religious superiority to Friday to be
a presumption that Defoe would share, I will take Crusoe as a reliable ‘‘I’’ even if
I myself abhor the views I believe Defoe to be endorsing through Crusoe, and I will
not be deterred from attaching the ‘‘I’’ of Defoe to the ‘‘I’’ of Crusoe. If, on the other
hand, I have reason to think Defoe critical of missionary colonialism, I might question
Crusoe’s reliability and consider a detached ironic relationship between author and
narrator-character. While reliability does not in itself determine attachment, in other
words, unreliability does tend to keep attachment from happening.
I will explore nonnarrativity or atemporality more fully below. Here I’m simply
noting that I mean this criterion to distinguish speech acts that do not recount the
actions, words, or behaviors of characters or events as these unfold in time. I believe
readers are more likely to make attachments between narrators and their author when
the narrating ‘‘I’’ is not reporting or enacting events. I am more likely, in other words,
to believe that Ann Beattie shares the homodiegetic narrator Ann’s views about
writing than to believe that Ann Beattie’s own father died on Christmas Day. I will
suggest, indeed, that equivocal genres are most likely to enact their now-attached and
now-detached character along lines of narrativity. That is, a literary text’s potential for
detachment stems from the pretended nature of some or most – though not necessarily all – of its temporal speech acts, while the same text’s potential for attachment
stems from the potentially nonpretended nature of at least some of its atemporal or
nonnarrative acts, even when these appear within fictional discourse. Indeed, James Phelan
has suggested that what I’m calling attached discourse (and what he calls ‘‘mask
narration’’) might be a particularly persuasive rhetorical device, for by ‘‘draw[ing]
upon the character narrator’s experiences as the ground for the utterance,’’ the author
can ‘‘not only make the direct communication plausible but also make it even more
forceful’’ (Phelan 2005: 201).
These five criteria combine to help explain the very different tendencies of specific
literary genres. Lyric poetry, with its conventional singularity, its commonplace
anonymity, its almost axiomatic reliability, its likelihood of evoking aspects of its
author’s identity, and its relatively low narrativity, is primed for authorial attachment.
Drama is, conversely, primed for detachment: apart from the provider of stage
directions, who ‘‘disappears,’’ as it were, in performance, the highest diegetic level
in a play usually entails a multiplicity of named characters engaged in a high level of
temporal behavior, so that unless criteria of identity are applied, no voice typically calls
out to be connected with an author’s ‘‘I.’’ Epistolary fiction, for example, seems to me
by virtue of its multiplicity of characters on the same diegetic level to resemble drama
214
Susan S. Lanser
in its detachment of all voices from the author’s voice; certainly I have never known a
reader to attach any voice in Les Liaisons dangereuses to the ‘‘I’’ of Laclos. Conversely,
‘‘third person’’ (heterodiegetic) narration by a single and unnamed public voice
encourages attachments similar to, if arguably more fragile than, the attachments
conventional to lyric modes. Thus, even though heterodiegetic narrators are technically fictional, some narrative theorists have acknowledged that readers have very little
incentive to distinguish the narrator of Northanger Abbey from Austen, Le Père Goriot
from Balzac, or Song of Solomon from Morrison. Homodiegetic fiction emerges as the
most equivocal of the equivocal genres, always technically detached and yet sometimes readily attachable.
Attachment and Narrativity
In order to understand the kinds of transgression that I believe are possible in the
reading of homodiegetic fiction, it is useful to begin with the lyric as an extended case
in point, for as Käte Hamburger noted decades ago in The Logic of Literature, ‘‘the
question of the identity or non-identity of the lyric I with the empirical I of the poet’’
has ‘‘not yet been answered by our structural analysis of the lyric statement’’ and
remains ‘‘the topic of much debate’’ (Hamburger 1973: 272).4 Let me take, for
example, John Keats’s ‘‘Bright Star.’’ Whether or not I read this sonnet as expressing
the author’s longing for Fanny Brawne, in which case I have surrounded the poem
with narrative elements to create fully attached autobiography, I am likely to make
some equivalence of vision between the poem’s unnamed persona and the ‘‘I’’ of John
Keats. That is because, as I have been suggesting and as Brian McHale has stated
outright, ‘‘the assumption of autobiographical authenticity, of an identity between the
poem’s ‘I’ and the poet’s self, is something like the ‘default setting’ for lyric poems’’
(McHale 2003: 235). Were I to read the ‘‘I’’ of ‘‘Bright Star’’ as wholly detached from
the ‘‘I’’ of Keats by positing an unreliable speaker and thus an ironic rendering of the
sonnet, I would not only be changing the meaning of the sonnet but, in transgressing
the conventions of lyric reading, turn ‘‘Bright Star’’ into a detached dramatic
monologue.
What I am proposing as the conventional reading of lyric would be relatively
unconcerned with the poem’s narrative elements, in this case with the hope to be
‘‘pillowed’’ and with the identity or even the real existence of the ‘‘fair love.’’ Let me
make this key point with a stronger example. Sharon Olds’s short poem ‘‘Son’’ begins
with an event: ‘‘Coming home from the woman-only bar,/ I go into my son’s room.’’
The poem goes on to describe the sleeping child, and it ends with the exhortation,
implicitly addressed to other feminist women, that ‘‘Into any new world we enter, let
us/take this man’’ (Olds 1984: 68). For the final line to have significance, it must
count as the author’s position in the same sense as does a letter to the editor; the
poem’s point of view must be guaranteed by a person, and the person the poem evokes
must be not a fictitious speaker but ‘‘Sharon Olds.’’ At the same time, however, I do
The ‘‘I’’ of the Beholder
215
not think this poem requires me to accept as a narrative truth that the writer Sharon
Olds came home from a woman-only bar on a particular evening and gazed upon her
sleeping son. Whether Olds actually experienced this moment or only imagined it
does not compromise my attachment of the poem’s thoughts and feelings to Sharon
Olds. That is, I may recognize the narrative elements as a convenient fiction – attached
therefore to a purely fictional narrator – while still attaching to Olds the author the
viewpoint of the final words.5
We can explain this readerly practice by recalling Peter Rabinowitz’s concept of the
‘‘rules of notice.’’ Rabinowitz reminds us that readers, unable to remember and attend
to the signification of everything in a text, must create a ‘‘hierarchy of importance’’ to
distinguish what matters to their reading from what does not (Rabinowitz [1987]
1998: 44). Such rules of notice ‘‘tell us where to concentrate our attention,’’ and those
elements on which we concentrate ‘‘serve as a basic structure on which to build an
interpretation’’ (Rabinowitz 1998: 53). Certain textual elements – titles, first and last
sentences, primary narrators, central characters – conventionally receive greater attention than other elements. But Rabinowitz recognizes that the grids of notice also
vary from genre to genre. I suggest that readers of lyric poetry so privilege nonnarrative thought, feeling, and vision that events can slide ‘‘beneath notice’’ when they
are not central to our understanding of the poet’s thought. Or, put differently, I can
attach the lyric ‘‘I,’’ as it were, to the author while rendering the narrative ‘‘I’’ fictional,
even though technically this ‘‘I’’ is a single entity. I can do so because I recognize that the
narrative apparatus of the poem exists in the service of nonnarrative purposes. On
somewhat different grounds but through a similar logic, I can detach as a fiction the
eye-witness claims George Eliot’s narrator makes in Scenes of Clerical Life – for
example, about being a member of ‘‘Shepperton Church as it was in the old days’’
(Eliot [1858] 1973: 42) – while attaching quite firmly to the author the philosophical
generalizations this same narrator makes about the world.
In the case of lyric, then, readers probably do make attachments and detachments
on more than technical grounds, overriding the technical singularities about ‘‘who’s
speaking’’ that structuralist narratology holds dear. I want to propose an inverse and
even more equivocal readerly behavior with respect to homodiegetic (and possibly also
fully focalized figural) fictions, in neither of which can one technically identify any
‘‘voice’’ that is authorial and thus ‘‘supposed’’ to be attached. It is important to
acknowledge that narrative fiction presents a tougher case. Brian McHale rightly
notes that if the ‘‘default setting’’ for lyric is ‘‘autobiographical authenticity,’’ the
‘‘default setting’’ for the relationship between fictional narrator-characters and their
authors is just the opposite: ‘‘impersonation,’’ not ‘‘authenticity’’ (McHale 2003: 235).
I would affirm this axiom, which is also held by nearly all narrative theorists:
impersonation, and thus what I am calling detachment between authorial and textual
‘‘I,’’ is indeed the default condition of narrative fiction. But I am also arguing that this
default position is frequently transgressed, and transgressed most often through the
same criteria that link author and voice in the lyric. In the case of narrative fiction, of
course, the story cannot quite fall beneath notice as it might with respect to Olds’s
216
Susan S. Lanser
poem. Yet readers may ignore the technical boundaries of fictional voice, in effect
doubling the ‘‘I’’ so that the narrator’s words sometimes belongs to the author as well as
to the narrating character and sometimes do not. Most commonly, as I have suggested,
this doubling occurs through a split between the text’s narrative and nonnarrative
textual elements, so that narrative elements are relegated purely to the fictional
narrator-character, while nonnarrative elements may be tied to the author as well as
to the narrator-character. If, however, readers believe that a fiction is autobiographical,
they can enact this doubling for most or all of a text without splitting the temporal
from the atemporal: if I have reason to believe that Ann Beattie’s own father died in
circumstances similar to those her character Ann narrates, even if not precisely on
Christmas Day, then I might read many elements of the text as potential autobiography. For as Dorrit Cohn admits in spite of her own admonitions to maintain
boundaries, ‘‘the distance separating author and narrator in any given first-person
novel is not a given and fixed quantity but a variable’’ that gives readers enormous
‘‘interpretive freedom’’ (Cohn 1999: 34). The most likely scenario, I suggest, is that
readers will detach many or all of the narrated events from the author but weigh
nonnarrative elements according to the other criteria I have set forth.
Transgression: The Human Stain
Let me take as an example Philip Roth’s novel The Human Stain, which uses the firstperson voice of the writer Nathan Zuckerman to tell the story of Coleman Silk, a
black man who passes for Jewish and whose distinguished academic career is ruined
by a misguided accusation of racism. As I read The Human Stain, I know that Nathan
Zuckerman is not Philip Roth. I do not assume that Roth has prostate cancer just
because Zuckerman does – though Norman Podhoretz does raise this question in a
review of the novel in which he notes that although the facts of Zuckerman’s life may
differ from the author’s, it has become ‘‘harder and harder . . . to distinguish’’ Roth
from Zuckerman (Podhoretz 2000: 36). But when Zuckerman the narrator writes a
several-line paean to male friendship, I do attach that paean to Philip Roth. When, in
the next moment, the same narrator re-enters narrative time with ‘‘This probably isn’t
usual for him, I was thinking’’ (Roth 2001: 27), I cease to align Nathan Zuckerman
with Philip Roth – or rather, Philip Roth dips once again beneath my notice as an
‘‘I’’. In the space of a paragraph, I have split the narrative from the nonnarrative
discourse of the same ‘‘I’’ narrator, attaching one to, and detaching the other from,
‘‘Philip Roth.’’ Later in the novel, I find myself associating with Philip Roth the
words of yet another character, the woman Faunia Farley, because it is by way of
Faunia’s musings, focalized through free indirect discourse, that the novel offers us the
meaning of its title phrase, ‘‘the human stain.’’ At that moment, in other words, I am
attaching together Nathan Zuckerman (who is allegedly ‘‘writing’’ Faunia’s story),
Faunia herself, and Philip Roth, even though a piece of me may want to resist the
connection between this female character and an author with a history of misogyny.
The ‘‘I’’ of the Beholder
217
(A similar process leads me, when I read Defoe’s The Fortunate Mistress, or Roxana, to
attach to Defoe – perhaps even more than to Roxana – the extensive critique of
marriage as the source of women’s miseries, a critique that Roxana herself does not
precisely heed.) But the split I’m suggesting readers of lyric make between temporal
and atemporal discourse is not always clear-cut: what Bakhtin has described as the
‘‘dialogism’’ of language, exploited especially in the novel, can arguably commingle
atemporal and temporal discourse, and attached and detached voices, in the same
sentence and even the same word.
Seeing ‘‘I’’ to ‘‘I’’
I have been claiming, then, that our reading of textual voice does not simply follow
the rules of discourse; it adheres to another logic that is not only formal and structural
but pragmatic and contextual, ‘‘staining’’ the divide between fiction and the real. The
farther a homodiegetic narrator wanders from the demands or details of story,
I speculate, the more likely that voice is to get authorially attached. I would even
hypothesize that when a narrator is exercising what Marie-Laure Ryan calls the
‘‘testimonial’’ function of narrative, only a perception of unreliability actively prevents
readers from attaching the author to the homodiegetic narrator’s voice. That is
perhaps one reason why unreliability preoccupies narrative theorists as it does not
preoccupy theorists of lyric or drama, and why it is usually homodiegetic rather than
heterodiegetic voice whose reliability worries us.
This perverse propensity for attachment may also help to explain why the inability
to attach an autodiegetic ‘‘I’’ to the ‘‘I’’ of an identifiable author can leave readers quite
adrift; I have discussed elsewhere (Lanser 2003) an anonymous work of 1744, The
Travels and Adventures of Mademoiselle de Richelieu, for which two sophisticated scholars,
Carolyn Woodward and Susan Lamb, have produced quite oppositional readings.
Woodward sees the work as a subversive lesbian and feminist novel; Lamb considers
it an antifeminist satire. Each scholar anchors her reading through a claim not simply
about the autodiegetic narrator but about the author’s politics and indeed the author’s
sex: Woodward’s imagined author is a feminist woman, Lamb’s a misogynist man.
Both scholars attach the narrator’s ideological pronouncements to their imagined
author, and yet both treat the events narrated by the autodiegetic narrator as pure
invention.
When it comes to the equivocal project of fiction, then, the beholder has a queer
and shifty eye, attaching and detaching by turns, and doing so despite the ‘‘fact’’ that
all fictional voices are fictions or that each ‘‘I’’ can technically represent only one
speaking entity. The Ann Beattie story with which I began offers metafictional
support for such behavior. The plot focuses on the narrator’s widowed mother,
whose involvement with another man after her husband’s death is viewed as tantamount to insanity by the daughter-narrator still mourning her father’s death. The
story’s title, ‘‘Find and Replace,’’ suggests the daughter’s bitterness at the mother’s
218
Susan S. Lanser
substitutionary act. But the narrator also proposes ‘‘Find and Replace’’ as a recipe for
composing fiction and thus arguably also for reading it. In conversation with a
stranger, Ann explains the mechanism by which writers turn autobiography into
fiction: ‘‘you just program the computer to replace one name with another. So, in the
final version, every time the word ‘Mom’ comes up, it’s replaced with ‘Aunt Begonia,’
or something’’ (Beattie 2001: 88). Later in the story, in order to avoid a speeding
ticket, the character herself enacts such a replacement when she tells a police officer
that her mother is dying. Her reasoning is arguably a program for fiction as well:
‘‘After all, it was a terrible situation. The easiest way to express it had been to say that
my mother was dying. Replace ‘lost her mind’ with ‘dying’’’ (Beattie 2001: 89). But
this one act of replacement leads Ann to other permutations of her story in order to
tell what she insists to the reader is nonetheless ‘‘the truth.’’ When I then reread
her opening sentence – ‘‘True story. My father died in a hospice on Christmas Day’’ –
I can, without a sense of contradiction, accept the claim ‘‘true story’’ as Ann Beattie’s
along with Ann’s (admittedly glib) comment on the project of fiction, while (at least
until I have reason to believe otherwise) attributing ‘‘My father died in a hospice’’ and
all that follows upon this assertion to the fictive ‘‘Ann.’’
I am not suggesting that the watchword of fiction, if I may pervert a phrase of
E. M. Forster, should be ‘‘only attach.’’ But I do think it worth probing further how,
when, and why attachment might happen despite formal imperatives against it, and
revising our narrative theories to allow for the ways in which readers may be breaking
the rules of structure as a way to give literary speech acts their fullest significance. In
‘‘What Is an Author?’’ Foucault imagines a time when ‘‘who’s speaking’’ would no
longer matter and ‘‘discourse would circulate without any need for an author [ . . . ]
unfolding in a pervasive anonymity’’ (Foucault 1977: 138). This world of wholly
detached genres may seem more dystopic than Foucault envisioned in an age of
anonymous websites and anthrax-laced letters with no return address. Yet complete
attachment is surely equally disturbing in an age of internet ‘‘cookies’’ and fingerprinting at airports. This may be just the moment to revel in the equivocality
of literature, with its attached and detached visions and the imaginative space to
‘‘find and replace,’’ and in that process to behold multiple others, real and invented,
‘‘I’’ to ‘‘I.’’
NOTES
1 Biographies and histories sometimes adopt
these same techniques of omniscience and focalization, but these methods are fictional liberties taken with historical material, and are
usually recognized and often deplored as such.
2 For a fuller sense of the debate surrounding
the ‘‘Social Text Affair,’’ see Professor Sokal’s
website: <http://www.physics.nyu.edu/faculty
/sokal/>.
3 I do not mean to suggest here that nonfictional
genres are not worthy of the designation
‘‘literature.’’ The constitutive definition of literature that I draw from Genette recognizes
literature as a set of genres that rely for their
construction, for their very existence, on
fictive or imaginary objects. In this constitutive
sense, ‘‘pulp’’ fiction is literature, but autobiography and essay are not. When literature
The ‘‘I’’ of the Beholder
is taken in its evaluative sense, however,
Rousseau’s Confessions and Montaigne’s Essais,
both attached texts, of course become literature, as indeed does Mary Karr’s The Liar’s
Club.
4 I am of course not using ‘‘lyric’’ to be synonymous with ‘‘poetry,’’ which runs the generic gamut from, say, Pope’s quite attached and
REFERENCES
AND
Beattie, A. (2001). ‘‘Find and Replace.’’ The New
Yorker, November 5, 84–9.
Booth, W. (1961). The Rhetoric of Fiction. Chicago:
University of Chicago Press.
Burke, K. (1941). ‘‘Literature as Equipment for
Living.’’ In The Philosophy of Literary Form: Studies in Symbolic Action (pp. 293–304). Baton
Rouge: Louisiana State University Press.
Carter, F. [A.] (1976). The Education of Little Tree.
New York: Delacorte Press.
Cohn, D. (1999). The Distinction of Fiction. Baltimore, MD: Johns Hopkins University Press.
Eliot, G. ([1858] 1973). Scenes of Clerical Life.
Harmondsworth, UK: Penguin.
Foucault, M. (1977). ‘‘What Is an Author?’’ In
Language, Counter-Memory, Practice: Selected Essays and Interviews, ed. D. F. Bouchard, trans.
D. F. Bouchard and S. Simon (pp. 113–38).
Ithaca, NY: Cornell University Press.
Gates, H. L., Jr. (1991). ‘‘ ‘Authenticity,’ or the
Lesson of Little Tree.’’ New York Times Book
Review, Nov. 24, 26.
Genette, G. ([1991] 1993). Fiction and Diction,
trans. C. Porter. Ithaca, NY: Cornell University
Press.
Griffin, R. (1999). ‘‘Anonymity and Authorship.’’
New Literary History 30, 877–95.
Hamburger, K. (1973). The Logic of Literature,
trans. M. J. Rose. Bloomington: Indiana University Press.
Lanser, S. (2003). ‘‘The Author’s Queer Clothes:
Anonymity, Sex(uality), and The Travels and
219
authorial ‘‘Essay on Man’’ to Browning’s quite
detached and dramatic ‘‘My Last Duchess.’’
5 It is possible that the poem would be compromised were I to learn that Sharon Olds did
not have a son. In a modest way, such knowledge might lead me to see the poem as a kind
of ‘‘lyric hoax’’ of the sort that Brian McHale
describes (McHale 2003).
FURTHER READING
Adventures of Mademoiselle de Richelieu. In R.
J. Griffin (ed.), The Faces of Anonymity: Anonymous and Pseudonymous Publication from the Sixteenth to the Twentieth Century (pp. 81–102).
New York and Basingstoke, UK: Palgrave Macmillan.
Lejeune, P. (1980). Le pacte autobiographique. [The
Autobiographical Compact.] Paris: Seuil.
McHale, B. (2003). ‘‘ ‘A Poet May Not Exist’:
Mock-Hoaxes and the Construction of National
Identity.’’ In R. J. Griffin (ed.), The Faces of
Anonymity: Anonymous and Pseudonymous Publication from the Sixteenth to the Twentieth Century
(pp. 233–52). New York and Basingstoke,
UK: Palgrave Macmillan.
Olds, S. (1984). The Dead and the Living. New
York: Knopf.
Phelan, J. (2005). Living To Tell About It: A Rhetoric and Ethics of Character Narration. Ithaca,
NY: Cornell University Press.
Podhoretz, N. (2000). ‘‘Bellow at 85, Roth at 67.’’
Commentary, July/August, 35–43.
Rabinowitz, P. J. ([1987] 1998). Before Reading:
Narrative Conventions and the Politics of Interpretation. Columbus: Ohio State University Press.
Roth, P. (2001). The Human Stain. London:
Vintage.
Santiago, D. [D. James] (1983). Famous All Over
Town. New York: Simon & Schuster.
14
Neonarrative; or, How to Render
the Unnarratable in Realist
Fiction and Contemporary Film
Robyn R. Warhol
In 1988, in the first special issue of Style devoted to ‘‘Narrative Theory and Criticism,’’
Gerald Prince published a typically ambitious brief essay called ‘‘The Disnarrated.’’
That issue of Style (the same number where Susan Lanser and Nilli Diengott held their
notorious debate over the possibility of bringing politics and narrative theory together in a feminist narratology) did for narrative studies in 1988 what the present
volume is to do for the early twenty-first century, showing the current state of the
field. I find myself still intrigued with Prince’s contribution to that issue, and with
the fate of the term he coined there, ‘‘the disnarrated,’’ which means ‘‘those passages in
a narrative that consider what did not or does not take place’’ (Prince 1988: 1). Prince
concluded his essay by allying himself with us feminist narratologists, on the one
hand, and, on the other hand, with Ross Chambers or Marie-Laure Ryan (who, Prince
says, were insisting on taking narrative situation always into account), pulling our
work together under the heading of ‘‘Pragmatics.’’ Calling this emphasis on pragmatics ‘‘the most striking difference between what might be called the classical narratology of the 1960s and 1970s and what might be designated as modern narratology,’’
Prince asserted that narratology turned toward pragmatics because of the repeated
(sociolinguistic) reminders about the importance of communicative contexts, because
of the great interest among literary critics and theorists in the receiver and his or her
decoding strategies, because of the growing number of cognition-oriented studies,
and because of the (renewed) awareness that narrative must be viewed not only as an
object or product but also as an act or process, as a situation-bound transaction
between two parties, as an exchange resulting from the desire of at least one of
these parties (Prince 1988: 7).
I believe that we are still practicing Prince’s ‘‘modern narratology,’’ and that the
concerns he cites still motivate what has evolved into a fundamentally pragmatic
approach to texts. In this spirit, I am returning to the topic of ‘‘the disnarrated,’’
Neonarrative
221
a term which fell into disuse soon after Prince coined it.1 I concur with Prince’s claim
that ‘‘the nature and content of the disnarrated . . . , the level at which it functions . . . ,
the relative frequency with which it appears, and the relative amount of space it
occupies can be a useful tool for characterizing narrative manners, schools, movements, and even entire periods’’ (Prince 1988: 6). Although Prince does not mention
them, literary genres can also be characterized by the degree to which they use
disnarration, and the nature of the material their narrators explicitly omit. I will
draw here on examples from literary narratives for the purposes of defining terms,
then look at a narrative form Prince did not consider – fictional film – for insight into
the role the disnarrated, as well as its associated tropes, ‘‘the unnarrated’’ and ‘‘the
unnarratable,’’ have played in filmic form. If the ‘‘disnarrated’’ describes those passages
in a narrative that tell what did not happen, what I call the ‘‘unnarrated’’ refers to those
passages that explicitly do not tell what is supposed to have happened, foregrounding
the narrator’s refusal to narrate. While Prince’s work describes ‘‘the disnarrated’’ as an
object in itself, I am interested in viewing disnarration and unnarration as narrative
acts, much in keeping with Prince’s vision of the pragmatics-focused future of
narratology.
I examine the disnarrated, as well as its affiliated figure, the unnarrated, and the
larger category to which they both belong, the unnarratable, for the ways they serve as
distinctive markers of genre. Narrative genres are known as much by what they do not
or cannot contain as by what they typically do contain: a novel or film belongs to a
particular genre as much for what it does not say or do as for what it does. The limits
of narratability vary according to nation, period, and audience as well as genre, but
they also stretch and change as genres evolve. Indeed, shifts in the category of the
unnarratable are, I would say, significant indicators of generic change itself, and they
both reflect and constitute their audiences’ developing senses of such matters as
politics, ethics, and values. The ‘‘disnarrated’’ along with ‘‘the unnarrated’’ are
instances of narrators’ making explicit the boundaries of the narratable; sometimes
disnarration and unnarration also become strategies for moving the boundaries
outward, and changing the genre itself. When disnarration and unnarration lead to
genre change, they are participating in what I will call ‘‘neonarrative,’’ or narratorial
strategies for making narrative genres new.
Varieties of the Unnarratable in Classic Realist Fiction
First, I want to make some terminological distinctions that Prince himself never
got around to making, although both his Narratology: The Form and Functioning
of Narrative and his Dictionary of Narratology do offer definitions upon which
I can build. Prince never defined ‘‘the unnarratable,’’ collapsing it instead with ‘‘the
nonnarratable’’ in defining his term, ‘‘disnarration’’; nor did he make the noun
forms he coined into verbs, that is, ‘‘to unnarrate’’ or ‘‘to disnarrate.’’ Starting with
his definition for the positive term, ‘‘narratable,’’ then, I will build a vocabulary
222
Robyn R. Warhol
for talking about what is not there in fictional narratives, both novelistic and
filmic.
Drawing on the work of Peter Brooks and D. A. Miller from the early 1980s,
Prince’s Dictionary of Narratology defines ‘‘the narratable’’ as ‘‘that which is worthy of
being told; that which is susceptible of or calls for narration’’ (Prince 1987: 56). If
‘‘unnarratable’’ (which Prince does not define in the dictionary) is to be the antonym
for this term, then it would mean ‘‘that which is unworthy of being told,’’ ‘‘that which
is not susceptible to narration,’’ and ‘‘that which does not call for narration’’ or perhaps
‘‘those circumstances under which narration is uncalled for.’’ Prince does define the
‘‘unnarratable’’ as ‘‘that which, according to a given narrative, cannot be narrated or is not
worth narrating either because it transgresses a law (social, authorial, generic, formal)
or because it defies the powers of a particular narrator (or those of any narrator) or
because it falls below the so-called threshold of narratability (it is not sufficiently
unusual or problematic)’’ (Prince 1988: 1). Because I think there are significant
pragmatic differences between that which ‘‘cannot be narrated’’ and that which ‘‘is
not worth narrating,’’ I will reorganize these definitions of the unnarratable to
highlight this distinction. For the sake of analysis, I will classify my illustrative
examples of the unnarratable from classic realist fiction in four categories. This is not
meant to be an exhaustive list of all possible forms of the unnarratable; indeed, I offer
these categories as four among many possibilities: that which, according to a given
narrative, (1) ‘‘needn’t be told (the subnarratable),’’ (2) ‘‘can’t be told (the supranarratable),’’ (3) ‘‘shouldn’t be told (the antinarratable),’’ and (4) ‘‘wouldn’t be told (the
paranarratable).’’ While that which is unnarratable may sometimes simply be left
implicit in a narrative (Prince 1982: 135), the ‘‘unnarratable’’ does not always lead to a
textual silence or to what Meir Sternberg has called ‘‘blanks’’ and ‘‘gaps’’2: in the
instances I will cite, it becomes explicit. I focus in my definitions on two narrative
strategies for representing the unnarratable: ‘‘disnarration’’ – telling what did not
happen, in place of what did – and ‘‘unnarration’’ – asserting that what did happen
cannot be retold in words, or explicitly indicating that what happened will not be
narrated because narrating it would be impossible.
The subnarratable: what needn’t be told because it’s ‘‘normal’’
The first definition, ‘‘that which is unworthy of being told,’’ is the equivalent of what
Prince calls ‘‘the non-narratable,’’ or the ‘‘normal’’ (Prince 1987: 52). These are events
that fall below the ‘‘threshold of narratability’’ because they ‘‘go without saying,’’
events too insignificant or banal to warrant representation. As Prince remarks, ‘‘If
I told a friend what I did yesterday, I would most probably not mention that I tied my
shoelaces’’ (Prince 1987: 1). Depending on the genre, however, the boundaries around
what is worth saying may cover a wide range of normal territory. I think of the scene
in the Richard Lester-directed Beatles’ movie, Help! (1965), in which Victor Spinetti
rolls his eyes and grimaces while Leo McKern tells him over the walkie-talkie, ‘‘I am
Neonarrative
223
moving my left foot. I am moving my right foot.’’ In realist Victorian fiction, as in the
mind of McKern’s character, the threshold for this kind of narratability is low indeed.
Novels such as Anthony Trollope’s are filled with narration detailing ‘‘the normal’’ at a
seemingly minimal level of significance, for example:
Lady Lufton got up and bustled about; she poked the fire and shifted the candles, spoke
a few words to Dr. Grantly, whispered something to her son, patted Lucy on the cheek,
told Fanny, who was a musician, that they would have a little music; and ended by
putting her two hands on Griselda’s shoulders and telling her that the fit of her frock
was perfect.
(Trollope [1861] 1987: 158)
Trollope’s narrator tells us that Lady Lufton was trying to make everyone comfortable
at her party, going about it the long way. Still, even in this passage, such details as the
motion of Lady Lufton’s feet as she bustles about, the blinking of her eyes, the beating
of her pulse ‘‘go without saying’’ – these details, along with myriad others that
Trollope does not include in sketching out the action of the scene, are not narrated
because they are taken for granted. So are the sequences of action that got Lady Lufton
from her bed that morning to the party that evening, all of which are absent from the
text. They are literally unremarkable – they need not be told. For this category,
I would substitute for Prince’s ‘‘nonnarratable’’ my new term, ‘‘subnarratable,’’ in
order more clearly to distinguish it, as Prince does not, from the larger term,
‘‘unnarratable.’’ In realist fiction, the subnarratable is not marked by either disnarration or unnarration; like Prince’s ‘‘nonnarratable,’’ its presence is typically marked by
its absence.
The supranarratable: what can’t be told because it’s ‘‘ineffable’’
The second variety of the unnarratable, ‘‘that which is not susceptible to narration,’’
comprises those events that defy narrative, foregrounding the inadequacy of language
or of visual image to achieve full representation, even of fictitious events. The
supranarratable is the category that often forms the occasion for what I am calling
‘‘unnarration’’ in classic realist texts, beginning with Tristram Shandy’s recourse to an
all-black page as the antiexpression of his grief for ‘‘poor Yorick’s’’ death (Sterne
[1759–67] 1967: 232) and carrying over in the sentimentalist tradition to narrators’
assertions that ‘‘I don’t think I have any words in which to tell’’ the specifics of highly
charged emotional scenes (Alcott [1868] 1983: 187). Louisa May Alcott’s sentimentalist narrator habitually unnarrates such moments, with comments such as ‘‘the
shock she received can better be imagined than described’’ (p. 359) or ‘‘Beth ran
straight into her father’s arms. Never mind what happened just after that, for the full
hearts overflowed, washing away the bitterness of the past and leaving only the
sweetness of the present’’ (p. 206). This is the unnarratable as the ineffable: events
that simply cannot be retold. The supranarratable is the one of these four types of the
224
Robyn R. Warhol
unnarratable that most consistently carries a textual marker in the form of the explicit
disclaimer I am calling ‘‘unnarration.’’
The antinarratable: what shouldn’t be told because of social convention
The antinarratable transgresses social laws or taboos, and for that reason remains
unspoken. This is the unnarratable as ‘‘that which does not call for narration,’’ in the
specific sense that it might prompt a response of ‘‘that’s uncalled for.’’ (I am thinking
of the phrase some people use to scold others for making impertinent or inappropriate
remarks.) Sex in realist Victorian novels, for instance, is always antinarratable, and can
only be known by its results as they play themselves out in the plot (for instance in
the presence of new babies, disillusioned hearts, or ruined reputations). As I have
written elsewhere, Victorian narrative uses euphemism, allusion, metaphor, and
especially metonymy to signify sexual connection between characters, but never
narration – and not even unnarration of the kind premodernist novels use to represent
sentimental affect.3 The same is true in Victorian fiction for most bodily functions,
not just copulation but especially excretion, so that when James Joyce places Leopold
Bloom on a toilet in Ulysses, he is making perhaps his most radical break with
Victorian limits of unnarratability by changing the boundaries of the antinarratable.
Twentieth-century novels like William Faulkner’s Absalom, Absalom! or Toni
Morrison’s Beloved, to give just two prominent examples out of many possibilities,
challenge the unnarratability of another kind of taboo, the silence that results from
trauma. These novels can be said to be about the struggle to tell what supposedly
shouldn’t be told, that which has been repressed both from history and from the
characters’ own consciousnesses. The much-commented-upon ambiguity of Beloved’s
narrator’s assertion that ‘‘It was not a story to pass on’’ combines, as Shlomith
Rimmon-Kenan says, ‘‘acceptance and rejection, an injunction to remember and a
recommendation to forget’’ (Rimmon-Kenan 1996: 123), thus pointing to the difficulties of embarking on a narrative of the antinarratable taboo. While Abdellatif
Khayati has quite persuasively written that Beloved traffics in the ‘‘unaccountable
language of the ineffable’’ (Khayati 1999: 315), which would imply that it can’t be
told (or that it is supranarratable), I would argue that Sethe’s story, and Beloved’s too,
eventually does get told in the course of the narrative: what has been repressed or
suppressed because it shouldn’t be told, gets expressed before the novel’s end because it
must be told for healing to occur and, for that matter, for the novel to get written. As
Khayati rightly argues, ‘‘Morrison’s grappling with the ‘unspeakable’ for her characters – namely, the fear that to evoke a past degradation may diminish them, humiliate
them, and shame them – is clear from the way in which they try to force forgetting
into a willed activity’’ (Khayati 1999: 319). Many commentators, Khayati among
them, have shown that what began as the unnarratable trauma in Morrison’s novel
ultimately becomes narratable, in a narrative process moving toward wholeness for the
characters and for the text itself.
Neonarrative
225
Trauma in Victorian realist fiction can also be figured as antinarratable, as belonging beyond the bounds of what should be told, but the boundaries of this kind of
narratability are stricter in Victorian than in twentieth-century texts. Many traumatic
experiences of characters in Victorian fiction – even in the sensation novels of Wilkie
Collins, for instance – happen outside the diegesis, and are narrated second-hand if
they are told at all. First-hand or focalized accounts of trauma in Victorian novels are
rare, making trauma the kind of antinarratable experience represented mainly by
silence or gaps in realist texts. Charles Dickens’s David Copperfield unnarrates the
childhood trauma of laboring in Murdstone and Grinby’s warehouse: ‘‘No words can
express the secret agony of my soul as I sunk into this companionship. . . . and felt my
hopes of growing up to be a learned and distinguished man, crushed in my bosom.
The deep remembrance of the sense I had of being utterly without hope now. . .
cannot be written’’ (Dickens ([1849–50] 1981: 210). Charlotte Brontë’s Villette
presents an interesting instance of a disnarrated trauma, in the famous passage
where the evasive narrator Lucy Snowe accounts for her teenage years:
I betook myself home, having been absent six months. It will be conjectured that I was
of course glad to return to the bosom of my kindred. Well! The amiable conjecture does
no harm, and may therefore be safely left uncontradicted. Far from saying nay, indeed,
I will permit the reader to picture me, for the next eight years, as a bark slumbering
through halcyon weather, in a harbour still as glass – the steersman stretched on the
little deck, his face up to heaven, his eyes closed: buried, if you will, in a long
prayer. . . . Picture me then idle, basking, plump, and happy, stretched on a cushioned
deck, warmed with constant sunshine, rocked by breezes indolently soft. However, it
cannot be concealed that, in that case, I must somehow have fallen over-board, or that
there must have been wreck at last. I too well remember a time – a long time, of cold, of
danger, of contention. To this hour, when I have the nightmare, it repeats the rush and
saltness of briny waves in my throat, and their icy pressure on my lungs. I even know
there was a storm, and that not of one hour nor one day. . . . In fine, the ship was lost, the
crew perished.
(Brontë [1853] 1979: 94)
I call this disnarration because Lucy tells what did not happen, in two different ways:
first she asserts that she will not contradict the ‘‘amiable conjecture’’ that she was
happy to return home, strongly implying that the conjecture is mistaken, though it
‘‘does no harm.’’ She says she will, in other words, not deny what did not happen. The
passage then moves into a more elaborate disnarration, as Lucy plays out the metaphor
of the storm. If she had been floating comfortably along (as she has implied she was
not), then she must have fallen overboard – the description of the shipwreck is only an
analogy for what is supposed to have ‘‘really’’ happened to Lucy. The metaphor (which
turns out to be drawn from a nightmare referring to the unspeakable trauma) is at the
same time a narrative, telling the story of a shipwreck that did not happen, which
Lucy offers in place of a narrative of what did happen. This is disnarration, a Victorian
novelistic solution for narrating the antinarratable, or what shouldn’t be told.
226
Robyn R. Warhol
The paranarratable: what wouldn’t be told because of formal convention
This last category of the unnarratable comprises that which transgresses a law of
literary genre without being recognizable as sub-, supra-, or antinarratable. For
example, in the feminocentric nineteenth-century novel, as Nancy K. Miller (1980)
so memorably pointed out, the heroine can in the end only get married or die. Laws of
literary generic convention are more inflexible, I believe, than laws of social convention, and have led throughout literary history to more instances of unnarratability
than even taboo has led. My students always look puzzled when I tell them that there
were many more possibilities for the life-stories of real Victorian women than there
were for Victorian heroines: it’s counterintuitive, because fiction would seemingly be
limited only by imagination, while ‘‘real life,’’ as the students call it, is bound to
follow the dictates of what undergraduates like to call ‘‘society.’’ I explain to them that
many, many, young women neither got married nor died, but (like Jane Austen, the
younger Brontë sisters and George Eliot, whose own lives followed trajectories very
different from their heroines’) found other ways to pass out of youth into maturity. For
an early to mid-nineteenth-century heroine, however, to find a different end from
marriage or death is unthinkable: how many of them publish a novel, or cheerfully set
up housekeeping with an unmarried sister, or pursue a professional career? As feminist
commentators have frequently observed, the influence of dominant ideology on
fictional form is too powerful. Such a story is paranarratable in that period: although
it certainly could happen ‘‘in life,’’ ‘‘in literature’’ it wouldn’t have been told.
For a Victorian novelist to choose an alternate outcome to a heroine’s marriage plot,
then, would be to attempt to narrate the paranarratable. Again Villette presents an
example of how Charlotte Brontë renders the unnarratable, this time not just by
disnarrating, but also by unnarrating it. Describing how she has waited for M. Paul’s
return by ship from South America, Lucy brings the narration to the point when he is
to set sail. She embarks on another description of a storm (metaphorical? literal? It
could be either and is probably both), but backs off from positively killing her hero,
for his death would be paranarratable, according to the laws of the genre in which she
writes:
Here pause: pause at once. There is enough said. Trouble no quiet, kind heart; leave
sunny imaginations hope. Let it be theirs to conceive the delight of joy born again fresh
out of great terror, the rapture of rescue from peril, the wondrous reprieve from dread,
the fruition of return. Let them picture union and a happy succeeding life.
(Brontë [1853] 1979: 596)
That’s Lucy’s last word on the fate of her marriage to M. Paul, and if Charlotte Brontë
wrote the conclusion this way to please Patrick Brontë – whose dismay at Lucy’s
unconventional ending led him to beg his daughter to leave some room for hope –
most commentators agree that the narrative provides little doubt that M. Paul is
Neonarrative
227
indeed to be presumed dead. As in the example of the antinarratable above, Lucy
disnarrates the ending, telling what did not happen (‘‘the rapture of rescue from peril,
the wondrous reprieve from dread, the fruition of return’’). This time, though, the
disnarration is introduced by unnarration, ‘‘There is enough said.’’ The unnarration
states what is not to be told, in this case not because it is insignificant, ineffable, or
taboo, but because in the Victorian novel it is paranarratable; it would not be told: it
simply would not do.
Neonarrative in Film: Stretching the Boundaries of the
Unnarratable
For a set of examples of how this vocabulary for the unnarratable might be useful in
describing shifts in generic form, I turn now to popular films that circulated in the
United States around the turn of the twenty-first century. Film makes literally
graphic the changes and developments in its form, and sometimes a particular
film’s experiments with widening the boundaries of the narratable can make us
conscious of unnarratability we hadn’t realized was previously there. The following
are examples of instances where films’ management of three of the categories of the
unnarratable draws attention to a genre’s limits and, in many cases, actually pushes
those limits out to widen the possibilities for filmic narration.4
The subnarratable
The threshold of narratability varies greatly among filmic genres, just as it does for
novelistic ones, and certain experimental films (Andy Warhol’s first film, Sleep [1963],
for instance, or Jim Jarmusch’s Stranger than Paradise [1984]) challenge the idea that
anything at all should be outside the realm of the narratable. Nothing on the level of
plot ever happens in these films, so every gesture, every breath or flick of an eyelid
becomes an event worthy of narrative representation. Such challenges to norms of the
subnarratable have been finding their way lately into popular film, sometimes for a
specifically aesthetic purpose – as in the scene in Sam Mendes’ American Beauty (1999)
where the camera lingers for 60 long seconds on a plastic bag being blown up against
a brick wall by a gusty wind, described by the young filmmaker in the movie as ‘‘the
most beautiful thing I’ve ever filmed’’ – and sometimes for more explicitly ethical or
political ends. An excellent example is Three Kings, the anti-Gulf War film David
O. Russell directed in 1999. As one character asks another to think seriously about
what happens when you shoot a gun at someone, the film shows a soldier being fatally
shot, but instead of focusing conventionally on the external view of the body falling
and hitting the ground, the frame jumps to a computer-animated internal image of
the flesh and organs being torn apart by the intruding metal, complete with sound
effects. The interior damage done by a bullet has up to this point been, for Hollywood
228
Robyn R. Warhol
film, a subnarratable set of details, something that goes without saying. By violently
anatomizing ‘‘the normal,’’ Russell’s technique broadens the range of the narratable for
contemporary film, calling into question what needn’t be told. As this technique
for imaging the injured body has been picked up on network television’s CSI, it has
become increasingly familiar, eventually losing the consciousness-raising value it
holds in Three Kings. When it happens in Three Kings, it is neonarrative; by the
time it has become a familiar feature of a network TV series, it is one of the
conventions of a newly changed genre.
The antinarratable
Examples of ways in which contemporary film has broken social conventions or taboos
by representing subjects previously considered beyond the pale are too numerous to
elaborate – so, for the purpose of illustration, I will cite only the most vivid example
I can think of, Boogie Nights, directed in 1997 by Paul Thomas Anderson. After a long
and complicated story hinging always on the reputed largeness of the porn-star
protagonist’s penis, the film concludes with a mirror-view of the organ itself: a rare
instance of making literal the phallic engine that propels not only this action film,
but the majority of Hollywood films in its genre. Since so much of this movie, like
other action films before it, is about the fact that the film cannot show the audience
‘‘how big it is,’’ the visual revelation at the end comes with a shock. That moment in
Boogie Nights functions as neonarrative for Hollywood film, but it would be unremarkable in, for example, a porn film, as it would in a photograph by Robert
Mapplethorpe. In stretching the boundaries of the antinarratable in one genre (in
this case, mainstream action film), neonarrative can borrow from another genre (here,
porn or gay art), and appropriate images or elements of story that represent a more
radical anticonventionality in their earlier forms. Once the phallus has made its
appearance in Boogie Nights, it enters the vocabulary of convention for Hollywood
films. It shows up again two years later, for example, in a locker room scene in Oliver
Stone’s 1999 football film, On Any Given Sunday.
The paranarratable
More interesting, for my present purposes, are those films that defy formal convention
by visually narrating incidents that may not be socially remarkable, but have been
hitherto unimaginable as filmic images. I read Spike Lee’s first feature-length movie,
She’s Gotta Have It (1986) as a formalist exercise in narrating scenes that had been
paranarratable for mainstream film up to that point: Lee’s film shows in detail a black
woman oiling a black man’s scalp, an African-American couple performing a Fred
Astaire/Ginger Rogers-style dance in 1950s Technicolor romantic movie style, and a
black man’s rape of a black woman, among many other images that mainstream films
had not previously framed. The introduction of these images into Lee’s movie is
neonarrative (or it becomes neonarrative when Lee’s film-school project gets released
Neonarrative
229
as a mainstream feature film), as it introduces them also into the realm of the
narratable for Hollywood film at large.
Conventions governing plotline and closure form another boundary of the narratable in mainstream film. The ‘‘narratable,’’ in this sense, is what can be imagined as
possibly happening within the plot of a given genre produced in a specific time and
place; as its opposite, the ‘‘paranarratable’’ could be defined as ‘‘the unthinkable.’’ For
example, in a classical Hollywood romantic comedy, it is not possible for the heroine
to die at the end, or for the romantic male lead to ditch her; nor is it possible for her to
decide to leave him at last for a better job in another city. Such outcomes are
paranarratable: they wouldn’t be told, because the conventions governing the genre
have proven too strong to allow them to be told without disrupting the genre
altogether. Several contemporary Hollywood comedies in this genre have pushed at
the boundaries of paranarratable endings (for instance, P. J. Hogan’s 1997 My Best
Friend’s Wedding, and Nicholas Hytner’s 1998 The Object of My Affection) by introducing gay male characters into the relationship mix and resolving the heroine’s story
with alternatives to marriage. When neonarrative enables a paranarratable plot closure
to become narratable in this way, the distinguishing characteristics of the genre of
romantic comedy have changed.
Disnarration and Unnarration in Film
If disnarration is simply telling of something that did not happen, many familiar
filmic conventions might qualify as disnarration, including fantasy or dream sequences and subjective points of view. Arguably, though, a fantasy sequence in
a film is a piece of ordinary narration: the film tells that the character is having
this fantasy or this dream, and the film signals (by the character’s suddenly jolting
awake, for instance) that the sequence is not to be taken as an authoritative set
of actions within the diegesis. Often, too, the focalization of events through a
protagonist’s point of view leads to implicit conclusions about what is ‘‘really’’
happening that are false, but get corrected by the film’s end: Richard Rush’s 1980
Peter O’Toole vehicle, The Stunt Man, for instance; or Francis Ford Coppola’s
The Conversation (1974), starring Gene Hackman; or, more recently, The Usual
Suspects, directed in 1995 by Bryan Singer. In each of these films, the first-time
viewer realizes – with a jolt – along with one of the focal characters that the diegetic
world is not what it has seemed to be. These films are not employing disnarration,
though – the filmic narrative has presented the same story all along, while the
difference comes in interpretation of what happened, not in presentation of what
happened.
I would reserve ‘‘disnarration’’ to refer to the technique used in several recent films
which tell alternate versions of the same story, without marking one version or the
other as ‘‘what really happens’’ diegetically until the end. Examples would include
Run Lola Run, Tom Tykwer’s 1998 German film released in the United States in
230
Robyn R. Warhol
1999, in which the titular character’s being late for a rendezvous leads to a sequence of
disastrously violent consequences adding up to a tragic ending. As in a video game
constantly reset by a persistent player,5 the action rewinds and plays out more than
once, and as the last version of the narrative concludes, the disasters most closely
touching the protagonist are ultimately averted. A similar structure based on the
contingent relationship of events governs the narration of Fisher Stevens’s Just A Kiss
(2002), in which an act of unfaithfulness leads to multiple suicides, accidents,
brutalizations, and murders before a reversal of time erases the unfaithful kiss,
resulting in a much less violent and happier ending. In both Run Lola Run and Just
a Kiss, I am arguing that the violent versions are disnarrated because the film gives
priority to the happier closure which comes at the movies’ end. Other, similarly
structured films (such as Peter Howitt’s Sliding Doors from 1998 and Spike Jonze’s
Adaptation from 2003) test the limits of my definition of disnarration, as they do not
offer a privileged narrative to choose from among the various versions of the diegesis
they present. Their postmodern technique more closely resembles alternate world or
parallel universe narratives typical of science fiction films and novels, in which the
question of what ‘‘really’’ is supposed to have happened is ultimately unresolvable,
even within the terms set out by the text.
The category I find most fascinating, and also most elusive, in this grammar of the
unnarratable in film, is ‘‘unnarration,’’ or the narrator’s assertion that what happened
cannot be rendered in narrative. The only clear examples I have found of unnarration
in film come from Hollywood classics of an earlier era, such as Leo McCary’s 1957
romance, An Affair to Remember. In that movie, the hero and heroine share a kiss offscreen while the camera frames their bodies up to mid-torso. The kiss itself is left to
the viewer’s imagination; the aggressively odd framing of the gesture is tantamount to
an assertion that the experience cannot be captured in narrative. This exemplifies the
supranarratable, or the unnarratable as the ineffable, that which cannot be told. Here
the kiss outside the filmic frame is being ‘‘unnarrated,’’ not out of prudery or
squeamishness or Hollywood censorship codes, but as a narrative means of indicating
that the emotion of the moment transcends representation. Alfred Hitchcock’s Frenzy
(1972) presents a similar case, in the scene where the camera tracks up a staircase,
stopping outside the door of the apartment where a woman is being brutally
murdered; so does John Ford’s 1956 Searchers, when John Wayne looks through a
window at the devastation wrought by an Indian raid on a settler’s household, the
camera focusing on his back rather than tracking with his gaze. In both cases, the
unnarration asserts that the scenes themselves are too horrific, supranarratable: they
can only be told by being not-told. The supranarratable in contemporary film seems
to reside primarily in horror movies, where the inability to see the terrifying object
can still be scarier than even the most vivid special effects, as in The Blair Witch
Project, directed by Daniel Myrick and Eduardo Sánchez in 1999. Using the hand-held
camera and the fictional documentary situation to misdirect the audience’s generic
expectations, The Blair Witch Project expands the genre of the horror film by refusing
to narrate the source of the horror.
Neonarrative
231
As I hope these examples from Victorian novels and contemporary film have shown, a
vocabulary for discussing the varieties of the unnarratable can be helpful in identifying specific changes in genre over time. Passages of disnarration and unnarration are
signposts pointing to the supranarratable, the antinarratable, and the paranarratable
in a given genre at a given moment, marking what cannot, shouldn’t, or wouldn’t be
told while still maintaining its unnarratability. My aim has been to provide a way to
talk about what narratives do not tell.6
NOTES
1 Brian Richardson addressed a similar topic in
‘‘Denarration in Fiction’’ (2001) but he was
coining a new term for the phenomenon he
observed there.
2 See The Poetics of Biblical Narrative (Sternberg
1985, especially pp. 235–58).
3 See ‘‘Narrating the Unnarratable’’ (Warhol
1994: 85–8).
4 For examples of the supranarratable in film,
see the section on unnarration in film, below.
REFERENCES
AND
Alcott, L. M. ([1868] 1983). Little Women. Toronto
and New York: Bantam.
Brooks, P. (1985). Reading for the Plot: Design and
Intention in Narrative. New York: Vintage
Books.
Brontë, C. ([1853] 1979). Villette. Harmondsworth, UK: Penguin.
Dickens, C. ([1849–50] 1981). David Copperfield.
Harmondsworth, UK: Penguin.
Khayati, A. (1999). ‘‘Representation, Race, and
the ‘Language’ of The Ineffable in Toni Morrison’s Narrative.’’ African American Review 33(2),
313–24.
Miller, D. A. (1981). Narrative and Its Discontents:
Problems of Closure in the Traditional Novel.
Princeton, NJ: Princeton University Press.
Miller, N. K. (1980). The Heroine’s Text: Readings
in the French and English Novel, 1722–1782.
New York: Columbia University Press.
Prince, G. (1982). Narratology: The Form and Functioning of Narrative. Berlin, New York, and Amsterdam: Mouton.
5 Tom Whalen (2000) has pointed out the parallels between the narrative of Run, Lola, Run,
and video game structure, seeing the audience,
along with Lola, as players. I thank Alan
Nadel for pointing me in the direction of
this article.
6 I am grateful for substantive help on this
project from Emma Kafalenos, Peggy Phelan,
Jay Clayton, Todd McGowan, James Vivian,
Peter Rabinowitz, and Jim Phelan.
FURTHER READING
Prince, G. (1987). A Dictionary of Narratology. Lincoln and London: University of Nebraska Press.
Prince, G. (1988). ‘‘The Disnarrated.’’ Style 22(1),
1–8.
Richardson, B. (2001). ‘‘Denarration in Fiction:
Erasing the Story in Beckett and Others.’’ Narrative 9(2), 168–75.
Rimmon-Kenan, S. (1996). A Glance Beyond
Doubt. Columbus: Ohio State University Press.
Sternberg, M. (1985). The Poetics of Biblical
Narrative: Ideological Literature and the Drama of
Reading. Bloomington: Indiana University
Press.
Sterne, L. ([1759–67] 1967). Tristram Shandy.
Harmondsworth, UK: Penguin.
Trollope, A. ([1861] 1987). Framley Parsonage.
Harmondsworth, UK: Penguin.
Warhol, R. R. (1994). ‘‘Narrating the Unnarratable: Gender and Metonymy in the Victorian
Novel.’’ Style 28(1), 74–94.
Whalen, T. (2000). ‘‘Run Lola Run.’’ Film Quarterly 53(3), 33–40.
15
Self-consciousness as a Narrative
Feature and Force: Tellers vs.
Informants in Generic Design
Meir Sternberg
Self, Action, and Transmission: Unifying the Narrative Field
In the narrative of human agency – the generic mainstream – private discourse is a
condition of narrativity itself. Strictly, the characters needn’t talk to each other at all
but must think in the acting. Typically, they do both, changing voices (along with
arenas, priorities, behaviors) on the social/secret axis. Most narrative texts, further,
combine the characters’ public and private life, verbal as otherwise, if only because
action logic involves psychologic, and vice versa. Definitionally so, hence even across
media, within any teleological model of agency: from Aristotle’s end-directedness at
large to cognitivism’s problem solving. The agent then doubles as subject, mimesis
plots embodied minds, self-address elucidates and/or counterpoints the overt dealings
with the world, dialogues not least. Just think of the role played in causal enchainment by the doer’s viewpoint – hope, fear, goal, motive, planning, knowledge,
ideology – as either cause or effect of doing. From the reader’s side, consider how
the emergence or opacity or timing of such hidden views and forces on the narrative
surface affects our processing of the enacted dynamics: between certitude and gapfilling hypothesis, irony and ignorance, suspense and curiosity or surprise, closure and
open-endedness, for example.
The manifestations of the secret life alone vary and its, or their, relations to the
social arena, the world in flux, the entire narrative text. Besides sequential disclosure,
the manifestations widely vary in the object rendered, as among private thought,
speech, and writing, or among the types of thought itself, including the operative
psychic model; likewise with the form of rendering: direct, indirect, free indirect,
telescoped, or latent in the external world or the artful design. (Some finer variants
will come later.) The relations, again, vary in the traffic or balance or hierarchy
between the public and the private domains, all the way to reverse subordinacies: as
Self-consciousness as a Narrative Feature
233
with the Odyssey versus Ulysses. Throughout, however, acting twins – even merges –
with reflecting, emplotment with perspective, due to their joint anchorage in the self.
It is on such a fundamental ground, and across such a range of variability, that
narrative, known as the most ‘‘dialogic’’ of genres, typically includes ‘‘monologic,’’ or,
better, what I call unself-conscious discourse. By self-consciousness I mean the discourser’s awareness of addressing an audience, transmitting a message. In this basic sense,
the term is independent of further traits associated with it in some common usages,
looser or value-laden.1 Sophistication, tight control, literariness, fictionality, reflexivity, for instance, may but needn’t go with self-consciousness, nor are they definitionally precluded from unself-consciousness. By this feature, instead, discoursers in
narrative polarize (or shuttle) between tellers and informants: those who communicate
with another about the world, as against those who lead their secret life and
unwittingly mediate in the process another’s higher-level communication.
On the one hand, we have Fielding’s persona in Tom Jones, Rousseau in his
Confessions, Humbert in Lolita, or for that matter, any dialogist within their world
and the dramatic monologist elsewhere; on the other hand, there is the secret diarist,
like Samuel Pepys in real life or Bridget Jones in fiction, the vocal soliloquist, usually
less regular, and the interior monologist, from sentence- to work-length. Where the
issue turns on transmission-mindedness, no paradox attaches to the fact that otheroriented discourse is self-conscious, self-discourse unself-conscious. But if the apparent paradox of selfhood in and out of touch with alterity troubles you, then replace the
terms by cognizant vs. uncognizant or addrecentric vs. egocentric (Sternberg 1983b).
In face of this polarity, we all sense, I believe, that the difference makes quite a
difference – if only from our knowledge of the chasm between what we tell and what
we keep to ourselves, or even from ourselves, complete with the respective modes of
presentation. It is therefore strange that, as narratologists, we have done so little to
act, or reflect, upon this tacit difference, regarding either our likes in everyday
(hi)storytelling or fictional creatures denied our privilege of keeping the secret life
secret, thanks to the bounds of human insight.
I have long tried to bring this feature into the prominence it deserves, but with
limited success thus far. In my first book on telling/reading in time, I introduced selfconsciousness as a key factor in the interplay between two omnipresent axes of
viewpoint, the discourser’s potential (or competence) and performance; hence also as
a key factor in the still larger, generic relation between point-of-view system itself and
temporal structure or even the entire narrative text.2 What with the self’s twin plotrole, therefore, the very integrity of the field is at stake.
In a nutshell, all discourse choices are ultimately motivated (determined, justified,
patterned, explained) in terms of the author’s silent communicative design; the author
in turn motivates those choices via the chain (as well as the object) of transmission
interposed, or mediating, between the authorial and the readerly end; and the
transmissional work done by each of the links in the chain crucially depends for its
rationale and shape and impact on their own awareness of transmitting, if only to an
audience other than the author’s.
234
Meir Sternberg
From the reader’s side, everything goes in reverse. Faced with the data as transmitted, we progressively infer (and if necessary, remake) some line of transmission
along which they assume operative shape: the one presumably designed by the
ultimate, reticent yet self-conscious, communicator, our opposite number. Originated
for this purpose, the in-between line we puzzle out alternates or crosses tellers with
informants to suit, among various functional relays (e.g., writing/speaking/thinking,
knowledgeable/unenlightened, right-/wrong-minded). Just as we motivators invest
the transmission with forms en route – down to the smallest, like pockets of direct,
indirect, free indirect quoting – so we endow the transmitters with features: by
whatever makes the wanted purposive sense, if only provisionally, of the text in
context. Reading, subtle or simple, thus constitutes a trial-and-error process of
motivating the givens to yield their best intermediate fit under the likeliest authorial
rules and goals.
Accordingly, with point of view understood as a (re)constructed end/means or
indeed end/mediacy system, the feature at issue gravitates from constancy at the
ultimate transmissional source to variability, even reversibility, along the mediating
chain. In fictional narrative, these options most proliferate between extremes, and
most invite crossing or enchainment, even of the extremes themselves. At one
extreme, a narrator is created in the author’s own image, endowed with all the relevant
powers, including equal audience-mindedness as well as omniscience, artistry, reliability, sense of communicative purpose. At the extreme opposite to omnicompetence,
the creaturely transmitter receives none of these authorial privileges – least of all, the
awareness of transmitting that makes (or here, in absentia, breaks) narratorhood. And
without such awareness, there also evaporates the drive toward interesting, moving,
guiding, persuading any outsider, let alone the implicit readership. Reduced to the
lowest potential, even below communicating humans in life and art, the informant
then mirror-images the superhuman teller as well as the originator of both.3
Such a gulf in competence will bear on the respective performances, notably their
distance from the powers delegated and from the delegating author behind the scenes.
Nothing in the text escapes these polar bearings, yet the handling of the event
sequence is the most fundamental of them all, because generic: the interaction of
narrated with narrating time distinctively constitutes narrative. It has therefore the
strongest claim to illustrating the principle at diametric work.
How, then, does enabled relate to encountered transmission, possible to actual as
well as to authorial stance, and what’s the role of (un)self-consciousness in their
(mis)match? Apropos temporal strategy, for example, the author-like narrator can
by virtue of omniscience tell us everything in time, just as things happened and as
would best serve their fullness, development, intelligibility in the reading selfconsciously managed. Omnicompetence indeed strategically matches and produces
omnicommunicativeness, hence orderly sequencing, throughout a narration like
Trollope’s, and for exactly these reasons. But then, given a diametric enterprise, the
quasi-author becomes suppressive instead: temporal performance runs counter, not
true, to the same narratorial powers in Fielding or Dickens. Bent on ambiguity
Self-consciousness as a Narrative Feature
235
instead of lucidity – on the artful disclosure rather than the natural-looking development of knowledge or judgment about the world – the narrator there will exploit
the very highest privileges to twist the event-line into the wanted surprise and
curiosity dynamics of reading.
In stark contrast to the omnicompetent narrator, the diarist quiet at the desk and
the monologist in thought are by nature unself-conscious, usually also liable to
ignorance and the rest of human fallibility. Here, therefore, it is all-round powerlessness that readily translates into a Fielding-like time performance, confronting us with
a multigap sequence to fulfill much the same generic ends – a reading unsettled all
along by curiosity or surprise – only from within the world, and the secret arena at
that. Having nobody but themselves to consider, Molly Bloom and Bridget Jones
plunge in medias res, then proceed toward further discontinuities, as befits the
workings of psychic license twinned with epistemic limitation. Midward, backward,
forward, sideward: the jumps reflect their mental state from moment to moment,
while newly forcing us into uneasy progressive closure, ever vulnerable to lingering
riddles or belated discoveries.
By a remarkable yet reasonable coincidence, superhuman and all too human
qualifications, public and private viewpoint, sheer world-representing and represented discourse, ironic teller and self-betraying informant, all meet here, in the
transmission of a winding, elliptic, ambiguous narrative. So they meet, but, against
appearances and current opinion, not necessarily. Even when the thinking or soliloquizing or journal-keeping informant could and would do otherwise in outbound
telling, it may appear, why do it in the free egocentricity of self-address?4 But, given
the will to Trollope-like lucidity, the author will nevertheless always find a way to
have the unwitting as well as the addrecentric transmitter begin at the beginning and
proceed in step with the world’s time, more or less, even from exposition through
complication toward resolution. Among narrative’s universal effects, future-bound
suspense dynamics will accordingly outweigh the backward-looking curiosity or
surprise to transform the entire reading experience. The very time-scheme of the
diary encodes this linear march between entries and enables it within each. Pepys thus
describes the relevant states of affairs (his own, his family’s, the nation’s) ‘‘at the end of
the last year,’’ with Cromwell gone and nothing certain, before recording the first
particular day, 1/1/1660, from morning to bedtime. Also, a review of one’s life, with a
view to making honest sense of its flow, will explain the retracer’s adherence to
chronology in solitary thought. ‘‘Just as it happened. In the right order’’: so the
widower in Dostoevsky’s ‘‘A Gentle Creature’’ understandably chooses to recall, ab ovo,
his catastrophic marriage.
Across the gulf, then, either perspectival extreme freely correlates with any temporal strategy; either form of mediacy can serve either basic set of narrative interests,
as with means and end in general. To the author, as to the reader, therefore, the most
high-powered and the most disempowered transmissions contrast not in the range of
time effects open to performance but in their logic of motivation: aesthetic and
mimetic, functional and fictional, respectively.5
236
Meir Sternberg
At the heart of this contrast, adjustable to either logic, self-consciousness operates
as a prime motivating factor. Endowed with it, omnicompetent telling, whether
arrowlike or crooked, is thus motivated in purely effective, aesthetic terms: the
‘‘communication has no existence, hence no role and reason, on the level of the fictive
reality. . . but takes place exclusively within the framework of the rhetorical relationship between narrator and reader’’ (Sternberg 1978: 247), author’s deputy and author’s
target. Even when less than immediate – outside metanarrative – their dealings need
only pass through the generic object, which the one renders and the other reads as a
first-order reality (un)made for some end, like the world of Tom Jones plotted for
comedy. In the absence of any lifelike subjective mediation, and so motivation, to
interrupt this contact, the discourse addresses itself to us straight – often indeed
against time, but always in line with authorized narratorial teleology. The ordering of
events goes by an order of sheer artistic priorities, with, say, lucidity or ambiguity,
development or disclosure, at their head.
Under the mimetic logic of transmission, however, equivalent (dis)orderings stand
at some further, substantial remove(s) from the author’s frame cum priorities. They no
longer reach, affect, enlighten, or trouble us virtually firsthand, objectified on the
highest communicative authority, but through the mediation of characters who
appear to be leading an independent existence in a world and discourse world other
than ours as frame-sharers. The characters’ very speaking, narrating, reflecting,
writing, belong to that existence, can always get it wrong and, as activities, change
it: transmissional joins with physical agency or eventhood, figural life-imaging with
living itself, under the umbrella of mimesis. The dynamics of the reading process,
orderly or tortuous, is then propelled not just in the guise of the dynamics of the
world-in-action but also of enacted discourse.
Like all enactment, the characters’ discoursive acts, public or private, narratorial or
self-oriented, mediate the author’s artful design by interposing some lifelike pattern
that embodies, camouflages, distances it: that both realizes and removes the operative
poetics. Here lies the key to the overdue theorizing and branching of this elusive
concept.6 Mediation, I would suggest, amounts to mimetic motivation, namely,
representation for an end beyond itself, below the worldlike surface. Russian warand-peace thus mediates Tolstoy’s ideology; Wonderland, Carroll’s nonsense;
Gulliver’s narrative of his travels, Swift’s satire; Emma Woodhouse’s unshared
thoughts, self-willed or contrite, Austen’s precarious rhetoric.
So the mediating fiction always stands to the mediated function as carrier and cover
at once, vehicle and veil. Only, when the characters join discoursive to physical agency
– Gulliver or Emma style – mediacy extends accordingly from the world in flux to
some world-bound perspective on the world: the transmissional chain lengthens, the
transmitted event-chain humanly wavers, and the object in general darkens, along
with the authorial intent governing them all at a commensurate remove. The mimesis
overlying the poesis is then not first- but second-order. It represents not (fictional)
existence proper but a (figural, vocal, silent, written) representation thereof, a subjective image of what counts as the tale’s objective reality.
Self-consciousness as a Narrative Feature
237
Therefore, I further argue, all such re-presentational mediacy entails quotation:
discourse within discourse, discourse about another’s discourse about the world
originally made from a viewpoint independent of the enclosing narrative frame that
transmits it at second hand, orchestrates it by remote control to new effect. The
inset as given has its artistic motivation (e.g., why work-length or piecemeal, why
verbatim-looking or edited or obliquely echoed, why more reliable or less) in the
quoting frame’s teleology; the frame owes its mimetic motivation (for whatever arts
designed) to the subjectivity (e.g., blindness) of the quotee’s inset discoursing. And
nowhere does this subjective, perspectival mimetic guise so thicken as with quoted
discoursers whose unself-consciousness runs to the limit of not even addressing their
fellow creatures: they think, speak, write along self-propelled lines, unmindful of
outsiders, let alone us readerly intruders on privacy by the frame’s courtesy.
By nature, all quotees remain ignorant of their framing between other parties for
other purposes. However, if acutely self-conscious narrators, once inset, turn informants relative to the higher-level discourse around and about them, then such informanthood will a fortiori characterize quoted voices that are egocentric. These
informants can therefore deploy broken time-lines too erratic and enigmatic for
would-be communicators, whether equally quoted or telling in propria persona.
Contrast Bloom’s stream of associations with the disciplined twists of a mystery
narrated by Watson or Fielding’s surprise plot. Nor need the egocentric transmitter
show more reader-friendliness when inclined toward chronological ordering –
elsewhere notoriously judged too simple, dull, artless, but here assuming subjective
interest as well as cover. Thus a life unself-consciously recorded seriatim, or retraced.
Either way, given pure informanthood, the master effects built into the genre’s
twofold happening/reading dynamics arise by a logic of motivation polar to the
omnicompetent narrator’s. Among eligible rationales, the mimetic now always overlies the aesthetic or otherwise communicative, and within mimesis itself, the perspectival displaces the objectively existential line of accounting for surface
(e.g., sequence) givens.
As with the interplay of times that defines narrativity, moreover, so with each and
every aspect of the narrative text. The polarized rule stretches from the (im)mediacy of
causal, analogical, thematic, ideological, intertextual sense-making down to the given
language: how the option between, say, well-formed and ungrammatical, even incoherent, discourse gets motivated by appeal to the respective transmitters. Compare
the Dickens spokesman truncating his sentences for sheer expressive impact with
Pepys’s or Molly Bloom’s verisimilar private shortcuts.7
Again, beyond temporality, the motivation for the author’s designs and effects may
change guises, even logics in sequence, as between narrated reality and some quoted
perspective on it. The mimesis overlying Carrollian nonsense thus gravitates, in late
retrospect, from an apparently first-order Wonderland to Alice’s dream of Wonderland. And as with global, story-length mediated transmission, so with local frame/
inset shifts. At will, Fielding’s deputy mimeticizes ad-hoc any artful choice by
changing roles and voices from narrator to quoter, toward enacted speech and
238
Meir Sternberg
thought, inset talk and inside view: whatever emerges or stays hidden to whatever
effect, it’s the current quotee’s (e.g., the soliloquizing Tom’s) responsibility, as it were.
Indeed, given the teleological action represented, its communicator must interpose and
enchain some re-presentations of agentive private views and scenarios: the plotters
within the plot necessarily double as informants to this extent, event line as relay line.
For global informants, again, changing tacks is easier still. Under some lifelike
pressure, the monologist or diarist (e.g., Dostoevsky’s Underground man) can break
into and out of chronology, well-formedness, soul-searching, other-echoing, or even
address to an imagined audience.
The ever-available shifts of or within the dominant logic of perspectival motivation, (un)self-consciousness and all, cannot be emphasized too strongly. Nor, inversely, can the systematic, often work-length interweaving of the two rationales in
further well-defined perspectival configurations. There, neither merging nor breaking
with the chief transmitter all along the line, the author rather fictionalizes the
narrative art by depriving that transmitter either of omniscience and the like or of
self-consciousness.
In the one instance, there arises the dramatized narrator, autobiographer, eyewitness. We encounter a representer re-presented at work, naturally limited but
communicating (e.g., sequencing) well or ill inside the represented world, at some
distance from the author’s own artistic frame and accordingly manipulable, even
exposable in its service. By remote control, therefore, the narration operates, shuttles,
wavers, divides, evolves between incompetence and authority, or what we read as such
for the best fit. Monologist-like self-betrayal itself persists – a Watson’s, Holden’s,
Humbert’s – now amid occupational, if always lower-level, self-consciousness, as well
as an apparent monopoly on discourse. When novelistic, such inset forms an enormous direct quote of a public narrative, oral or written, which freely encloses in turn
quotes of all kinds, styles, levels, voices, beginning with the narrator’s experiencing
self. But direct quotation that runs and recounts and relays from cover to cover, minus
title page, is a quotation under a director still.
In the other instance, the Jamesian paradigm emerges: a self-conscious omniscient
narrator self-restricted to a humanly unself-conscious ‘‘reflector’’ or ‘‘vessel of consciousness,’’ figural, fallible, thinking and observing, and so mediating the world for
us in multiple (discoursive cum epistemic) lifelike ignorance, as wanted by the
framing communicator. A human mediator quoted (typically nondirectly, preferably
free indirectly) by a superhuman mediator, that vessel, like its (directly) speaking
fellow, can in turn quote ad lib to remove the artist’s design further – only with the
greater latitude and naturalness kept for egocentricity. James’s Lambert Strether,
monologist-like, reflects in every sense; Holden also tells, not least self-conscious in
telltale defiance of autobiographical decorum.
The variety-in-unity of these two interpolar forms of transmission should leap to
the eye by now, what with their (over)privileged status in modern writing and
theory.8 Instead, let us move toward a higher vantage point, a more comparative
research program trained on this feature and its fortunes.
Self-consciousness as a Narrative Feature
239
Some Obstacles and Further Ramifications: A Comparative
Balance Sheet
The original theory as just outlined has since been taken up by some narratologists,
with notable results. For example, David Bordwell (1985: 57ff.) and Kristin
Thompson (1988) have boldly extended this idea of self-consciousness to cinematic
narrative, where narratorhood itself is in question. Again, in redefining (un)reliability
as a text-integrating measure, Tamar Yacobi (e.g., 1981, 2000, 2001) has located the
unreliable narrator vis-à-vis the author at the poles of giveaway information and goaldirected communication. Symptomatically, though, her own followers on the inferred
status of the narrator’s (un)reliability have often overlooked this rudiment, to the
limit of erasing the authorial reference-point itself. (But see Cohn 2000: 73, 89, 148.)
Reducing mediators instead, the author-oriented Booth (1961: especially 149–54),
with parallels before and since, erases the narrator/informant line, as a mere technicality of ‘‘person.’’ Inversely, the few analogues to our transmissional divide, though
better than the common oblivion, or erasure, leave more holes and oddities than they
repair.
So the task ahead is twofold. On the widest front, there remain both a great deal of
misunderstanding, even resistance, to be overcome and still more work of exploration,
pulling together, finer line-drawing to be carried out on the topic itself. A condition
and test of the advance proposed, moreover, uncovering the impediments to it also
freshly illuminates the nature, centrality, and range, or subgrouping, of the phenomena at issue. So the levels of analysis will intercross below. But I don’t have the space
now for going into both the metatheory and the theory required, except on a
programmatic scale. Let me therefore critically adumbrate the former and, constructively, suggest in the process how the latter may develop the opening argument.
What tendencies, then, have conspired against this key feature? First, the partitioning of the generic whole. Recall the basics of narrative economy with which
I started: the actional value of the subjective life, and vice versa, as well as their
protean manifestations. In E. M. Forster’s bare ‘‘plot’’ exemplar, ‘‘The king died and
then the queen died of grief,’’ (Forster 1962: 93) how to separate the public from the
psychic, or enchainment from mind-imaging? The uncommon effect of the one death
on the survivor not only generates the other death but also individuates the sufferer,
via miniature thought-quotation: the widow’s collapse is the action’s coherence.
(Contrast ‘‘of grief ’’ with ‘‘of typhus’’.)
Note also that, with every mimetic perspectival guise assumed by the author’s
designs on us, big or small, another agent, another force for and possibly object of
change, enters the represented arena. A quotee leads a life, a dramatized subject
(mouth, hand, eye, ear, mind) is dramatizable at all levels, to all purposes. So
the slightest move or shallowest motive of a doer necessarily reflects a viewpoint on
the world; the barest viewpoint of an observer refracts, impels, or itself enacts some
movement in the world. Beyond twinship, either operates as a two-in-one. Yet the
240
Meir Sternberg
former, if registered, would traditionally be consigned to the sphere of ‘‘action,’’ or
character in action; the latter, to ‘‘narration’’ or ‘‘focalization.’’
Such compartmentalizing issues from an age-old split in theoretical agenda between plot and perspective, doing and discourse, representation and communication,
even fabula and sjuzhet – down to what defines narrative, hence where its study should
center. For the respective camps along history, think of Aristotle and Plato;
E. M. Forster or Vladimir Propp and Percy Lubbock; Ronald Crane and Wayne
Booth; Claude Bremond and Gérard Genette; cognitivist and pragmaticist story
analysis. Recent narratologies, despite increasing gestures toward holism, still cut
asunder what narrative joins together. Lately, the above miniature plot, interlinking
the two royal deaths from within, has even been found wanting by zealots of both
parties. It counts as underplotted in Ryan (1991), because driveless, planless, unconflictual; and as short on ‘‘experientiality’’ (i.e., underquoting, overobjective), by
modernist standards, in Fludernik (1996). With narrativity itself barely allowed to
the exemplar on either ground, at that, the convergent negatives alleged only
radicalize in little the persistent split.
The losses suffered, and the challenges posed, largely mass at this juncture, namely,
the interplay of transmissional feature with actional force, either way. Unselfconscious discourse, where treated at all, has got partitioned between the camps.
The less classifiable or elaborate or prestigious forms (e.g., the telescoped inside view
latent in Forster’s queen dying ‘‘of grief ’’) are assimilated to emplotment; the more
determinate or expressive or just novel-looking (e.g., interior monologue, free indirect
thought, however causative) are assigned to perspectivity, specifically mind representation. Forms of private happening like these cry out for treatment as such or as a whole,
always in their two-in-one re(-)presentationality. Inversely, for that matter, with selfconsciousness. Despite theory’s heavy investment not only in narration but also in
speech representation, what do we know about the kinds and movements and
workings of enacted dialogue, from local communicative interaction among speakers
in and viewpoints on the world to plot value and oblique rhetoric within the authorial
frame? A staple of narrative, it would appear to have fallen between the discipline’s
stools.
However, even in the perspective-oriented camp, especially that born of Structuralist narratology, this hole-ridden compartmentalizing goes with a second obstacle.
I refer to the influence of the linguistic model. Here it compounds other, betterknown restrictive effects in privileging not simply communication, but face-to-face
communication. Such talk exchange, with the partners changing roles between
speaker and addressee, marks of course the antipole to unself-conscious discourse, a
fortiori if soundless and wordless as well as addresseeless.
Since Saussure, this oblivion to the noncommunicative half of our experience, lived
or lifelike, has become the rule in various fields – pragmatics, discourse analysis,
philosophy of language, as well as linguistics and semiotics – with repercussions on
Structuralist narratology and beyond. Unparalleled in impact is Emile Benveniste’s
notorious opposition of tellerless ‘‘(hi)story’’ to subjectivity-flaunting ‘‘discourse’’;
Self-consciousness as a Narrative Feature
241
but, for its lesson and lineage to be duly appreciated, we must trace them back to
the roots.
Throughout Benveniste’s important work on person deixis as/and the subjectivity
of language, note first the extreme interpersonal bias of his master coinage ‘‘ ‘instance
of discourse,’ . . . by which language is actualized in speech by a speaker’’ (1971: 217).
More insistently yet, ‘‘discourse is language put into action, and necessarily between
partners’’: from primordial time, we find ‘‘man speaking to another man,’’ so that
‘‘language provides the very definition of man’’ (pp. 223, 224).
This forcing of a speech partnership on language use to constitute person(ality),
subjectivity, even humanity itself, on top of discoursivity, artificially restricts the
whole lot. Oddly, ‘‘subjectivity’’ comes down to humans at their least subjective (least
private, hence least freely egocentric, individual, off-guard). Surface-deep, it then
invests us at our most other-minded, because communicative, hence most disinclined
to express or reveal our hidden selves, best forearmed against self-exposure as informants. The theory accommodates only the self-conscious ‘‘enunciatory’’ half of our
discourse activity, or rather the social tip of the iceberg, for even the most talkative
among us think more (and frequently, of course, other) than they speak, not least in
speaking. Everything else drops out by fiat.
Thus, where is all silent self-communion, which eludes an outsider’s access in reality
but not one’s own introspection? Oblivious to empirics, further, the linguist also
typically blanks out at least two ranges of accessible unself-consciousness, one lived
and spied upon, one lifelike and reportive. One absentee is all spoken and written
private discoursing, given in principle to overhearing and, as it were, overreading,
respectively. If you don’t listen at keyholes, you look over Pepys’s shoulder. The other
is our licit second-order experience – all everyday and literary re-presentations of all
modes of privacy in all quoting forms. With Benveniste’s ‘‘discourse is language put
into action . . . between partners,’’ contrast my inclusive threefold definition of (verbal)
discourse, as a piece of language that renders, or relays, a piece of world from a certain
standpoint.
In turn, the narrow scope leads to outright misconception of the forms in play,
especially the personal deictics and the situational roles they encode. False Gordian
knots are tied between the partners supposed to coform the instance of discourse.
Thus the knot of role-reversibility: ‘‘the one whom ‘I’ defines by ‘you’ thinks of
himself as ‘I,’ and ‘I’ [then] becomes ‘you’ ’’ (Benveniste 1971: 199). An invariant of
two-way communication alone, however, such turn-taking grows variable in communication at large – a novelist’s audience can’t speak back – and wholly avoidable out of
it. There, in ego-land, real or represented, our own or a storied informant’s, no ‘‘you’’
has to be faced and defined, much less inverted into ‘‘I.’’ Even where present, this
‘‘you’’ remains optional, interior, self-made and self-voiced or self-silenced. Discourse,
thus remodeled, would join bipolarity and flexibility to inclusiveness: the systemic
gains compound, proportionately to the opposite minuses.
Likewise with the alleged mutual implicativeness of the deictic terms, as contrasting members of Benveniste’s ‘‘correlation of personality,’’ exclusive of the ‘‘he,’’ the
242
Meir Sternberg
nonpresent nonperson. Now, from the tolerable premise that ‘‘ ‘I’ designates the
one who speaks,’’ it does not follow that ‘‘ ‘you’ is necessarily designated by the
‘I’ ’’( Benveniste 1971: 197), so that ‘‘neither of the terms can be conceived of without
the other’’ (p. 225). For unself-conscious discourse may keep out even the self-made
‘‘you’’ altogether: either to focus on the ego as ego, the usual way, or to correlate
instead with a ‘‘he,’’ allegedly outside the deictic circle. In this opaque third-person
guise, Blazes Boylan haunts the mind of the cuckolded Bloom. The principle holds
even on the belief, often enacted in literature since antiquity, that thought is the soul
discoursing with itself: the self-discourse may well remain unaddressed to any psychic
entity, as always in the Bible’s inner speech. By a logic that accommodates discourse as
a whole, therefore, it is rather ‘‘you’’ (exterior or interior to the self) that necessarily
designates an ‘‘I,’’ self-conscious or otherwise, but not vice versa. The ‘‘correlation of
personality’’ needn’t work both ways, just the one least in keeping with Benveniste’s
rationale and paradigm. In turn, this one-way implication brings home the true
universal dependency-relation. It confirms and explains precisely the dispensability
of the ‘‘you’’ in our mental representations, as the special, secondary term of the pair,
hinging on whether the ‘‘I’’ opts for or against address, self-address to a countervoice
in private dialogue included.
Last but not least fallacious among the package deals is that of superficial,
especially pronominal, reflex with underlying situational role. The linguist occupationally adheres to the surface – as if it were freezable outside Utopia’s regimented
Unispeak – often conflating the variable, multiform, even dispensable pronouns I/you
with the universal and protean discourse-roles speaker/addressee. Conflating reflex
with role, indexical with indexed, worsens the earlier imbalance between the two
discourse-situations. Even in terms of this very fallacy, the two’s shared repertoire of
person-deictic indicators, headed by the fancied I/you ‘‘correlates,’’ renders the monopoly given in theory to one sharer all the less tenable, visibly so. Why ignore private
expression, if copronominalized for role-playing? And with the fallacy overcome, the
difference made by (un)self-consciousness to the roles (from their obligatory number
upward, as just argued) springs into higher relief against the surface pronominal
commonalities.
Now, within social language itself, the same formalist reflex/role conflation also
breeds Benveniste’s impossible ‘‘(hi)story’’ – purely objective, tellerless, noncommunicative – to invalidate the dichotomy with so-called ‘‘discourse.’’ His otherwise
mandatory correlates I/you (here, referred to narrator/narratee) abruptly turn contingent in, or on, the language of narrative: present in ‘‘discoursed,’’ absent in ‘‘(hi)storied,’’ events. Yet the incongruous genre-specific binarism just splits the theory’s
overall package dealing, and so founders in turn, now on public realities. Briefly,
suppose a historical account dispensed with any visible ‘‘correlation of personality,’’
how would it follow that events would then unroll ‘‘without any intervention of the
speaker in the narration’’ (Benveniste 1971: 206), ‘‘so that there is then no longer even
a narrator’’ (p. 208)? How, in reason, would events ‘‘narrate themselves’’ and to whom
but a narratee?
Self-consciousness as a Narrative Feature
243
Ironically, these rhetorical questions become genuine only with the polar change of
rationale to our secret life. There alone can (e.g., perceived) events find (in the mind’s
unreflective or camera eye) an image both unknowing and unknown: a limit of unselfconsciousness imageable in turn, though never reachable, via story’s mimesis of that
world-reflecting yet unreflective mind as an inset advisedly quoted between discoursive knowers.
Even so, the rationales polarize by awareness of self and other in event-imaging, not
by person-deixis. The most impersonal-looking history, with no I’s and no you’s, like
the Bible from Genesis to Kings, still entails, if public, both a speaker and an
addressee, and may indeed pronominalize their roles at any turn. Just as the surfacing
of the pronouns does not establish interpersonal contact, so their burial never
precludes it: the manifest/latent variants rather enrich the cross-typology of selfconsciousness and transmission or, regarding ‘‘history,’’ narration itself, within (narrative) discourse.
Echoes and parallels and follow-ups to Benveniste abound in narratology: little
wonder, since even the more critical references (e.g., Culler 1975: 197–200) generally
fail to target and redress the discoursive imbalance. The field, instead, is torn between
comparable alternative extremes (often – on which later – between theory and practice
as well). Apropos public storytelling itself, witness the throwback to pseudo-Jamesian
immediacy of transmission, under fresh disciplinary auspices. Benveniste’s reifying of
surface deixis has revived, or reinforced, the modernist belief in the logical monster
of ‘‘history’’-like, narratorless linguistic narratives – and within written fiction at that.
This revival has not passed unopposed, of course, especially when putative addresseelessness in these narratives (Banfield 1982) followed suit. But then, among the very
opponents, the genre is alternatively theorized by its public discourse, correlated roles
and all. Less a mistheorizing, this, than an undertheorizing, deficient in scope and
power rather than downright illogical.9 Unlike mine, the sundry communication
models of narrative are invariably symmetrical throughout: real author corresponds to
real reader, implied author to implied reader, narrator to narratee, dialogist – if
mentioned – to interlocutor. So, instead of replacing Benveniste’s illusory noncommunicative form (‘‘history,’’ with precursors and offshoots) by the genuine, secret-life
article, they reject it only to reimpose, actually multiply his speech partnership
between exterior subjects. The transmitter/receiver symmetry all along the chain
officially neither allows nor accounts for the varieties of asymmetrical, egocentric,
receiverless, unself-conscious discourse incorporated in narrative – not even those
somehow cropping up in the narratology.
It’s like omitting motive from the action chain. How, where, why do sheer
informants enter, and at times apparently replace, the official line of communicators?
Or how do so-called focalizations (especially if internal, or if coming and going) enter
the flow of narration? Once the question arises, the answer enforces itself to reconceptualize the whole narrative model. They enter, as explained, the way all intermediate transmitters do and must, the only possible way: via quotation by the
anterior higher frame, with all the built-in choices and montages and subordinacies
244
Meir Sternberg
available to it, except that the inset concerned quotes an informant, re-presents an
originally private single-party discourse event.
At the receiving end, likewise, standard typologies of reader, audience, narratee
omit to mark the polar extreme of absence (i.e., addresseelessness) for two-way
comparison. To enlarge and refine the comparison, as already noted apropos Benveniste, an inclusive typology will also embody the modulations within unselfconsciousness itself toward self-discourse proper: from ‘‘no addressee’’ to ‘‘no outside
addressee,’’ to private dialogue, to the surfacing, even alternating of ‘‘you’’ forms,
always self-made.
Furthermore, such reconceptualizing can alone accommodate the intermediate,
mixed variants of narrative awareness that apply to both transmitter and receiver,
often in significant contrast to their absolutely or comparatively (dis)privileged
mates. Otherwise, at discourse’s either end, the gradability of viewpoints (hence of
contrast) that we outlined in between the poles will then also escape notice, or at least
systematic reference to this feature. Nabokov outreaches the telling Humbert as well
as opposes the experiencing Humbert in self-consciousness, while his reader enjoys
more protection against novelistic ironies than any dramatized addressee within the
novel’s world.
Again, the wanted rethinking in the light of this feature, with new gains in
coverage, discrimination, and explanatory power at once, stretches to the very what’s
of quoted privacy. Theories of reported discourse have long opposed speech to thought
as objects of report, subjective objects in effect, but without due attention to the
reported vis-à-vis the reporting subject’s self-consciousness. From life to narrative,
from inset to frame, the divide between vocal speech and mental thought does not
entail that between social and secret re(-)presentation, as perhaps appears, but flexibly
correlates with it. Crosses and compounds result and alternate, as do the two obvious
pairings, all enchainable in sequence.
Thus Elizabeth Bennet, holding a dialogue framed within Jane Austen’s large
novelistic dialogue, indeed polarizes with Molly Bloom in silent thought vis-à-vis
Joyce’s communication. Clean against such automatic fourfold sorting, however,
Elizabeth in equally vocal self-address (e.g., in exclamatory response to Darcy’s letter)
polarizes with herself as the Austen-like dialogist, and groups instead with Molly the
silent monologist reported to us in Joyce’s novel. The differences in transmitted
expression – from world-making to ordering to wording – regroup to suit: privacy,
as always, motivates discoursive egocentricity. Further, the very reporter may turn
unself-conscious, even interior: we witness those heroines looking back or forward in
private speech or thought to speech or thought – or for that matter, writing – with a
freedom denied to their inventors and higher-level reporters, never mind their own
public selves. Apart from signifying along the traditional outer/inner line, our feature
multiply cuts across it to diversify the genre’s resources and record.
So does another, still more principled, breach of that line. It is time to appreciate
the uniqueness of the hearing act, first-order or reported, as a cross between speech
and thought: an operation on the former public utterance in the latter’s privacy. Given
Self-consciousness as a Narrative Feature
245
a reported encounter, between every two turns of dialogue there lurks a pocket of
monologue – usually gapped, elsewhere surfacing in ‘‘x heard that’’ form, but always
an eventful subjective nexus. In that pocket, the hearer (mis)perceives and (mis)interprets the foregoing utterance with a view to a vocal response, one possibly,
sometimes advisedly, divergent from what has just been heard, let alone spoken. As
an emplotted inside view, hearing within an interpersonal speech-event thus corresponds to motive between two physical events.
Much the same holds for every two letters exchanged in a correspondence, and, on a
wider scale, for every dramatized narratee, whether responsive or formally reticent. All
along, then, what I call ‘‘the world from the addressee’s viewpoint’’ (Sternberg 1986)
extends the range of unself-consciousness opened by addresseeless discourse to the
very heart of self-consciousness territory. Just like goal-directed action in general,
dramatized communication itself subsumes, provokes, and enacts reflection, inset
telling multiplies informants of its own. Another revealing measure, this, of the
power of a unified theory to encompass, indeed to enrich the narrative field, while
both integrating and differentiating the parts in systematic operational terms: largely
by appeal to our feature-cum-force.
Among further obstacles to such a versatile approach, another array runs through two
books contemporaneous with mine (Sternberg 1978), and apparently informed for a
change by a distinction analogous to self-consciousness. In Franz Stanzel’s A Theory of
Narrative ([1979] 1984), the first constitutive element of narrative situations is
‘‘mode,’’ opposing ‘‘narrator’’ (Cid Hamete in Don Quixote, Zeitblom in Doctor Faustus)
to ‘‘reflector’’ (Strether, Molly). The Jamesian terminology augurs well for its theorizing. Dorrit Cohn’s more specialized Transparent Minds (1978) traces the narrative
rendering of ‘‘consciousness,’’ as against outer affairs and accounts. Either work also
rightly criticizes Booth’s grouping of minds with mouths under narratorhood. Between them, however, the promised distinction gets so mishandled as to confuse the
issue and its wider implications in various respects: some peculiar, some already more
familiar, some newly typical and beyond German scholarship. (Thus Genettian Mood/
Voice, focalizing/narrating.) Sorting out this tangle will redraw, even retest, the
difference the negative way, on a broad front. Especially, (un)self-consciousness will
prove resistant to marriage, a fortiori to interchange, with other narrative attributes; a
fortiori with its own binary; a fortiori where such attempts on it either load its typology
or force some type(s) into or out of mediacy, even narrativity, regardless.
This ascendant order of distinctiveness gains relief, point by point, from the opposite
analyses cited. Regarding the customary narratological split, both analysts favor
transmission over action. Stanzel even defines narrative by its (undefined) ‘‘mediacy of
presentation’’ alone (1984: 4ff.). Within the favored half, therefore, you would at least
expect a sharp polarizing of ‘‘teller’’ and ‘‘reflector’’ modes, to suit with Jamesian usage:
A teller-character narrates, records, informs, writes letters, includes documents, cites
reliable informants, refers to his own narration, addresses the reader, comments on that
246
Meir Sternberg
which has been narrated . . . By contrast, a reflector-character reflects, that is, he mirrors
events of the outer world in his consciousness, perceives, feels, registers, but always
silently, because he never ‘‘narrates,’’ that is, he does not verbalize his perceptions,
thoughts and feelings in an attempt to communicate them.
(Stanzel 1984: 144)
James would hardly approve. Far from being criterial – necessary or sufficient – the
two bundles of traits put forward are mostly sharable by the alleged binaries. Thus,
except for the circular or tautological ‘‘narrates, records, informs’’ and the equivocal
‘‘addresses the reader’’ (overtly?), none of the activities associated here with the ‘‘teller’’
(letter-writing, documenting, quoting, self-reference and self-commentary) is in fact
obligatory, much less as a set. Or else Hamsun’s Hunger, or Joyce’s Portrait, remains
untold. Nor does reflection, least of all written, Pepys style, bar any of these discourse
activities.
Inversely, a reflector like Strether indeed ‘‘mirrors . . . perceives, feels, registers,’’ but
will the same necessarily apply to his vocal or journal-writing counterparts? Or to the
(day)dreamer? Mirroring, perception, emotion, registration itself, anything beyond
private world-making, are all dispensable to the reflector mode at large. (‘‘Informant’’
is duly noncommittal.) And even where vessel or soliloquist or diarist indulges in all
these activities, doesn’t the same mental repertoire extend to the narrator proper (as,
elsewhere, to the crossmodal focalizer)?
Hence also the falsity of the categorical ‘‘always silently,’’ imposed on the reflector’s
discourse as such, yet again geared to the nonspeaking, nonwriting variety alone. But
then, ‘‘does not verbalize’’ is hardly the law of even this reflectorial variety – Strether,
Molly, Faulkner’s Benjy – which famously ranges over the axis of mute self-expression,
as between word and image.
So the one relevant distinction (‘‘attempt to communicate’’ present or absent) gets
lost in the mishmash on and between the modal flanks. No wonder taxonomic
incongruities abound. Beckett’s Malone swings from ‘‘interior monologue’’ to ‘‘firstperson narrator’’ and back to ‘‘reflector’’ (Stanzel 1984: 61, 94, 145), then ‘‘oscillates’’
between the roles (pp. 150, 226). Across the same barrier, ‘‘epistolary and diary
novels’’ are lumped together, the second being mediated by a ‘‘teller’’ amid selfaddress (pp. 202, 211, 225–6). Inversely, ‘‘the dramatic monologue is non-narrative’’
(pp. 225–6), though both addrecentric and actional by definition. The nonnarrativity
verdict is even generalized for all direct quoting, as if ‘‘first person narrative’’ itself
didn’t manifest this form in large. Wavering, contradictory bracketing, arbitrary
exclusion from the genre: a signal negative lesson. Together with the overemphasis
on interiority – abetted by the very term reflector – it highlights both the cutting
edge and the scope of our (un)self-consciousness model.
If this lust for feature-clustering already defeats itself, worse ensues. The uttermost
package-dealer in narratology – witness his entire ‘‘typological circle’’ – Stanzel
proceeds to conflate the two modes with an assortment of further binarisms, mainly
presentational and all in truth independent, crosspolar, reversible. Narrator vs.
reflector allegedly goes with:
Self-consciousness as a Narrative Feature
247
1 Telling vs. showing;
2 Commentary present vs. absent;
3 Account explicit, generalized, and purportedly complete vs. implicit, specific and
conspicuously incomplete;
4 Indeterminacy in communication vs. built into the human condition;
5 Preliminaries first vs. none;
6 Beginning with identifiable ‘‘I’’ vs. referentless ‘‘he’’;
7 Reference to hero by name vs. opaque pronoun;
8 Stable vs. open ending;
9 Unreliability (if character) definitional vs. irrelevant;
10 Past tense having vs. shedding past meaning; and so forth
(Stanzel 1984: 146–70).
Never mind consistency, or the persistent bias for interior unself-consciousness,
or the bizarre idea of reliability: the packaging of these 10 extra oppositions with
the modal one at issue is again all demonstrably wrong in reason and fact alike. So
indeed my argument already predicts about the lot, or has even demonstrated
regarding the majority, concerned in effect with temporal structure between lucidity
and ambiguity. Just recall how omnicompetent narrators, Fielding’s suppressor
or Trollope’s all-communicator, meet their operational equivalents among unprivileged informants, dechronologizing or chronologizing, with a view to either basic
time experience. Poles apart in aesthetic/mimetic motivation, crucially hinging
on self-consciousness, the respective transmitters yet co-implement the desired
authorial strategy via the means appropriate to it: Stanzel’s assorted opposites,
inter alia.
Worst of all, these false concomitants in Stanzel not only overload and defocus
the pertinent feature, to the loss of cutting edge, but may even substitute for it to
metamorphic effect both ways: either polar discourser transforms into the other,
regardless of self-consciousness itself. And this beyond the plainer, if symptomatic,
oversights that befall Malone or the diarist, the dramatic monologist, the direct
quoter. Incredibly, the theory boasts the logical (typological, a fortiori teleological)
monster of ‘‘reflectorization,’’ whereby the ‘‘teller temporarily assumes certain characteristic qualities of a reflector-character’’ (Stanzel 1984: 173).
Even an authorial narrator will become ‘‘reflectorized’’ by dint of what is actually
nothing but intermittent, ever-eligible, at most reportive shifts in communicational
tactics or posture or distance. If we look back to the gratuitous tenfold opposition
above, these reflectorizers involve a switch of discourse from left to right column - all
amid ongoing self-consciousness, clean forgotten by the taxonomist (Stanzel 1984:
156–200 passim). So much so that reflectorizing, down to free association, overtakes
the very highest communicator, and nowhere else than in Ulysses, taken as ‘‘the
interior monologue of the author’’ at ‘‘composition’’ (pp. 178–80).
Author apart, were such ‘‘transformation’’ conceptualized as devolving from
omniscient to humanly fallible narrator, it would at least avoid sheer unreason, if
248
Meir Sternberg
only inconsistency with the distinctive feature of their shared narratorhood vis-à-vis
reflectorhood. So, late in Nabokov’s Pnin, the impersonal teller emerges as the
hero’s earthly opponent, while remaining self-conscious, unreflectorized. Again,
were the ‘‘transformation’’ described as a narrator fronting and framing some reflector,
instead of ‘‘temporarily becoming’’ one, it might even fit the case here and there,
even if nothing like news, let alone a general explanatory rule: not even for the
listed discourse breaches. Transmissional motivating options would still extend
from, say, bare narratorial games of withholding or code-switching to inset reflectorial
peculiarities of style, judgment, knowledge, beholding, thinking, hearing. So Fielding’s narrator pretends human ignorance at will, and James’s systematically
limits himself to rendering the vessel’s human outlook – limited and secret by nature
– without any metamorphosis in either teller’s absolute powers, whether omniscience,
however suppressive, or audience-mindedness, however disguised. Nor does such
metamorphosis attend even an unexpected overall shift of motivation: from Wonderland told to Wonderland dreamed, with the teller removed intact, frameward, and the
tale alone, if you will, reflectorized in the dreaming, via Alice.
As it is, instead, Stanzel’s transmitter abruptly changes poles, complete with
realm of existence, across a qualitative barrier impassable on his own premises.
With self-consciousness forgotten in favor of its imaginary correlates, the artistically
motivated devices of the most author-like of tellers (and, in Ulysses, also the most
knowing of authors), at will operating under subjective guise, veer round into
the perspectival lapses and liberties betrayed by the most creature-like, disempowered, egocentric of reflectors. Discourse surface gets mistaken for discoursive reality,
contrivance for self-exposure, performance for (in)competence.
Similarly with ‘‘first-person narrative.’’ Changes in pronominal, I/he self-reference
(though, Benveniste fashion, never affecting speech role), or in distance from
the experiencing self, will reflectorize it (Stanzel 1984: 104, 208–9). Conversely,
the interior monologue turns ‘‘narrative’’ – though ‘‘of course’’ still unaddressed to any
listener – where its reflector changes tenses from present to past (pp. 207–8, 227).
Reminiscing monologists or soliloquists, most nondirectly quoted vessels, and diarists at large would thus collocate, by force of tense, with the standard retrospective
narrator.
Either way, the polar modes transform into each other on the strength of
their noncriterial features – if theirs at all - against the one that makes the difference:
self-consciousness present or absent. It is like saying that, as a woman eats and
plays and menstruates and reasons, an eating or playful or reasoning man becomes
womanized.
So the attack on Booth’s modal indifference boomerangs with a vengeance. Yet
the narrator/reflector miscorrelations and the reflectorizing illogic circulate nevertheless, even in the better hands of Cohn (1981: 170–4) or Fludernik (1996). The
allure of package-dealing dies hard, the oblivion to self-consciousness comes easy,
given the ongoing taxonomic spirit unmindful of the Proteus Principle: the
living interplay of features with forces, re(-)presentational forms with communicative
Self-consciousness as a Narrative Feature
249
functions, that makes and make sense of discourse as transacted, varieties included.
If teller and reflector can interchange as alternative perspectival means to a narrative
end – mediators, hence motivators, of generic (dis)ordering, above all – they
can neither stand opposed nor change to each other by force of any adjoinable
superficiality.
Indeed, as well as endorsing Stanzel’s typological fallacies here, Cohn (1978)
outspecifies them in her reflector-focused study, otherwise the best of its kind. Just
observe the tendentious packaging drive that runs, on various scales, through the
book’s Part II to overdiscriminate among narrative forms, all in honor of one idealized
private mode. Common options specialize, harden, and bunch into constants, may’s
into pseudo-differential must’s, some already encountered.
Thus, the monologist vs. first-person narrator divide is reyoked with variables
beyond communicativeness, especially present vs. past tense and anachrony vs. chronology. On the monologic flank, in turn, such overlinkage apparently rationalizes
Cohn’s (likewise standard, here titular) bias for unvoiced self-discourse, as against
written journal and spoken thought. Besides their alleged past-reference, these
inferiors willy-nilly gather further narration-like markers, or forfeit their ‘‘properly’’
monologic opposites: randomness, emotivity, ellipsis, pronominal opacity. All, again,
counter to the expressive freedom shared perforce by all egocentrics. Last, within
unvoiced self-discourse itself, her specialty, Cohn even outdoes Stanzel. She invents
otherwise useful types like the orderly retrospective ‘‘autobiographical monologue’’
(Dostoevsky’s ‘‘A Gentle Creature’’) and the disorderly, free-associational yet still pastoriented ‘‘memory monologue’’ (The Sound and the Fury), not as variants of, or at least
parallels to, but as variants from ‘‘genuine interior monologue.’’ Note the essentialist
rhetoric enlisted in ‘‘genuine’’ to valorize and purify, hence reify, the latter monologue
form – exclusive of its nearest equivalents in unself-consciousness, plus inner speech,
plus work-length ‘‘autonomy,’’ plus temporality.
Monologism, then, progressively reduces to privileged contingencies. So, by
the end of the road, at a prohibitive cost rising with each fork, the monologic –
instead of encompassing the secret life, or even the hushed kind, or the directly
quoted subkind thereof, or the extended subsubkind – finds its rare quintessence in
Molly Bloom, as desired. No interpretive finesse can save this web of fissures and
(con)fusions by fiat, either; nothing short of the reorientation outlined, via (un)selfconsciousness, will do justice to the private domain as such, and to the transparent
mind within it.
The less so because the macromonologue loses in Cohn its generic tellerliness
(reminiscent of modernism’s ‘‘dramatized’’ tale, or Benveniste’s ‘‘history,’’ or structuralism’s ‘‘reproductive’’ quoting below) along with its informant-specific unity.
Even the disfavored types, once grown text-length, count here as equally ‘‘autonomous’’: immediate, single-voiced, directly quoted yet narrator-free to the point of
‘‘antinarrative,’’ sui generis. This intersects afresh with Stanzel, who generalizes the
vanished mediacy of narration in narrative for all direct citation of speech and
thought: passing echo, scene, text, interior and dramatic monologue. (Compare
Meir Sternberg
250
Narrative as action-representing communication
Author
Authorized narrator
Pure aesthetic
(metanarrative) motivation
Mimetic (intermediate)
motivation
Existential (objective, representational, well-told)
Perspectival (subjective, re-presentational, framed)
Self-conscious (second-order quoted narrator)
Unself-conscious (quoted informant)
Mental
Vocal
Written
Second-order narratee
Narratee
Reader
Figure 15.1 Narrative as action-representing communication
Genette denarrativizing such formal directness into ‘‘pure mimesis’’ or ‘‘immediate
discourse,’’ again literally and hermetically reproductive of the original. Elsewhere,
the narratorless purity has long subsumed free indirectness and the ‘‘third-person’’
Jamesian model.) Inconsistently selective to boot, Cohn and her followers would keep
the vanishing narrative act for ‘‘autonomous’’ quoted self-discourse, as though sheer
extent made all the difference in transmissional, hence generic, kind.
So the contradiction in terms only sharpens, along with the counterevidence.
Logically, direct quotation entails a quoter, regardless; and the quoter must be
acknowledged to re(con)textualize, reperspectivize, subordinate the original utterance
or thought in the framing, however quietly, on pain of what I call the direct speech
fallacy. Indeed, how else, if not through the narrator turned quoter, do we gain access
to the monologist’s consciousness, deemed here ‘‘the special preserve of narrative’’ visà-vis ‘‘dramatic and cinematic’’ fictions (Cohn 1978: vi, 5) at that? Again, who else
decides where to begin, or end, re-presenting the secret flux? And if you would
Self-consciousness as a Narrative Feature
251
attribute both access and cutoffs to the author instead – already in breach of putative
autonomous monologic univocality – there still remain in the inset marks of framing
textual (rather than arguably contextual) interference that are definitionally given to a
performing narrator alone. Who else verbalizes nonverbal mentation? Who segments
the interior speech or thought continuum with written aids to understanding like
chapters, paragraphs, sentences, punctuation? Who introduces and deciphers and
annotates (let alone expurgates) Pepys’s self-writing, or, in effect, the episode-length
Molly paradigm itself set within the Ulysses whole?
Here, therefore, incurring the direct speech fallacy means that the unself-conscious
mind somehow, by wondrous default, quotes, inscribes, edits, publishes, as well as
expresses itself; the self-centered reflector also plays self-teller to us outsiders. By this
self-contradiction, moreover, the theories (Genette included) divide against themselves, joining their very targets of attack: from Booth et al.’s insideviewee-as-narrator
to Banfield et al.’s narratorless subjectivity.
Inset experience, like all re(-)presentation, never displaces but unknowingly
mediates the framing communication on every axis. Taken together, the consequences,
for good or for ill, are radical. To dovetail my argument’s constructive and critical
thrusts, Fig. 15.1 will serve. (You’ll recall that the generic branching increasingly
enables in practice crosses, shifts, enchainments, further ramifications, unrealized
possibilities.)
NOTES
1 In Booth (1961: 155), it thus depends on the
narrator’s awareness of composing a literary
work and reference to the problems involved:
these stipulations exclude Huck Finn and the
barber in Lardner’s ‘‘Haircut,’’ both selfconscious, as addressors, within my framework. Likewise restrictive is the common
phrase ‘‘self-conscious artistry,’’ echoed in,
e.g., Cohn (1978: 152, 180, 187, 199, 265).
2 Sternberg (1978: 246–305, 1983a: esp.
172–86; also, e.g., 1986, 1991, 2001: esp.
167ff.), on how the same functional, holistic
(re)construction produces in little the quoting
forms of speech and thought, or indeed speech
vs. thought.
3 Narrative (in)competence defines itself relative
to the author – across all exterior value judgments – as its linguistic parallel does to the
relevant speech-community.
4 Accordingly, below we will find Stanzel packaging the reflector, and Cohn the interior
monologist, with anachronic deployment, a
miscorrelation that goes as far back as Edouard
Dujardin.
5 Again, these two logics govern all representation – across kinds, arts, media, ontologies,
truth claims – but we’ll proceed with the exemplary, because most protean, case of narrative
fiction.
6 Never defined, often unstable, where most invoked: in Lubbock, Stanzel, Genette and company.
7 Interestingly, Ulysses imitates Pepys’s style in
Oxen of the Sun.
8 Details in Sternberg (1978: 276ff.), and below.
9 Illogic ensues, as will appear, when certain
discourse types, often private ones (above all,
interior monologue) get officially excepted
from narrativity.
252
Meir Sternberg
REFERENCES
AND
Banfield, A (1982). Unspeakable Sentences. Boston:
Routledge.
Benveniste, E. (1971). Problems in General Linguistics. Coral Gables: University of Miami Press.
Booth, W. C. (1961). The Rhetoric of Fiction. Chicago: University of Chicago Press.
Bordwell, D. (1985). Narration in the Fiction Film.
Madison: University of Wisconsin Press.
Cohn, D. (1978). Transparent Minds: Narrative
Modes for Presenting Consciousness in Fiction.
Princeton, NJ: Princeton University Press.
Cohn, D. (1981). ‘‘The Encirclement of Narrative:
On Franz Stanzel’s Theorie des Erzählens.’’ Poetics
Today 2, 157–82
Cohn, D. (2000). The Distinction of Fiction. Baltimore: Johns Hopkins University Press.
Culler, J. (1975). Structuralist Poetics. London:
Routledge.
Fludernik, M. (1996). Towards a ‘‘Natural’’ Narratology. London: Routledge.
Forster, E. M. (1962). Aspects of the Novel. Harmondsworth, UK: Penguin.
Ryan, M.-L. (1991). Possible Worlds, Artificial
Intelligence, Narrative Theory. Bloomington:
Indiana University Press.
Stanzel, F. K. (1984). A Theory of Narrative. Cambridge, UK: Cambridge University Press.
Sternberg, M. (1978). Expositional Modes and Temporal Ordering in Fiction. Baltimore: Johns Hopkins University Press.
FURTHER READING
Sternberg, M. (1983a). ‘‘Mimesis and Motivation:
The Two Faces of Fictional Coherence.’’ In
J. Strelka (ed.), Literary Criticism and Philosophy
(pp. 145–88). University Park: Pennsylvania
State University Press.
Sternberg, M. (1983b). ‘‘Deictic Sequence: World,
Language and Convention.’’ In G. Rauh (ed.),
Essays on Deixis (277–316). Tübingen: Gunter
Narr.
Sternberg, M. (1986). ‘‘The World from the
Addressee’s Viewpoint: Reception as Representation, Dialogue as Monologue.’’ Style 20,
295–318.
Sternberg, M. (1991). ‘‘How Indirect Discourse
Means: Syntax, Semantics, Pragmatics, Poetics.’’ In R. Sell (ed.), Literary Pragmatics
(pp. 62–93). London: Routledge.
Sternberg, M. (2001). ‘‘Factives and Perspectives:
Making Sense of Presupposition as Exemplary
Inference.’’ Poetics Today 22, 129–244.
Thompson, K. (1988). Breaking the Glass Armor:
Neoformalist Film Analysis. Princeton, NJ:
Princeton University Press.
Yacobi, T. (1981). ‘‘Fictional Reliability as a Communicative Problem.’’ Poetics Today 2, 113–26.
Yacobi, T. (2000). ‘‘Interart Narrative: (Un)Reliability and Ekphrasis.’’ Poetics Today 21: 708–47.
Yacobi, T. (2001). ‘‘Package-Deals in Fictional
Narrative: The Case of the Narrator’s (Un)Reliability.’’ Narrative 9, 223–29.
16
Effects of Sequence, Embedding,
and Ekphrasis in Poe’s ‘‘The Oval
Portrait’’
Emma Kafalenos
Edgar Allan Poe’s initial consideration when beginning to write, he says in ‘‘The
Philosophy of Composition,’’ is the effect, or impression, he wants the story or poem
to have on its readers. Only after he has chosen the effect, he continues, does he
‘‘consider whether it can best be wrought by incident or tone . . . afterward looking
about me (or rather within) for such combinations of event, or tone, as shall best aid
me in the construction of the effect’’ (Poe 1977: 550). ‘‘The Philosophy of Composition,’’ published in 1846, is the essay in which Poe offers a description of how he
wrote ‘‘The Raven.’’ For that poem, Poe says, he chose ‘‘a novel, first, and secondly a
vivid effect’’ (1977: 550). Poe’s stories, too, often produce vivid and novel effects, and,
like ‘‘The Raven,’’ may have been written to do so.
The strong effect that Poe’s stories and poems produce is often, in part, the result of
supernatural elements. But even in his stories that include the supernatural, the effect
that the supernatural produces is often heightened by narrative strategies that, unlike
the supernatural, can be explained through analysis. Poe’s story ‘‘The Oval Portrait,’’
published in 1842, takes its title from a portrait that in the story is represented
through ekphrasis, the rerepresentation in words of a visual representation. Poe’s use
of ekphrasis draws attention to the power of language to guide interpretations by
what is omitted as well as by what is said. The embedding in the story of both the
oval portrait and a document that purportedly describes how it was made demonstrates the power of narratives to shape interpretations of causality by controlling the
sequence in which information is revealed.
‘‘The Oval Portrait’’ is very short – a mere six paragraphs. For the purposes of the
present analysis I think of the story as made up of three sections of about the same
length: the first two paragraphs, the next three paragraphs, and the final long
paragraph. In the first of the three sections an intradiegetic narrator introduces
himself and his situation. In the second section the narrator recounts his experience
254
Emma Kafalenos
viewing the oval portrait. The final section is a quotation, set off by quotation marks
at the beginning and the end, of a paragraph describing how the portrait was made,
which the narrator reports that he is about to read. Thus readers read the quoted
paragraph as if at the same time that the narrator does, and with knowledge of the
narrator’s situation and of the experience that has led him to pick up the volume and
turn to this specific passage. I begin my analysis by looking at this quoted paragraph,
to consider how we would interpret the causal relations among the events it reports if
we encountered the paragraph as an independent text, complete in itself. At the end of
this essay I will compare this interpretation with our interpretation of exactly the
same text when we read it as an embedded segment in Poe’s story. The difference in
the two interpretations illustrates the degree to which narrative strategies can guide
readers’ interpretations of an event’s causes and effects.
To be able to talk about and compare interpretations, I draw upon a vocabulary
I have developed, primarily from work by Tzvetan Todorov (1969a, 1969b) and
Vladimir Propp (1968), for naming interpretations of causal relations. The set of 10
functions that I have defined names stages in a causal sequence that leads from one
equilibrium to another equilibrium. Readers can use these names to express their
interpretations of causal relations in a story or in part of a story.
I list here the five functions that provide the necessary vocabulary to formulate and
compare interpretations of the quoted paragraph at the end of Poe’s story, when read
by itself and when read in the context of the complete story. These five functions
represent stages from the disruption of an equilibrium to the establishment of a new
equilibrium:1
Five functions, with definitions
EQUILIBRIUM
A
destabilizing event that disrupts an equilibrium
C
decision by C-actant (the character who performs function C) to
attempt to alleviate the function-A event
C0
C-actant’s initial act to alleviate the function-A event
H
C-actant’s primary activity to alleviate the function-A event
I or Ineg
success (or failure) of the function-H activity
EQUILIBRIUM
Turning to the quoted paragraph at the end of Poe’s story and reading it by itself as if
it stood alone, most readers would probably interpret the initial situation as a
moment of equilibrium. The first sentence reads: ‘‘She was a maiden of rarest beauty,
and not more lovely than full of glee’’ (1977: 105). Because this first sentence
describes an ongoing situation in which unsurpassable beauty and pleasure coexist,
readers are guided to interpret that all is well in the narrative world – that the
narrative world is in a state of equilibrium. As we read on, because the first character
to whom we are introduced is the girl, we are guided by the primacy effect to
interpret what follows in relation to how it affects her and her situation.2
Effects of Sequence, Embedding, and Ekphrasis
255
For many readers, the combination of superlative beauty and pleasure that the
first sentence describes will suggest that something disruptive is about to occur.
In fact, the second sentence immediately introduces a destabilizing set of events
that disrupts that equilibrium: ‘‘And evil was the hour when she saw, and loved,
and wedded the painter.’’ Soon the ‘‘evil’’ is specified; the painter has ‘‘already a
bride in his Art’’ (1977: 105). My interpretation – and, I suggest, most readers’
interpretation – of the causal relations among the reported events can be represented
by these functions:
Analysis 1 (quoted paragraph, first three sentences)
EQ maiden is full of glee
A
she marries a painter who is already married to his art
Once we have understood that it is the painter’s devotion to his art that disturbs the
girl’s otherwise unqualified happiness, probably most of us read on wondering how, or
whether, the painter will reassure his wife that she is as important to him as his art is.
But as the story continues we learn that she ‘‘dread[s] only the pallet and brushes and
other untoward instruments which deprived her of the countenance of her lover. It
was thus a terrible thing for this lady to hear the painter speak of his desire to portray
even his young bride’’ (1977: 105). Because the girl is made so unhappy by the
painter’s desire to paint her, most readers probably interpret his request as a second
function-A disruptive event:
Analysis 2 (quoted paragraph, first four sentences)
EQ maiden is full of glee
A1 she marries a painter who is already married to his art
A2 painter requests that the girl sit for a portrait (for the girl, his request is ‘‘a terrible
thing . . . to hear’’)
Moreover, as the weeks pass during which the girl sits for her portrait, she grows
‘‘daily more dispirited and weak’’ (1977: 106), a third function-A event. And then,
although some who saw the unfinished portrait ‘‘spoke of its resemblance in low
words, as of a mighty marvel,’’ the painter ‘‘would not see that the tints which
he spread upon the canvas were drawn from the cheeks of her who sate beside
him’’ (p. 106, Poe’s italics), a fourth event that disrupts the initial equilibrium.
Probably most readers continue to think about how, and by whose efforts, the
disruptive situation can be resolved. We wonder whether there isn’t someone who
can point out to the painter how damaging painting her portrait is to the girl he is
painting.
But then at last the painter applies the final brushstrokes. The quoted paragraph –
and Poe’s story – conclude:
. . . and, for one moment, the painter stood entranced before the work which he had
wrought; but in the next, while he yet gazed, he grew tremulous and very pallid, and
256
Emma Kafalenos
aghast, and crying with a loud voice, ‘‘This is indeed Life itself!’’ turned suddenly to
regard his beloved: – She was dead!
(1977: 106, Poe’s italics)
Reading the quoted paragraph in its entirety and considering it as an entity complete
in itself, without reference to the story in which it is embedded, I interpret (and
assume that most readers interpret) the causal relations among the reported events
according to these functions:
Analysis 3 (quoted paragraph, read as an entity complete in itself )
EQ maiden is full of glee
A1 she marries a painter who is already married to his art
A2 painter requests that the girl sit for a portrait (for the girl, his request is ‘‘a terrible
thing . . . to hear’’)
A3 the girl grows dispirited and weak
A4 painter won’t recognize that he is killing the girl by painting her
A5 upon completion of the painting the girl is dead
In the quoted paragraph an initial equilibrium is followed by one event after another
that disrupts the girl’s previously happy life. These disruptive events destabilize the
situation and culminate in the death of the girl at the very moment that the painting
is completed. The girl’s death is the awful final event, the event we learn about even
after we learn that the painting has been successfully completed. We interpret the
death of the girl as the concluding tragic event in a series of disruptive events. The
quoted paragraph offers an example of a narrative that traces only a part of a complete
sequence – in this case the move from an equilibrium to a disruptive situation, a
pattern that is relatively uncommon in narratives and that most readers probably find
dissatisfying.
Narratives guide readers’ interpretations of the function of events in a number of
ways, including ways that make even tragic concluding events seem less thoroughly
disruptive than the quoted paragraph makes them seem. One way is to encourage
readers to align their interpretations of the function of events with the interpretations
of one character rather than another, for instance, by providing information from the
perspective of one character rather than another. In the quoted paragraph we are given
information about how the girl responds – to seeing that her husband is already
married to his art, to being asked to sit for her portrait, to being painted. As a result,
readers tend to interpret the function of events according to their effect on the girl,
that is, as a series of function-A occurrences.
These same events that the quoted paragraph reports could have been told instead
from the perspective of the painter. It is much more common, in fact, in narratives in
which one character accomplishes something and a character that that character cares
about dies, that the story be told in a way that guides readers to interpret the function
of events in accordance with the character who accomplishes something (in this case,
the painter who paints a marvelous painting) rather than, or as much as, with the
Effects of Sequence, Embedding, and Ekphrasis
257
character who dies (in this case, the girl). We can imagine the events that the quoted
paragraph reports told in a way that would guide readers to admire the painter’s
accomplishment and, yes, to recognize the tragedy of the girl’s death, but at the same
time to appreciate the complexity of the concluding situation in which the painter
recognizes simultaneously his success as a painter and his failure as a husband. Readers
might interpret the conclusion of this imagined story as offering a balance between
the two simultaneous final events: the function-A death of the girl and the function-I
successful completion of the painting.
Analysis 4: (imagined retelling of the events that the quoted paragraph reports)
A
painter desires to portray his bride
C
painter decides to ask her to sit (the painter is the C-actant – the character who
performs function C)
C0
painter asks her to sit
H
painter paints her portrait
I
painter completes the painting, but
A
the girl is dead
Another way that narratives guide readers’ interpretations of the situation at
the end of a story is by appending a concluding comment by a character or the
narrator. In some cases these comments reveal the character’s or the narrator’s interpretation of what has occurred; in other cases such comments may merely indicate
that life in the narrative world is continuing. The quoted paragraph offers an example
of a narrative in which the events that have occurred make the resolution of
the motivating function-A situation impossible. After the death of the girl there is
no way to bring her back to life and make her happy again. In this situation, in some
narratives, a character or a narrator will say something that guides readers to recognize
that the narrative world is again in equilibrium. At the end of the quoted paragraph,
for instance, one further sentence could tell us that the painter regretted for the rest
of his days his blindness to his wife’s deteriorating condition. That one sentence
would draw our attention from the final function-A event (the girl’s death) to the
stability – the new equilibrium – of the resultant ongoing situation. If readers find
the quoted paragraph dissatisfying when read as if it stood alone, I suggest, it is
largely because the paragraph guides readers to interpret the function of events from
the perspective of the girl, to whom nothing good happens to balance the bad,
and because it withholds all comment about the state of the narrative world after
her death.
But the quoted paragraph, which thus far we have analyzed as if it were a separate
entity, is the final section of Poe’s story. Poe’s story ends with the same words – ‘‘She
was dead’’ – and the same event – the girl’s death – that the quoted paragraph does.
I want now to look at Poe’s story in its entirety, ultimately to show how differently
readers interpret the causal effects of the girl’s death when we read about it in the
larger context of the complete story, which shifts readers’ primary attention from the
short life and the death of the girl to the context in which Poe’s narrator views
258
Emma Kafalenos
(presumably) her portrait and to the questions the portrait raises in his mind and
in ours.
Poe’s story begins at what can be called an achieved equilibrium: an equilibrium that
is the result of events that we learn about at least in broad outline. The opening
sentence reports a set of events that trace an entire sequence. The sentence reads:
The château into which my valet had ventured to make forcible entrance, rather than
permit me, in my desperately wounded condition, to pass a night in the open air, was
one of those piles of commingled gloom and grandeur which have so long frowned
among the Apennines, not less in fact than in the fancy of Mrs. Radcliffe.
(Poe 1977: 103)
Once we organize in chronological sequence the events this sentence reveals, we will
probably interpret the causal relations according to these functions:
Analysis 5 (how the initial equilibrium came about)
A
the narrator is too desperately wounded, his valet thinks, to spend the night
outdoors in the Apennines
C
the narrator’s valet decides to locate shelter for the narrator (the valet is the
C-actant – the character who performs function C)
C0
the valet breaks into a chateau
H
the valet establishes the narrator in a room inside the chateau
I
the valet succeeds in finding shelter for the narrator
EQ the narrator is now ensconced indoors for the night
In the rest of the first two paragraphs (the initial section of the story), the further
information reinforces readers’ initial interpretation that the opening situation is an
equilibrium. The narrator describes the ‘‘remote turret’’ (1977: 103) to which his valet
has brought him. He is in a bed that has black velvet curtains. A candelabrum provides
enough light that he can contemplate his surroundings.3 The walls, in his words, ‘‘were
hung with tapestry and bedecked with manifold and multiform armorial trophies,
together with an unusually great number of very spirited modern paintings.’’ A ‘‘small
volume’’ he found on his pillow ‘‘purported to criticise and describe’’ the paintings (p.
103). For hours, the narrator explains, he gazed at the paintings and read from the book,
until finally it was midnight. In this first section of the story, readers have been given
only one piece of information that suggests that the ongoing equilibrium is less stable
than it otherwise would have seemed. The narrator has mentioned that his ‘‘incipient
delirium’’ (his words) perhaps explains his ‘‘deep interest’’ in the paintings (p. 103).
In the next section of the story (the third, fourth, and fifth paragraphs), the
prevailing equilibrium is disrupted when the narrator moves the candelabrum.
With the resultant change in lighting, he is able to see a painting he had not
previously noticed: the oval portrait to which the story’s title refers. About
the painting, after this first glimpse, the narrator tells us no more than that it is
Effects of Sequence, Embedding, and Ekphrasis
259
‘‘the portrait of a young girl just ripening into womanhood’’ (1977: 104). His own
response, however, he describes in some detail. After glancing at the painting, he says,
he immediately closed his eyes, then began to analyze why he had done so. In his
words, ‘‘It was an impulsive movement to gain time for thought – to make sure that
my vision had not deceived me – to calm and subdue my fancy for a more sober and
more certain gaze’’ (p. 104). Since the narrator has previously mentioned his ‘‘incipient delirium,’’ and since he now describes ‘‘the dreamy stupor which was stealing over
[his] senses’’ (p. 104), readers probably gather that his initial view of the portrait has
led him to fear that he has in fact been overtaken by delirium and that he cannot trust
his sensory perceptions. We may interpret his response to the portrait according to
these functions:
Analysis 6 (narrator’s first response to the portrait)
EQ narrator is ensconced in the chateau
A
narrator fears he has become delirious
C
narrator decides to determine whether he is delirious (the narrator is the C-actant)
C0
narrator ‘‘calm[s] and subdue[s] [his] fancy for a more sober . . . gaze’’
H
narrator studies the painting at length
Thus far, all that we know about the portrait is that it has made a strong impression
on the narrator and that its subject is a young girl. We have no additional information
beyond what the narrator tells us, of course, because we cannot see the portrait, which
is embedded in Poe’s story through ekphrasis. As Tamar Yacobi has discerned, when
an artwork (or other document) is embedded in another work, two media need to be
considered if we are to understand how the embedded artwork is perceived by readers
or viewers in our world. On the one hand there is the medium we would perceive if we
could enter the narrative world and experience the artwork there – in the case of the
oval portrait, paint. On the other hand there is the medium in which the embedded
artwork is represented in the work in which it is embedded – in this case (as always in
the case of ekphrasis), words.4 Readers’ response to Poe’s story is affected by the
interplay between the two media: the visual representation (for which, on the basis of
our prior experience with paintings, we can conceive certain characteristics, but not
others) and the verbal rerepresentation that at best can convey to us one among other
possible descriptions.5
The narrator, after looking again at the painting, now judges that he ‘‘saw aright’’
(1977: 104). But because readers cannot see the portrait, we cannot gauge whether the
narrator’s initial shocked response was the result of something unusual about the
portrait, or whether his perceptions are less trustworthy than he now thinks they are.
In other words, at this point in our reading, some readers may be unsure whether to
analyze the causal relations among the events we have thus far learned in accordance
with Analysis 7 below (the narrator is in full control of his faculties and now
understands his earlier response) or Analysis 8 (the narrator’s shocked response to
the portrait may be an effect of his delirium).
260
Emma Kafalenos
Analysis 7 (the interpretation the narrator proposes)
EQ narrator is ensconced in the chateau
A
narrator fears he has become delirious
C
narrator decides to determine whether he is delirious
C0
narrator ‘‘calm[s] and subdue[s] [his] fancy for a more sober . . . gaze’’
H
narrator studies the painting at length
I
narrator decides he ‘‘saw aright’’
Analysis 8 (some readers’ interpretation)
EQ narrator is ensconced in the chateau
A
narrator may have become delirious; readers cannot trust his perceptions
Once the narrator judges that he ‘‘saw aright,’’ he devotes a few words to describing
the painting before analyzing why he had been ‘‘so suddenly and so vehemently
moved’’ (1977: 105). The portrait, which is set in an oval frame, depicts no more than
a head and shoulders, the body disappearing into the deep shadow of the background.
Without the frame, the head and shoulders disappearing into the shadows might well
have been understood as an apparition. Here Poe’s story flirts with the supernatural. It
is because of the frame, the narrator says, and because only a part of the body is
depicted, that he decides that even though his ‘‘fancy [had been] shaken from its
half slumber,’’ he could not have ‘‘mistaken the head for that of a living person.’’
Continuing to stare at the painting for perhaps another hour, he at last concludes:
‘‘I had found the spell of the picture in an absolute life-likeliness of expression,
which at first startling, finally confounded, subdued and appalled me’’ (p. 105,
Poe’s italics).
The word ‘‘appalled’’ guides us to recognize that the narrator is interpreting
viewing the painting as a function-A disruptive event. Because the narrator describes
the ‘‘life-likeliness’’ of the depicted girl’s expression as appalling, we understand that,
for him, deciding that he has viewed the painting ‘‘aright’’ does not re-establish the
equilibrium established in the first section of the story. Rather, the narrator has now
explained, at least to his satisfaction, why seeing the painting has been so disturbing
an event:
Analysis
EQ
A1
C
C0
H
I
I ¼ A2
9 (narrator’s interpretation of his experience in response to the portrait)
narrator is ensconced in the chateau
narrator fears he has become delirious
narrator decides to determine whether he is delirious
narrator ‘‘calm[s] and subdue[s] [his] fancy for a more sober . . . gaze’’
narrator studies the painting at length
narrator decides he ‘‘saw aright’’ – that the ‘‘life-likeliness’’ of the depicted
expression is appalling (the resolution of the function-A1 situation creates a
new function-A situation)
the ‘‘life-likeliness’’ of expression is appalling: the equilibrium is again destabilized
Effects of Sequence, Embedding, and Ekphrasis
261
Because the painting is represented in the story only through ekphrasis, which
makes it impossible for readers to envision it except through the double lens of the
narrator’s perceptions and the narrator’s words that describe his perceptions, some
readers may question whether visual representation can ever be sufficiently lifelike to
be considered appalling. For these skeptical readers, the narrator’s finding the ‘‘lifelikeliness’’ of expression appalling may corroborate and extend their earlier worry that
the narrator is delirious and that his perceptions – and now his conceptions also –
cannot be trusted.
Analysis 10 (skeptical readers’ interpretation)
EQ narrator is ensconced in the chateau
A
narrator may have become delirious; readers may not be able to trust his perceptions
A
visual representation so lifelike as to be appalling is impossible; readers cannot
trust narrator’s conceptions
It is exactly because skeptical readers question the accuracy of the narrator’s perceptions and conceptions, however, that they share his desire to understand why the
depicted girl’s expression has seemed to him appallingly lifelike. For this reason, at
this point in the story, the narrator himself, readers whose interpretations agree with
the narrator’s, and skeptical readers are similarly motivated to investigate the causes of
the narrator’s experience. Immediately after being told that the narrator finds the
portrait’s ‘‘life-likeliness’’ of expression appalling, we read:
With deep and reverent awe I replaced the candelabrum in its former position. The
cause of my deep agitation being thus shut from view, I sought eagerly the volume
which discussed the paintings and their histories. Turning to the number which
designated the oval portrait I read there the vague and quaint words which follow.
(1977: 105)
Although the narrator and skeptical readers are attempting to resolve different issues,
readers make the same decision that the narrator does: to read on to resolve the
questions that the narrator’s experience has raised.
Analysis 11 (Poe’s story, the first two sections)
A
(for the narrator) the ‘‘life-likeliness’’ of the portrait is appalling
A
(for the skeptical reader) the narrator’s incipient delirium may have affected his
perceptions and conceptions
C
a decision is made – by the narrator and by readers – to look for an explanation
through reading (narrator and readers are C-actants)
C0
narrator picks up book; readers read on
H
the narrator and readers read the quoted paragraph simultaneously and for the
same purpose: to find out why the narrator has found the portrait so appallingly
lifelike
262
Emma Kafalenos
Thus readers read the quoted paragraph not only as if at the same time the narrator
does, but also with the same function-C motivation that the narrator has: to look for
information that explains the narrator’s response to the painting. As a result, most
readers as they read, I suggest, devote more attention to the painting than to the
painter, and to the painter than to the girl – a reversal of the attention we pay to
the girl when we read the quoted paragraph as if it stood alone.6 In fact, when
we reach the final words and hear the painter, looking at his painting, say ‘‘ ‘This is
indeed Life itself!’ [before] turn[ing] suddenly to regard his beloved: – She was dead!’’
we find the explanation we sought. Even skeptical readers recognize that a painting
with the history of the one in question is no ordinary painting and that the
circumstances in which the painting had been made lend support to the accuracy of
both the narrator’s perception of the depicted girl’s lifelike expression and his
conception that that ‘‘life-likeliness’’ was appalling. After reading the complete
story, most readers, I propose, interpret the function of the reported events according
to these functions:
Analysis
EQ
A1
C
C0
H
I
I ¼ A2
C
C0
H
I
12 (Poe’s story in its entirety)
narrator is ensconced in the chateau
narrator fears he has become delirious
narrator decides to determine whether he is delirious
narrator ‘‘calm[s] and subdue[s] [his] fancy for a more sober . . . gaze’’
narrator studies the painting at length
narrator decides he ‘‘saw aright’’ – that the ‘‘life-likeliness’’ of the depicted
expression is appalling
the ‘‘life-likeliness’’ of expression is appalling
both the narrator and readers decide to look for an explanation through
reading (narrator and readers are C-actants)
narrator picks up book; readers read on
the narrator and readers read the quoted paragraph simultaneously and for the
same purpose: to find out why the narrator has found the portrait so appallingly lifelike
the circumstances in which the painting was made can be understood to
explain the narrator’s experience
Because the quoted paragraph is the concluding section of Poe’s story, both the
paragraph and the story end with the same words and the same event: the completion
of the painting and the subsequent death of the girl. Let us now contrast analysis 12
to analysis 3. When we read the paragraph by itself as if it stood alone (analysis 3), we
interpreted the death of the girl at the moment when the painting was completed as a
major disruptive event – in my terminology a function-A event. When we read the
identical paragraph in the context of Poe’s story (analysis 12), we interpreted the
death of the girl at the moment when the painting was completed as a successful
outcome to our efforts to explain why the narrator found the portrait so lifelike – in
my terminology a function-I event. Function analysis demonstrates the magnitude of
Effects of Sequence, Embedding, and Ekphrasis
263
the effect that the way events are told can have on interpretations of the causes and
effects of those events, in this case the death of the girl.7 I draw attention to five
narrative strategies that help to explain these very different interpretations that the
girl’s death elicits in the two contexts.
First, and perhaps most obvious, there is the primacy effect. The quoted paragraph
introduces the girl in the first sentence, whereas in the first sentence of Poe’s story the
narrator introduces himself.
Second, the primacy effect is reinforced in both cases by the perspective from which
information is presented. In the quoted paragraph, the information comes to us
through a lens that is positioned near the girl. When we read the quoted paragraph
by itself, as we saw at the beginning of this essay, the information we are given about
how the girl responds to the events in her life guides us to align our interpretation of
the function of the events with hers. Our interpretation stops at function A because,
from the perspective of the girl, after her death there is nothing more to interpret. In
the first two sections of Poe’s story, in contrast, it is the narrator himself whose
perceptions and thoughts we learn. As a result, we tend to align our interpretation of
the function of events with his interpretation. When we reach the quoted paragraph,
knowing that we are reading it as if at the same time that he is, we tend to align our
interpretation with what we presume is his interpretation. From his perspective, we
know him well enough to assume, the juxtaposition of the girl’s death and the
painting’s completion will provide reassurance that he was correct (function I) in
analyzing the painting’s ‘‘lifelikeliness’’ as the cause of its strong effect.
Third, in addition to which character is introduced initially we need to consider the
sequence in which events are reported. In the quoted paragraph the events in the girl’s
life are told chronologically. We are introduced to her, then told that she meets, falls
in love with, and marries the painter, and then that he paints her portrait. The first
and second sections of Poe’s story too, after the opening flashback, are told more or less
chronologically. The narrator lies in his bed, looks about the room, reads, then looks
at the oval portrait, then reads about the oval portrait. But in Poe’s story the quoted
paragraph that the narrator reads provides information about events that are chronologically prior to, rather than subsequent to, the night during which the narrator is
reading it.8
As Meir Sternberg has shown, the sequence in which information is provided to
readers affects the form of attention readers accord to it. The quoted paragraph, when
read by itself as if it stood alone, exemplifies suspense. We read it with the expectation
that we are going to learn what lies ahead: what is going to happen as the events
continue to unroll. Given this expectation our attention is focused on the outcome –
in this case the completion of the painting and the subsequent death of the girl.
When the quoted paragraph is read as a part of Poe’s story, on the other hand, it
exemplifies curiosity. We read it with the expectation that we are going to learn about
some previous occurrence that will help to explain the present situation in the
narrative world. Given this expectation, our attention is focused primarily on the
relation between the earlier events and the present situation – in this case the relation
264
Emma Kafalenos
between, on the one hand, the completion of the painting and the subsequent death of
the girl, and, on the other hand, the narrator’s appalled response to the portrait.9 The
sequence of the telling establishes the temporal relation between a present situation in
the narrative world and the events of another time period, previous or future, to which
readers’ attention may be drawn.
Fourth, in Metahistory, Hayden White (1973) recognized that the historian chooses
a segment of the historical record to narrate, and that the choice of one segment rather
than another segment influences interpretations of the events the historian reports.
White gives as an example the death of a king: the ‘‘death of the king may be a
beginning, an ending, or simply a transitional event in three different stories’’ (1973:
7). In other words, whether an historical event is interpreted as a beginning, as a
transition, or as a concluding event depends on its chronological position in the set of
events being told, which depends on which segment of the historical record the
historian chooses to treat. In fiction, as in accounts of historical events, interpretations
of causal relations among reported events are influenced by the chronological position
of the event in the set of events being reported. In the quoted paragraph at the end of
Poe’s story, when read by itself as if it stood alone, the segment of time that is reported
extends from the girl’s period of happiness to her death. Because her death is
chronologically the final event in the set of reported events, we interpret it as the
consequence of the earlier events and even as the culminating event that explains why
the events that lead up to it have been reported.
In the complete story as Poe tells it, however, the events that are reported scenically
take place during a segment of time that includes only a few hours. In the opening
paragraphs the narrator is lying in bed, looking at and reading about paintings. He
views the portrait, then reads about it in the book. During this temporal segment
the girl exists only as a portrait and as the subject of a paragraph. By limiting the
scenic representation to these few hours, Poe shifts the emphasis from the girl’s short
life and early death, as well as from the violence the narrator has previously suffered,
to the narrator’s experiencing the girl’s portrait, which is the only event that disturbs
the equilibrium during this temporal segment.
If we consider not just the events that are reported scenically but the entire set of
events that the story reveals about the narrator, the temporal segment that is reported
extends from the narrator’s being wounded somewhere in the Apennines to his being
brought for shelter to the château, to rest and recuperate. In the context of this
temporal segment, which Poe could have chosen to represent scenically in its entirety,
beginning with the wounding and the breaking into the château, the girl is again
only a portrait and a paragraph, and even the narrator’s shocked response to the
portrait is no more than a minor distraction during the period of his recovery.
Fifth, from the perspective of contemporary interarts studies, the apparition of a
head and shoulders that in Poe’s story is contained in a frame draws attention to
parallel temporal, ontological, and epistemological issues that embedded artworks
and literary representations of the supernatural both raise. Just as any artwork can be
thought of as a conduit to the earlier time in which it was made, and any portrait as a
Effects of Sequence, Embedding, and Ekphrasis
265
link to the time in the life of the person that is portrayed, the shock that the
appearance of a ghost creates is surely in part the result of the clash between two
temporal periods, the present in which it is perceived and the earlier time in which
the apparition is presumed to have been alive.
In addition to bringing into the present the prior temporal period in which both
the painted person and the ghostly apparition once presumably lived, both the
portrait and the ghost further complicate perception by introducing a tension
between contradictory interpretations of the ontology of the image. When a portrait
is thought of as an artwork, we assume an ontological boundary that we cannot cross.
We cannot enter a painting and, for instance, have dinner with the painted figure.
When a portrait is thought of as comparable to a photograph, a representation of
someone who once lived, then logically only the temporal gap – but not the
ontological gap – separates the person portrayed from the person viewing the
painting.10 Nonetheless, experience teaches us that we respond with different emotions to a painting that represents a person in pain, even when we know that the
representation is of an historical figure, than we do to a living person in front of us
who is in pain, or, arguably, even to a photograph of a person who is in pain. Part of
the difference is presumably the result of the temporal gap; there is nothing we can do
in our time period to relieve the suffering the painted figure once endured. But part of
the difference is the effect of the ontological separation indicated by the medium; oil
paint does not feel pain. Similarly, the continuing fascination with ghosts is in part
the effect of the tension between considering an apparition as an image of someone
who once lived among us and from whom we are separated only by time, or as an
emanation from an extraworldly realm to which we have no access.
The epistemological tension that the ontological issue introduces is compounded
by ekphrasis, which, as we saw earlier, allows readers to know only what the viewer
of the artwork in the narrative world sees and chooses to describe. In contrast to
the portrait that readers see only through the narrator’s words, the quoted paragraph
is in the same medium in the narrative world as in the story we read: language.
Because we read the very same account that the narrator does about the girl whose
husband paints her to death, we are more ready to accept than we would be if the
information were rerepresented to us by the narrator, that in the narrative world an
unusual relation of near-identity exists between the object of representation (the girl
who gives herself entirely to her husband’s art), the representation (the portrait that
conveys her ‘‘life-likeliness’’), and its perception (the narrator’s response as if to a
living person).
As a function analysis demonstrates, Poe’s complete story shapes readers’ interpretations of the effects of the girl’s death very differently than the quoted paragraph does.
The narrative strategies in the complete story that guide our interpretations include
the selection of the very short temporal segment that is scenically represented and
the sequence in which information is provided, both of which are made possible by
embedding the events pertaining to the girl. The two media through which the girl’s
story is represented contribute too. Because the portrait is represented only through
266
Emma Kafalenos
ekphrasis, readers can question the accuracy of the narrator’s scarcely believable
perceptions and conceptions – a questioning that for readers is a function-A disruption that motivates us to read the quoted paragraph to resolve it. Then, because
readers can join the narrator in reading the words of the quoted paragraph, we can
substitute our experience for his and read the passage as he presumably does: as
evidence of the accuracy of his perceptions.
NOTES
1 These five functions are selected, for the purposes of this essay, from the set of 10 functions
I use to analyze causal relations in any narrative (Kafalenos 1995, 1999).
2 Both Meir Sternberg (1978: 93ff.) and Menakhem Perry (1979: 53ff.) have summarized
and analyzed psychologists’ studies that document the primacy effect: our tendency to accept as valid the information we are initially
given, even when that information is contradicted later in the same message.
3 Similarities between this turret and the turret
in which the painter paints his wife permit
interpreting the quoted paragraph as a mise en
abyme – a contained work that mirrors or duplicates aspects of the work that contains it.
Moshe Ron (1987: 427) proposes that the
repetition of the word ‘‘turret’’ in this story
can in itself ‘‘prompt readers to posit a mise en
abyme relation with very little additional substance.’’ J. Hillis Miller drew my attention, in
conversation in response to an earlier version of
this essay, to the parallel between the candelabrum that lights the painting in the narrator’s
turret and the light in the painter’s turret that
‘‘dripped upon the pale canvas’’ (1977: 106,
italics added). I am grateful to Brian McHale,
who was the first to point out to me that the
quoted paragraph could be read as a mise en
abyme and who brought Ron’s essay to my
attention.
4 Yacobi perceives that
In ekphrasis . . . the represented as well as
the representing domains are representations of an object. . . . Like all ‘‘quotation,’’
therefore, ekphrasis bundles together no
less than three, rather than two, domains:
one first-order, strictly ‘‘represented’’; one
second-order, ‘‘representational’’ in the vis-
ual mode; one third-order, ‘‘re-presentational’’ in the linguistic discourse. From
the side of the reader of literature, we encounter the original object [in Poe’s story,
the girl] at a twofold remove . . . mediated
by the pictorial image that the language
itself mediates, re-images, quotes for us.
(1997: 36, her italics).
5 In Poe’s story (1) the girl is (2) represented in a
portrait, which is (3) rerepresented in words.
For analysis of the rerepresentation in one
medium of a representation in another medium, other than the visual-to-verbal move of
ekphrasis, see Kafalenos (2003).
5 Roland Barthes (1985: 150) describes ‘‘a picture [as] never anything but its own plural
description,’’ effectively suggesting the limitless ways a given visual depiction can be
described. Further, Barthes argues, ‘‘the identity of what is ‘represented’ is ceaselessly
deferred . . . the analysis is endless; but this
leakage, this infinity of language is precisely
the picture’s system: the image . . . is not
the repository of a system but the generation
of systems.’’ See also Spence (1982), particularly chapter 3, ‘‘Putting Pictures into
Words.’’
6 Alan Nadel pointed out in conversation, in
response to an earlier version of this essay,
that some readers when they begin to read
the quoted paragraph still expect (and want)
to be told how the narrator became wounded.
Unlike skeptical readers, for whom the story of
how the painting was made offers information
that helps them gauge the validity of the
narrator’s perceptions and conceptions, these
readers will not have their question answered.
But these readers’ curiosity about the narrator’s earlier adventure – like skeptical readers’
Effects of Sequence, Embedding, and Ekphrasis
interest in gauging the narrator’s perceptions –
by focusing their attention on something
other than the girl and her fate will lead
them to interpret the girl’s death differently
when they read the quoted paragraph as part of
Poe’s story than when they read it as if it stood
alone.
7 Functions name interpretations of causal relations among events in the narrative world; the
comparison I am drawing is between interpreting the death of the girl as a major disruptive event that ends her life (analysis 3) and
as an event that successfully resolves an enigma (analysis 12). For both the trusting
reader and the skeptical reader, however, learning about the girl’s death produces a stronger
emotional response when we read the quoted
paragraph as part of Poe’s story than when we
read it as if it stood alone. Stating the case for
the trusting reader (in a response to an earlier
version of this essay), James Phelan points to
the significance of
the way that the context intensifies the
horror of the embedded story. ‘‘The Oval
Portrait’’ seems to me to work by constructing a reliable witness to the effect of the
horror narrated in the embedded story. The
whole thing builds to the last two lines,
and the punch packed within those lines
depends on our having seen the result of the
REFERENCES
AND
Barthes, R. ([1969] 1985). ‘‘Is Painting a Language?’’ In The Responsibility of Forms: Critical
Essays on Music, Art, and Representation, trans.
R. Howard (pp. 149–52). New York: Hill and
Wang.
Kafalenos, E. (1995). ‘‘Lingering Along the Narrative Path: Extended Functions in Kafka and
Henry James.’’ Narrative 3, 117–38.
Kafalenos, E. (1999). ‘‘Not (Yet) Knowing: Epistemological Effects of Deferred and Suppressed
Information in Narrative.’’ In D. Herman (ed.),
Narratologies: New Perspectives on Narrative Analysis (pp. 33–65). Columbus: Ohio State University Press.
Kafalenos, E. (2003). ‘‘The Power of Double Coding to Represent New Forms of Representation:
The Truman Show, Dorian Gray, ‘‘Blow-Up,’’ and
267
painter’s effort in the narrator’s prior account of the painting’s effect.
For the skeptical reader too, whose position I
more comfortably share, the amazement that
a portrait can possess the ‘‘life-likeliness’’ that
the last two lines suggest that in fact it does,
is similarly intense.
8 Menakhem Perry discerns that one of the
common ways in which accounts vary from
the chronology of external events is by including ‘‘a block of information transmitted
from one character to another’’ (1979:
39–40).
9 Meir Sternberg describes the interplay between the two sequences of narrative (story
and discourse, a told and telling/reading
sequence) as the source of ‘‘the three universal narrative effects/interests/dynamics of
prospection, retrospection, and recognition
– suspense, curiosity, and surprise, for
short’’ (2001: 117). In this essay I draw attention to two of the three, suspense and
curiosity.
10 See Marie-Laure Ryan’s analysis of ‘‘boundaries within the representing discourse, and
boundaries within the represented reality
(the ‘semantic domain’ of the text); boundaries with gates to get across, and boundaries
with only windows to look through’’ (1990:
873).
FURTHER READING
Whistler’s Caprice in Purple and Gold.’’ Poetics
Today 24, 1–33.
Poe, E. A. (1977). The Portable Poe, ed. P. A. D.
Stern. New York: Penguin.
Perry, M. (1979). ‘‘How the Order of a Text
Creates its Meanings (With an Analysis of
Faulkner’s ‘A Rose for Emily’).’’ Poetics Today
1, 35–64, 311–61.
Propp, V. ([1928] 1968). Morphology of the Folktale,
revised and ed. L. A. Wagner, trans. L. Scott,
2nd edn. Austin: University of Texas Press.
Ron, Moshe (1987). ‘‘The Restricted Abyss: Nine
Problems in the Theory of Mise en Abyme.’’
Poetics Today 8, 417–38.
Ryan, M.-L. (1990). ‘‘Stacks, Frames and Boundaries, or Narrative as Computer Language.’’
Poetics Today 11, 873–99.
268
Emma Kafalenos
Spence, D. P. (1982). Narrative Truth and Historical
Truth: Meaning and Interpretation in Psychoanalysis. New York: W. W. Norton.
Sternberg, M. (1978). Expositional Modes and Temporal Ordering in Fiction. Baltimore, MD: Johns
Hopkins University Press.
Sternberg, M. (2001). ‘‘How Narrativity Makes a
Difference.’’ Narrative 9, 115–22.
Todorov, T. (1969a). Grammaire du Décaméron
[The Grammar of the Decameron]. The Hague:
Mouton.
Todorov, T. (1969b). ‘‘Structural Analysis of Narrative,’’ trans. A. Weinstein. Novel: A Forum on
Fiction 3, 70–76.
White, H. (1973). Metahistory: The Historical Imagination in Nineteenth-Century Europe. Baltimore, MD: Johns Hopkins University Press.
Yacobi, T. (1997). ‘‘Verbal Frames and Ekphrastic
Figuration.’’ In U.-B. Lagerroth, H. Lund, and
E. Hedling (eds.), Interart Poetics: Essays on the
Interrelations of the Arts and Media (pp. 35–46).
Amsterdam and Atlanta, GA: Rodopi.
17
Mrs. Dalloway’s Progeny:
The Hours as Second-degree
Narrative
Seymour Chatman
Michael Cunningham’s The Hours was a great critical and popular success when it
appeared in 1998, winning the Pulitzer and the PEN Faulkner prizes. When it was
turned into a movie, directed by Stephen Daldry, in 2002, sales of the novel and of
Mrs. Dalloway skyrocketed. This enthusiasm once again demonstrated the permeability of the borders between high and popular culture. Quite apart from its merits as
a novel, the popular reception of The Hours must be counted an important cultural
event, one that justifies an interest in the formal dimensions of Cunningham’s project.
What kind of a narrative is The Hours? How best to describe its relation to Mrs.
Dalloway – narratologically, stylistically, thematically? And what does that description tell us about both novels’ aesthetic and thematic achievements? Answering these
questions may shed light on the current popularity of the derivative or imitative text,
a kind of text that some critics believe especially characterizes the postmodern era.
The reviewers of The Hours could not agree on how to classify it. One thought it a
‘‘sequel.’’ Others evoked a musical metaphor, likening it to ‘‘variations on a theme,’’ or
‘‘extended riff.’’ More matter-of-factly, it was called a ‘‘retelling’’ or a ‘‘reworking.’’
Cunningham himself (in a question-and-answer session at a bookstore reading) said he
intended not an ‘‘imitation,’’ an ‘‘annex,’’ or a ‘‘pastiche’’ but a ‘‘rewriting.’’ Still others
settled for ‘‘emulation,’’ ‘‘replay,’’ ‘‘echo,’’ or ‘‘parallel.’’ Clearly our critical vocabulary
lacks distinctive terminology for what Gérard Genette has dubbed ‘‘second-degree’’
narratives, those bearing a more than passing resemblance to some original. To situate
novels like Cunningham’s, we need to understand the complex varieties of seconddegree texts. Fortunately, we have Genette’s excellent guide Palimpsests, which sifts
eruditely through literary tradition. Genette (1997) identifies five major types,
which, as is his custom, he distinguishes with daunting combinations of classical
morphemes: (1) ‘‘Intertexts’’ – these are older texts encapsulated within newer, as in
quoting, plagiarizing, and alluding; (2) ‘‘Paratexts’’ – the framing elements of a core
270
Seymour Chatman
text, like titles, prefaces, notices, and forewords; (3) ‘‘Metatexts’’ – literary criticism
and other kinds of external commentary; (4) ‘‘Architexts’’ – ‘‘the entire set of . . .
categories – types of discourse, modes of enunciation, literary genres – from which
emerges each singular text’’; and (5) ‘‘Hypertexts’’ – the products of ‘‘any relationship
uniting a text B, the ‘hypertext,’ with an earlier text A, the ‘hypotext,’ upon which it
is grafted in a manner that is not that of commentary.’’ Unfortunately, as users of the
internet know, ‘‘hypertext’’ has come to mean something quite different, namely,
‘‘nonlinear’’ or ‘‘linked.’’ So I’ll use ‘‘original’’ and ‘‘imitative’’ to refer to these different
texts.
Of course, every text arguably imitates one or more originals. But of principal
concern to this essay are those which are massively imitative. The Hours, like Max
Beerbohm’s parodies, or what Genette calls ‘‘serious transformations,’’ like Ulysses,
imitates another text on sustained and explicit grounds. It announces a contract with
Woolf’s novel: not only are some of its chapters entitled ‘‘Mrs. Dalloway,’’ one
character named ‘‘Virginia Woolf’’ and another ‘‘Mrs. Brown’’ (presumably after that
elusive ordinary person whom, Woolf argued, Arnold Bennett could never capture),
but ‘‘The Hours’’ was the original working title of Woolf’s manuscript.
The vagaries of traditional usage tend to muddy rather than clarify discussion.
Especially in English, different kinds of imitation are lumped indiscriminately
together. If it does nothing else, Genette’s taxonomy usefully clarifies what The
Hours is not. It distinguishes among texts that are satirically, playfully, and seriously
derivative. It further divides satires into three subtypes: (1) ‘‘strict parody,’’ (2) ‘‘satirical pastiche’’ or ‘‘caricature,’’ and (3) ‘‘travesty.’’ A strict parody is very short; it makes
fun of the original through wordplay, sticking as closely to it as possible: Thomas
Gray’s line becomes ‘‘the short and simple flannels of the poor’’; Hemingway’s stories
are retitled ‘‘The Snooze of Kilimanjaro,’’ ‘‘The Crullers,’’ or ‘‘A Clean-Well-Sighted
Ace.’’ Strict parody is but a rhetorical figure, a one-liner, and so tells us little about
extensive second-degree texts.
Satirical pastiche goes beyond minimal alteration. For instance, an elevated style
may be applied to trivial or wildly inappropriate events, as in ‘‘The Rape of the Lock.’’
And whereas ‘‘parody. . . deforms a text . . . pastiche . . . borrows a style – and all that
goes with it’’ (Genette 1997: 78). The opposite also occurs: a lowbrow style may be
used to convey ‘‘noble’’ events. This is ‘‘travesty’’ or ‘‘burlesque travesty’’ (a type rarer
today than in the eighteenth century). A modern English example is the reconstruction of the letter from Goneril to Regan (King Lear I, iv) as it might have been written
by a twentieth-century society matron: ‘‘Dearest Regan . . . We have been having the
most trying time lately with Papa . . . ’’ (by Maurice Baring, reprinted in Macdonald
1960: 307–11).
Pastiches need not be caricatural: they may be merely ‘‘playful,’’ a stylistic imitation aiming ‘‘at a sort of pure amusement or pleasing exercise with no aggressive or
mocking intention . . . it [is] the ludic mode of the imitative text’’ (Genette 1997: 27).
Classic examples appear in Proust’s Pastiches et Mélanges. The line between satire and
play is fine but perceptible: the satirist means to deride the original, its author, or
The Hours as Second-degree Narrative
271
someone else; the nonsatiric pasticheur means only to tease without really challenging
the value of the original. Both kinds entail critique, but ludic pastiche also implicitly
offers homage. It is gentle and shows evident respect for the original. In ‘‘The
Guerdon’’ and ‘‘The Mote in the Middle Distance,’’ for example, Max Beerbohm
teases Henry James’s excessively delicate representation of characters’ states of mind,
but clearly does not seriously deride the Master’s manner or content. Sometimes, as in
Proust, the motive for pastiche seems a mere flex of stylistic muscle.
Clearly, The Hours is neither caricature nor travesty: the reader will find no
ridiculing of Virginia Woolf, of her novel Mrs. Dalloway, or of Cunningham’s own
major characters (though it does satirize minor characters with satirized counterparts
in Mrs. Dalloway – Walter Hardy is Cunningham’s version of Hugh Whitbread,
Oliver St. Ives of Lady Bruton, Mary Krull of Miss Kilman). But The Hours profusely
imitates aspects of the style of Mrs. Dalloway, and to that extent entails pastiche. Still,
a novel whose intention were mere pastiche – satiric or playful – would not win a
Pulitzer Prize.
So we must turn to Genette’s category of the ‘‘serious’’ derivative, a category
heretofore unnamed by literary criticism. Genette distinguishes between serious
imitation and serious transformation. Serious imitation is forgerie, a word obsolete
in French which Genette’s translators render as ‘‘forgery.’’ Unfortunately, in English
that term connotes an illicit text which passes itself off as a newly discovered work by
a famous author or source, like James Macpherson’s Ossian poems or Thomas
Chatterton’s ballads. But forgerie can also mean simply ‘‘that which is wrought or
crafted’’ (the actual French equivalent of English ‘‘forgery’’ is falsification.) An English
example of this more neutral sense is the James Bond novels written after Ian
Fleming’s death by Kingsley Amis (pseudonym ‘‘Robert Markham’’), John Gardner,
and Raymond Benson.
But The Hours does not fit into that category either: it does not turn on further
adventures of Clarissa Dalloway – hence the inappropriateness of the term ‘‘sequel.’’
(Another novel published in 1999, Mr. Dalloway, by Robin Lippincott, is such a
sequel. It retains many of the characters, and its events occur in 1927. Additionally, it
‘‘transfocalizes’’ Woolf’s novel, in other words, tells the story from the perspective of
an erstwhile minor character, Richard Dalloway [as Rosencrantz and Guildenstern Are
Dead transfocalizes Hamlet]). The category most appropriate to The Hours, rather, is
‘‘serious transformation’’ or ‘‘transposition,’’ the category of Ulysses, Doktor Faustus,
and Mourning Becomes Electra. These are texts that do not continue the lives of the
original characters, but rather use them as patterns for new characters whose experience is somehow parallel. They preserve important elements of the story, but move it
to a new spatiotemporal world, a new diegesis (Genette 1997: 295). Genette finds
transposition the most important kind of hypertext, both historically and aesthetically. Paraphrasing and expanding his definition, we can say that transposition is an
‘‘imitation in a serious mode, usually signed by its real author and often marked
‘contractually’ (e.g., by a name) whose dominant function is to pursue or extend a
preexisting literary achievement while proclaiming itself as a new effort.’’
272
Seymour Chatman
The Hours is indeed an ‘‘extension,’’ in particular, what Genette might call a
complement of Mrs. Dalloway. It makes a new whole with its original, not merely
allusively but formatively. The engaged common reader of The Hours (and not only the
scholar studying second-degree texts) will (re)read Mrs. Dalloway for a fuller comprehension of how both novels work. Mrs. Dalloway then becomes an explanatory source
for what motivates Mrs Brown, why Clarissa Vaughan should be happy despite the
death of Richard, how gay marriages can be as ordinary as straight ones. The Hours fits
into the category of complement because it presupposes its readers’ familiarity with
Mrs. Dalloway and also their tacit agreement that the structure and style of the
original permit an alternative sexual ethos.
Obviously, declaring The Hours a ‘‘complementary transposition’’ is only a categorization, not an explanation or analysis. How precisely does it adapt Virginia Woolf’s
novel, and what are the aesthetic and thematic consequences? Does it constitute an
independent classic? I suggest that Cunningham’s failure to capture one important
aspect of Woolf’s technique limits his novel’s scope and so misses something of the
universality of Mrs. Dalloway’s appeal.
Let us first consider the obvious similarities of content. These are to be found
exclusively in the chapters that Cunningham labels ‘‘Mrs. Dalloway.’’ A generalized
plot summary fits both the original and Clarissa Vaughan’s story: on a given day in a
large metropolis, a female protagonist plans a party. A mentally ill man with whom
she feels close kinship kills himself in despair. The heroine is deeply shocked, but,
with the support of a stable partner, finds that this death has given her courage to live
on. (Mrs. Dalloway: ‘‘She felt somehow very like him – the young man who had killed
himself. She felt glad that he had done it; thrown it away. . . He made her feel the
beauty; made her feel the fun’’ [Woolf 1925: 186]; The Hours: ‘‘Clarissa Vaughan
thinks: ‘It is, in fact, a party, after all. It is a party for the not-yet-dead; for the
relatively undamaged; for those who for mysterious reasons have the fortune to be
alive’ ’’ [Cunningham 1998: 226].)
At a more detailed plot level, the correspondence is equally close. At the outset
each Clarissa feels playfully exhilarated as she goes out shopping for flowers (‘‘What a
lark! What a plunge!’’ [Woolf 1925:5]/‘‘What a thrill, what a shock . . . ’’ [Cunningham 1998: 10]). Each is observed sympathetically by a neighbor, a walk-on who never
reappears (Scrope Purvis/Willie Bass). Each reminisces fondly about adolescence in a
country home (Bourton/Wellfleet), and recalls a boyfriend’s playfully cynical response
to her love of natural beauty (Peter Walsh: ‘‘I prefer men to cauliflowers’’ [1925: 3]/
Richard Brown: ‘‘Beauty is a whore, I like money better’’ [1998: 11]). On their
errands, each runs into a bluff, well-known, and well-connected male friend (Hugh
Whitbread/Walter Hardy), men who do good works but who are more facade than
substance, men who normally reside in the country but are in town today for the sake
of a sickly partner (Evelyn/Evan). Each visits the shop of a female florist who is an old
friend (Miss Pym/Barbara). Each hears a noise out in the street associated with ‘‘a
face of the greatest importance’’ (the Prime Minister? the Queen?/Meryl Streep?
Vanessa Redgrave? Susan Sarandon?). Throughout the ‘‘Mrs. Dalloway’’ chapters,
The Hours as Second-degree Narrative
273
Clarissa Vaughan thinks the same thoughts as Clarissa Dalloway, sometimes in the
very same words (‘‘Heaven only knows why we love it [the city] so’’ [1925: 4/1998:
226]). What goes oddly unexplained, however, is why Clarissa Vaughan, despite her
manifold associations with Clarissa Dalloway, including Richard’s calling her by that
name, should be oblivious of these parallels. Are they supposed to be coincidental? If
not, what was Cunningham’s purpose in exercising this artistic license?
The parallelism among other characters is equally patent: Clarissa Vaughan also has
her faithful mate, Sally Lester. Like Peter Walsh, Richard Brown has been left by a
Clarissa who cannot handle his tumult; like Septimus Smith he commits suicide after
suffering an illness that causes him to hear voices. Richard’s lover Louis’s plight also
resembles Peter Walsh’s: rejected, he has gone off to San Francisco (as Walsh has gone
off to India), returning a bit the worse for wear and agonizing about an affair he is
having with a young person. Both Peter and Louis have nervous tics: Peter plays with
his pen knife, Louis counts his steps. Like Clarissa Parry, Clarissa Vaughan has chosen
a more stable if less exciting partner over a teenage sweetheart. Ms. Vaughan’s choice
can be read as the fulfillment of the (only latent) lesbian desire evident in Clarissa
Dalloway’s and Sally Seton’s kiss. (Both Clarissas remember that passionate kiss – to
the ‘‘wrong sex’’ – as the happiest moment of their lives.) Sally Lester, like Richard
Dalloway, attends a luncheon to discuss a silly project with Oliver St. Ives, the
counterpart of Lady Bruton. And so on.
Of course, these parallels concern only one of the three stories in The Hours; Mrs
Brown’s and Mrs Woolf’s stories are newly minted. So the novels really have quite
different structures and themes. Mrs. Dalloway turns on two main story threads –
Clarissa Dalloway’s preparations for her grand party, and Septimus Smith’s drift
toward suicide. Woolf joins these protagonists only contingently. Except for a critical
moment in Mrs Dalloway’s consciousness, the stories do not intersect. Clarissa never
meets Septimus but only hears about his suicide at her party.
Clarissa Vaughan, on the other hand, personally witnesses Richard Brown’s suicide.
Until the end, The Hours seems to be weaving three story threads. But the plot is not
really tripartite. The ‘‘Mrs. Brown’’ and ‘‘Mrs. Dalloway’’ chapters, in a surprise (and
rather un-Woolfian) twist, belong to the same story, since Richard, the adolescent
lover and lifelong friend of Clarissa, is revealed to be the son of Mrs Brown. Further,
the characters reflect two aspects of Clarissa Dalloway’s character. Laura Brown, who
prefers a good book to the company of her doting husband and apprehensive child,
corresponds to the retiring, ‘‘nun-like’’ Clarissa Dalloway constrained by fatigue and
illness. Conversely, Clarissa Vaughan’s busy life in New York corresponds to Clarissa
Dalloway’s activity as accomplished hostess. We are led to believe that Laura is
frustrated by unrequited lesbian desires, desires which she suppresses for a solitary
life in Canada – as opposed to Clarissa Vaughan, who has acknowledged her lesbianism, achieved liberation, and enjoys a satisfying domestic relationship. The chronology here is relevant: the gray 1950s of the Mrs. Brown chapters is still closet-time,
the time of repression, while the 1990s generally accepts more diverse roles for
women and is witnessing a waning of homophobia.
274
Seymour Chatman
But what role do the ‘‘Mrs. Woolf’’ chapters play in Cunningham’s complementary
novel? The biographical fragment is not popular with some Woolfians, who argue
that Virginia was not dominated by Leonard, that she was not mean to her servants,
that Vanessa’s children did not make fun of her, and so on. Of course, the novel is not a
biography. The ‘‘Mrs. Woolf’’ sections seem designed less to reconstruct the actual
genesis of Mrs. Dalloway than to draw situational parallels with Cunningham’s
protagonists – Virginia’s ‘‘entrapment’’ in Richmond with Laura Brown’s entrapment
in a Los Angeles suburb, and her yearning for an active metropolitan life of the sort
enjoyed by Clarissa Vaughan. Indeed, the juxtaposition of Mrs Woolf, Mrs Brown,
and Ms Vaughan appears more admonitory than historical. Though acknowledging
that the world is a vale of tears (AIDS corresponds to World War I shell-shock), The
Hours suggests that a degree of salvation and happiness can be achieved by being true
to one’s sexual inclinations. Cunningham’s ‘‘Virginia Woolf’’ sublimates her lesbian
impulse, creates a novel, and the novel solaces a reclusive woman ill-suited to
conventional marriage but unready to move to a different sexual orientation. For
Laura, the consequence is celibacy. Clarissa Vaughan is clearly the happiest of the
three: growing up in a more tolerant age, she finds contentment with another woman,
and can celebrate life and Richard’s poetic achievements even with the mother whom
he publicly reviled.
The explicit connection between the plots, of course, isn’t Mrs Woolf but Virginia
Woolf’s Mrs Dalloway, a fictional character seemingly projected as a ‘‘working out’’ of
Woolf’s suicidal depressions, a projection crystallized in her decision to kill off not
Clarissa but Septimus (‘‘a greater mind than Clarissa’s’’ [Cunningham 1998: 154]).
Insofar as they recount the writing of Mrs. Dalloway, the ‘‘Mrs. Woolf’’ chapters
presumably relate to the sjuzhet or discourse of The Hours. Here the ‘‘real’’ author is
portrayed as deciding just how to tell a story, a story which inspires one reader, Mrs
Brown, to change her life and somehow serves (mysteriously) as a pattern for Clarissa
Vaughan. These two threads presuppose an earlier, generative, part, a moment in the
life of Virginia Woolf and her evolving creature, Clarissa Dalloway. There is, of course,
nothing new about using a real historical figure as a character in a fiction. Indeed, the
technique occurs frequently. One reviewer, Jonathan Dee (1999), cites the recent
‘‘torrent of such novels, by Russell Banks (on John Brown), Pat Barker (Wilfred Owen
and Siegfried Sassoon), Jay Parini (Walter Benjamin), Thomas Pynchon (Mason and
Dixon), Susan Sontag (Lord Nelson), John Updike (James Buchanan).’’
At the level of style, that is, of homage pastiche, The Hours (and Lippincott’s Mr.
Dalloway too) offers a plenitude of Dallowayisms. Here are a few examples:
.
.
Exclamations, especially of joie de vivre: Mrs. Dalloway: ‘‘What a lark! What a
plunge!/The Hours: ‘What a thrill, what a shock’ (but without exclamation points:
Woolf uses them to mark at least 25 of Clarissa’s thoughts in the first 10 pages of
the novel; Cunningham is mostly content with a simple period).
Interruptions of the flow of thought – marked by commas, parentheses, or dashes
– as each Clarissa tries to recover a memory, searches for a word, feels uncertain.
The Hours as Second-degree Narrative
275
(The self-questioning tic is prominent in Woolf’s diary, e.g., ‘‘All my nerves stood
upright, flushed, electrified (what’s the word?) with the sheer beauty [of nature],’’
August 19, 1924, Woolf 1996: 64). A favorite form is ‘‘Was that it?’’ – Mrs.
Dalloway: ‘‘Peter Walsh said ‘Musing among the vegetables?’ was that it?’’
(1925:3); The Hours: Clarissa wonders what to buy Evan, like Richard an AIDS
victim: ‘‘Not flowers; if flowers are subtly wrong for the deceased they’re disastrous for the ill. But what?’’ (1998: 21). (About these parentheses, see John
Mullan’s (2003) review ‘‘But I digress . . . ’’).
. Repetitions that register the full implications of what each Clarissa has just
thought or perceived, a form that would be otiose in ordinary prose: Mrs. Dalloway: ‘‘Clarissa read on the telephone pad, ‘Lady Bruton wishes to know if
Mr. Dalloway will lunch with her today.’ ‘Mr. Dalloway, ma’am, told me to tell
you he would be lunching out’ ’’ then: ‘‘Lady Bruton, whose lunch parties were
said to be extraordinarily amusing, had not asked her’’ (Woolf 1925: 29); The
Hours: ‘‘Clarissa pushes the rewind button. If Sally forgot to mention her lunch
with Oliver St. Ives it’s probably because the invitation was made to Sally alone.
Oliver St. Ives, the scandal, the hero, has not asked Clarissa to lunch’’ (Cunningham 1998: 93–4).
. Frequent ‘‘near’’ deixis through words like ‘‘here,’’ ‘‘now,’’ ‘‘this,’’ to highlight the
immediacy of the characters’ (rather than the narrator’s) immediate presence on
the scene rather than the narrator’s. Mrs. Dalloway : [about the air at Bourton]
‘‘How fresh, how calm, stiller than this, of course’’ (1925: 3)/The Hours: ‘‘New
York . . . always produces mornings like this’’ (1998: 9); Mrs. Dalloway: Peter
thinks: ‘‘Life itself, every moment of it, every drop of it, here, this instant, now,
in the sun, in Regent’s Park, was enough’’ (1925: 79)/The Hours: ‘‘If she were to
express it publicly (now, at her age), this love of hers would consign her to the
realm of the duped and the simpleminded’’ (1998: 12).
But there are important nuances of Mrs. Dalloway’s style which Cunningham seems to
miss or ignore. Their absence affects the degree and nature of our immersion in his
characters’ psyches. Cunningham’s are too often depicted from an outside, narrator’s,
perspective (or ‘‘slant’’). Perhaps Cunningham did not grasp the extent of Woolf’s
internalizing method, or perhaps he had other goals. Whatever the reason, he made
quite different (and, I believe, technically less interesting) lexical and syntactic
choices.
The principal differences concern syntactic focus. Woolf studiously avoided presenting Clarissa from an external narrator’s vantage, in the manner she associated with
Arnold Bennett’s fiction. How Clarissa gets around town is syntactically minimized, a
mere frame for her musings – her memories of Bourton and of Peter Walsh’s sarcastic
remarks, her sense of the hush that follows Big Ben’s sound, her love of the bustle of
London. When Woolf does provide an external view of Clarissa she embeds it in the
consciousness of another character, often a walk-on, a personage who, despite his or
her irrelevance to the plot, could function as ‘‘chorus’’: ‘‘She stiffened a little on the
276
Seymour Chatman
kerb . . . A charming woman Scrope Purvis thought her’’ (Woolf 1925: 4). In her
notebook Woolf tells herself: ‘‘no chapters,’’ and then ‘‘possible choruses.’’ Scrope
Purvis is such a chorus. Deep into Aeschylus at the time, Woolf decided to use
observing bystanders instead of the narrator to describe aspects of the protagonists’
appearance and behavior: ‘‘Why not have an observer in the street at each critical point
who acts the part of chorus – some nameless person?’’ (1996: 419). Equally explicit is
her diary entry of September 5, 1923: ‘‘Characters are to be merely views: personality
must be avoided at all costs . . . Directly you specify hair, age, etc. something frivolous,
or irrelevant gets into the book.’’ ‘‘You’’ here clearly refers to an external narrator.
Since she was writing a novel, not a play, she did not need to have these walk-ons
literally speak; she could simply enter their minds. Henry James had already questioned narratorial omniscience; Woolf was going a step farther in minimizing the
narrator’s independent spatial vantage on the scene. As I shall argue below, the device
has important thematic implications.
Cunningham’s sentences, however, more regularly feature Clarissa as the subject of
a visualizable predicate, and hence the narrator as interpreter of the spectacle:
‘‘Clarissa . . . runs out, promising to be back in half an hour’’ (1998: 9), ‘‘She delays
for a moment the plunge’’ (p. 9), ‘‘As Clarissa steps down from the vestibule her shoe
makes gritty contact with the red-brown, mica-studded stone of the first stair’’ (p. 10),
‘‘Clarissa crosses Eighth Street’’ (p. 13). By Woolf’s standards, these sentences presuppose a narrator who moves Clarissa from place to place, instead of letting her
thoughts simply flow along, without much concern for her whereabouts. We see
Woolf erasing such expressions from her manuscript. For instance, from the original
sentence ‘‘she had burst open the French windows . . . & stepped out on to the terrace
of Bourton’’ she deletes everything after ‘‘&.’’ Assigning the narrator the task of
tracking Clarissa’s stroll (Woolf would say) interferes with our immersion in her
mind. Cunningham’s syntax, by Woolfian standards, excessively features his protagonist’s occupations at the expense of her preoccupations. At the very beginning of the
‘‘Mrs. Dalloway’’ section, for example, Vaughan is the subject of a stream of active
predicates evoking an externally observable scene: ‘‘Clarissa feigns exasperation . . .
leaves Sally cleaning the bathroom, and runs out, promising to be back in half an
hour’’ (1998: 9).
A related stylistic detail: Woolf often prefers -ing verb forms – both present
participles and gerunds – to completed predicates. These emphasize the continuing
flow of experience and the co-occurrence of events inside and outside the mind, the toand-fro of memory, opinion, perception (fittingly, as we learn, since Clarissa Dalloway
has a life-and-death interest in continuation). Typical is the novel’s third paragraph,
which, after the exclamations, moves to a participial string that could, theoretically,
go on forever: ‘‘How fresh . . . the air was . . . solemn, feeling as she did, standing there
at the open window, that something awful was about to happen; looking at the
flowers, at the trees with the smoke winding off them and the rooks rising, falling;
standing and looking . . . ’’ (1925: 3). Actually, in her diary Woolf expressed some
uneasiness about her penchant for present participles: ‘‘It is a disgrace that I write
The Hours as Second-degree Narrative
277
nothing, or if I write, write sloppily, using nothing but present participles’’ (September 7, 1924). But Clarissa’s need to live, even if in a state of sensation and dread, could
not have been more effectively represented. Present participles enabled Woolf to
highlight meditation and minimize banal movements through the streets – ‘‘‘Such
fools we are,’ she thought, crossing Victoria Street’’ (1925: 4). Absorbed selfcommuning remains the focus; -ing forms, casually attached, are only there to mark
Clarissa’s physical progress.
Although he does of course use some present participles, Cunningham regularly
favors a verb form we don’t find in Mrs Dalloway – the historic present tense: ‘‘The
vestibule door opens onto a June morning’’ (1998: 9), ‘‘Clarissa pauses at the
threshold’’ (p. 9), ‘‘Clarissa crosses Eighth Street’’ (p. 13) and so on. Whether the
historic present suggests a greater immediacy than the simple preterite is debatable,
but its use is clearly less evocative of the eddying dynamics of the protagonist’s inner
life than Woolf’s present participles.
The Hours generally differs from Mrs. Dalloway in its treatment of time. Of course,
both novels rely on temporal unity: the whole of Mrs. Dalloway takes place on a single
day – the day of Clarissa’s party – while The Hours lasts a single day in each of the lives
of the protagonists. But the movement through ‘‘the hours’’ of these days varies
noticeably. In Mrs. Dalloway the exact diegetic moment is frequently stated or
implied, providing a detached real-life chronological armature for the multiplicity
of consciousnesses. A nexus is formed by clocks, which figure significantly in the
background and at moments of transition – not only the city-wide booms of Big Ben
but also the registrations of local clocks, like that above the shop of Rigby and
Lowndes (1925: 102). Clocks tangibly demarcate the thought flows of characters. As a
clock strikes, the discourse often seeps from Scrope Purvis to Clarissa, from Regia to
Hugh Whitbread, from Sir William Bradshaw to Peter Walsh. The clocks facilitate
daringly unpredictable shifts from one consciousness to another.
A larger purpose underlies this scrupulous fixing of chronology: Woolf is preoccupied with the whole life of London and beyond it, the British Empire. Her city is
more than just background. The Hours, on the other hand, concerned as it mostly is
with Greenwich Village, uses the ‘‘hour’’ as a thematic leitmotif. It contrasts the
despair that Richard Brown feels about how he will fill the few hours of life left to
him with Clarissa Vaughan’s optimism about the future – ‘‘there will be ecstatic hours
and darker hours, but still we cherish the city, the morning; we hope, more than
anything for more’’ (1998: 225).
Woolf’s innovative contribution to fiction is a great fluidity of mental representations, in indirect or direct style, to form a path among unrelated as well as related
characters. Chance proximity is sufficient to trigger one. A typical example: Peter
Walsh, consoling himself in Regent’s Park, watches a little girl, Elise Mitchell (given
a name but only a momentary presence), who ‘‘scudded off again full tilt into a lady’s
legs,’’ a sight which makes him laugh out. Not missing a beat, the next paragraph
enters a totally different mind, Lucrezia’s, preoccupied as it is with the unfairness of
her plight – until the same Elise ran ‘‘full tilt into her, fell flat, and burst out crying.’’
278
Seymour Chatman
‘‘That was comforting rather. She stood her upright, dusted her frock, kissed her’’
(Woolf, 1925: 65).
Spatial proximity can even facilitate telepathy:
For Heaven’s sake, leave your knife alone! [Clarissa] cried to herself in irrepressible
irritation; it was [Peter’s] silly unconventionality, his weakness; his lack of the ghost of a
notion what any one else was feeling that annoyed her, had always annoyed her; and now
at his age, how silly!
I know all that, Peter thought; I know what I’m up against, he thought, running his
finger along the blade of his knife.
(Woolf 1925: 46)
In Mrs. Dalloway, shifting mental filters through chance proximity has important
political as well as aesthetic implications. The technique emphasizes the democracy of
the city street (in the face of England’s official class rigidity), a place where everyone,
even the youngest or most modest denizen, enjoys the dignity of a name and even a
partial view of the scene. The aesthetic effect, early perceived by David Daiches
(1942), is to make Woolf’s art impressionist. What is important is not how things
are in themselves but how they appear to characters. The novel registers these
impressions virtually simultaneously but from different angles. To use Woolf’s own
word, though in a different sense, the novel moves through ‘‘haloed’’ mental registrations, through a space that is permeable among agents regardless of their importance to the plot. The narrator does not independently concretize how things look,
sound, or feel ‘‘out there.’’ Rather she travels from one character’s angle to another –
that of Peter, Rezia, Edgar J. Watkiss, Sarah Bletchley, Maisie Johnson, Mrs Dempster
– whoever the random witness is. The nexus between politics and aesthetics is
implicit in Daiches’s observation that characters are ‘‘introduced whose sole function
is to have fleeting impressions of the principals . . . yet not solely, for this sudden flash
of light reveals him [or her], too, as an independent person with a life of his own
somewhere in the background, with experiences, prejudices, a texture of living’’
(Daiches 1942: 56). ‘‘Independent’’ is worth repeating: the technique democratically
reminds us that there is more to the vibrant world of London than Clarissa’s party and
Septimus’s death. The crowd in the street jumbled together without regard to class
distinction is not mere background. It represents, not only bodies brushing past each
other, but also minds and plights. By the same token, British history, no longer
locked behind dusty academic doors, opens out into this microscopically examined
hour of city life.
Woolf enhanced this sense of psychic fluidity among her characters by forgoing
chapter breaks, separating the novel’s parts only by gaps in the space on the page. The
Hours, on the other hand, is divided into a prologue and 22 chapters clearly labeled
with one of the three major characters’ names. Further, it attenuates the number and
abruptness of shifts among consciousnesses. Most of the book is confined to the
consciousnesses of the protagonists or to major characters like Richard Brown,
The Hours as Second-degree Narrative
279
Louis Waters, or Sally Lester. Except for that of Willie Bass, no consciousness from the
outside world is followed. When Clarissa Vaughan shares diegetic space with random
minor characters, there is usually no corresponding shift in filter. Compare, for
example, the ‘‘unknown famous face’’ sequences in both novels. In Mrs. Dalloway,
named individuals vividly experience this ‘‘face of the very greatest importance’’
(1925: 14), this symbol of Empire. They hear the ‘‘voice of authority, the spirit of
religion.’’ This face, this ‘‘greatness,’’ the narrator predicts, as if speaking for the
crowd, ‘‘will be known to curious antiquaries, sifting the ruins of time, when London
is a grass-grown path and all those hurrying along the pavement this Wednesday
morning are but bones with a few wedding rings mixed up in their dust and the gold
stoppings of innumerable decayed teeth’’ (Woolf 1925: 16). Anyone who knows
Woolf’s essays, diaries, and letters, of course, cannot believe that the real author
unironically endorses the narrator’s effusions. The paratextual evidence alone – the
author’s very name – calls into question any rosy prognosis of the British empire’s
future. And as if to corroborate our skepticism, the crowd’s awed attention quickly
moves to the smoke letters of the skywriting airplane. (Woolf’s diary entry of June 19,
1923: ‘‘I want to give life and death, sanity and insanity; I want to criticise the social
system and to show it at work, at its most intense,’’ [Woolf 1996: 56]).
In the corresponding scene in The Hours, however, we stay only in Clarissa’s
consciousness. Two young bystanders are mentioned, but we learn neither their
names nor their unspoken thoughts. We hear only the dialogue between them,
through Clarissa’s ears. The one with hair dyed canary yellow is arbitrarily named
‘‘Sun’’ by Clarissa and the other, the platinum, ‘‘Moon.’’ The only mind entered is
Clarissa’s, as she silently agrees with Sun that the famous face belongs to Meryl Streep
and not Susan Sarandon. It is true that Cunningham also puts the episode in historical
perspective: Sun and Moon
will grow to middle age and then old age, either wither or bloat; the cemeteries in
which they’re buried will fall eventually into ruin, the grass grown wild, browsed at
night by dogs; and when all that remains of these girls is a few silver fillings lost
underground, the woman in the trailer, be she Meryl Streep or Vanessa Redgrave or even
Susan Sarandon, will still be known. She will exist in archives, in books; her recorded
voice will be stored away among other precious and venerated objects.
(Cunningham 1998: 50–1)
But since his novel lacks Woolf’s breadth of social and political resonance, it is
difficult to know precisely what it is satirizing. Is it the adoration of celebrity? Is it
archivists? Since The Hours seems so much less a social critique than Mrs. Dalloway,
does the author even mean this sentence ironically?
Certain of Woolf’s ‘‘choruses’’ converge at Clarissa Dalloway’s party (not only in the
novel but in various short stories written during this period). Clarissa reflects upon
her guests, upon her own past, and upon the world outside her window; and
the characters who know her best – Sally Seton, her husband Richard, and Peter
280
Seymour Chatman
Walsh – reflect upon her. The last to do so is Peter, for whom Clarissa remains . . . indescribable: ‘‘For there she was.’’ The words that end The Hours are only superficially
similar, their force is quite different. Despite his consistent usage of indirect free style
to represent characters’ thinking, Cunningham seems uninterested in Woolf’s concept
of character as ‘‘chorus.’’ It is Clarissa, not some outsider, who thinks ‘‘here she is,
herself, Clarissa, not Mrs. Dalloway anymore; there is no one now to call her that.
Here she is with another hour before her’’ (1998: 228). Quite apart from the difference
in content, that’s a significant alteration of Mrs. Dalloway’s overall point-of-view
strategy and the implications about history and culture that it engenders.
Judging Cunningham’s overall success is not the task of this essay, but placing it in
the universe of second-degree texts would be arid without drawing some critical
inferences. Despite his successful transposition of many features of Mrs. Dalloway, we
see important significant stylistic differences that bear on Cunningham’s project, a
project which might be characterized as transferring the spirit, and, indeed, much of
the letter of Woolf’s 1923 London to 1998 Greenwich Village. Cunningham says he
was inspired by Woolf’s ability to ‘‘write the epic stories of ordinary people.’’ The
word shows up at a critical moment in his ‘‘Mrs. Woolf’’ section when Virginia
decides that Septimus should die instead of Clarissa, who shall spend her life ‘‘loving
her life of ordinary pleasures’’ (Cunningham 1998: 211). But despite his apparent
range – the evocation of Woolf in Richmond, of a lost housewife in a 1950s Los
Angeles suburb (in some ways the most achieved part of the book), and of a book
editor in Greenwich Village – Cunningham does not achieve the broad cultural sweep
of Mrs. Dalloway. His Village has been likened to Woolf’s Bloomsbury. But Mrs.
Dalloway is not set in Bloomsbury; it is not concerned with a colony of artists and
intellectuals; its purview is the whole of London’s population, indeed the whole of a
British empire in decline. It is a design finely etched, as we’ve seen, by devices like the
‘‘chorus.’’ Both explicitly and implicitly, Woolf conjures up a simulacrum of English
society of the 1920s – from the Prime Minister to the ‘‘veriest frump’’ in the street.
Cunningham’s limitation to a small segment of American culture, to a group of
professionals in the entertainment industry living in Greenwich Village – however
accurate the depiction – against its achieving the pantheon of transposed novels, a
pantheon dominated by Ulysses. It is true that the equivalencies work – that Richard’s
suicide is made plausibly comparable to Septimus’s, and that the tension between
Clarissa Dalloway’s and Sally Seton’s kiss and the humdrum domestic contentment of
the Dalloways is satisfyingly reflected in the relationship between Clarissa Vaughan and
Sally Lester, that, in short, Cunningham has gone beyond gender questions to address
the commonality of human emotions, to what James Phelan aptly calls his characters’
‘‘responses to the wonder and chaos and pain of existence’’ (private correspondence,
2003). To that extent, The Hours can be accounted a small ‘‘epic of ordinary lives.’’
But what circumscribes the novel is its apparent lack of engagement with the rest
of the American, even of the New York, world, a world with which gay intellectuals,
like everybody else, must interface. The Hours seems too sealed off. Like her counterpart, Clarissa Vaughan feels, at least occasionally, anxiety and a sense of failure. But
The Hours as Second-degree Narrative
281
Mrs Dalloway’s plight has an important social cause – the exclusion of women,
including those of the upper-middle class, from matters of national significance.
And while it may be true that the AIDS epidemic has been as great a scourge as
World War I, it has not led to the sort of political and social consequences that
afflicted Britain in the 1920s.
The Hours persuasively offers a ‘‘postcloset’’ American rewriting of Mrs. Dalloway,
addressing an implied reader with enlightened attitudes about same-sex relationships.
For such readers – and happily they are increasingly numerous and mainstream – the
correspondences with Mrs. Dalloway are quite happy. But to match the scope of
the original, Cunningham needed a broader canvas and a set of Woolfian techniques
that he didn’t utilize.1
NOTE
1 I am grateful to the editors and to Emma
Kafelanos, Gerald Prince, and Alex Zwerdling
for help with this essay.
REFERENCES
AND
Chatman, S. (2001). ‘‘Parody and Style.’’ Poetics
Today 22, 25–39.
Cunningham, M. (1998). The Hours. New York:
Picador USA.
Cunningham, M. (2002). ‘‘Talk at Kelly Writers
House Fellows Seminar (February 11, 2002).’’
<http://www.english.upenn.edu/whfellow/
cunningham.html>.
Daiches, D. (1942). Virginia Woolf. Norfolk, CT:
New Directions.
Dee, J. (1999). ‘‘Review of The Hours.’’ Harper’s
Magazine 298 (June), 76.
Genette, G. ([1982] 1997). Palimpsests: Literature
in the Second Degree, trans. C. Newman and
C. Doubinsky. Lincoln and London: University
of Nebraska Press.
Lee, H. (1997). Virginia Woolf . New York: Knopf.
FURTHER READING
Lippincott, R. (1999). Mr. Dalloway. Louisville,
KY: Sarabande.
Macdonald, D. (1960). Parodies: An Anthology from
Chaucer to Beerbohm – and After. New York:
Random House.
Mullan, J. (2003). ‘‘‘But I digress . . . ’: a review of
The Hours.’’ The Guardian 22 February.
Wood, M. (1998). ‘‘Parallel Lives.’’ The New York
Times Book Review 22 November, Section 7,
Column 1, p. 6.
Woolf, V. (1925). Mrs. Dalloway. London: Harcourt.
Woolf, V. (1996). Virginia Woolf ‘‘The Hours’’: The
British Museum Manuscript of Mrs. Dalloway, ed.
H. Wussow. New York: Pace University Press.
Zwerdling, A. (1986). Virginia Woolf and the Real
World. Berkeley: University of California Press.
PART III
Narrative Form and its
Relationship to History, Politics,
and Ethics
18
Genre, Repetition, Temporal
Order: Some Aspects of Biblical
Narratology
David H. Richter
The term ‘‘biblical narratology’’ is an oxymoron, particularly if we come to biblical
narrative not from an ideological perspective but from the angle of the formal features
that are peculiar to it. Contemporary narratologies, both rhetorical and structuralist,
in the style of Phelan and the style of Fludernik, were created to operate on the
complexities of works like Absalom, Absalom! rather than the book of Samuel. That is,
their analyses were designed for works of narrative artistry that are wholes rather than
totals, that are written by identifiable authors about whose lives and attitudes
information can be discovered, or – in the case of anonymous works – by authors
who can be placed with some confidence both geographically and historically. They
presume that the texts of these works can be established, that omissions, transpositions, and additions imposed by later redactors have not warped them almost beyond
recognition. They presume that we can easily intuit whether a given narrative is
intended to be read as fiction or as fact or an intricate combination of the two.1 They
further presume that we can understand in at least a rough and ready way the system
of genres within which a given narrative text has its place. And they presume that we
are free to locate the meaning of a text using rules of notice, signification, configuration, and coherence, that is, the usual rules for the interpretation of secular narratives
identified and elucidated by Rabinowitz, rather than special rules of interpretation
that derive from exterior systems of belief. None of these things is true of biblical
narrative – which is kind of scary. Here I would like to focus on three of these
interconnected issues: genre, repetition, and temporal order.
Genre
Let me take as my first text the familiar story of Jonah, who, summoned to prophesy
the doom of Nineveh as punishment for its wickedness, attempts to flee his mission
286
David H. Richter
by taking ship in the opposite direction. God thwarts this escape by causing a
tempest, and Jonah, admitting that the tempest is his own fault, is thrown overboard,
and swallowed by a fish who, three days later, vomits him up on the seashore. Sent by
God once more to Nineveh, Jonah fulfills his mission. His prophecy is believed by the
inhabitants, who repent, fast and pray, so that God relents and does not destroy them.
This is the first three-fourths of the story, the part that has entered orthodox
Judaism, Christianity, and Islam, each of which has different ways of accounting for
the one rough patch in that part of the story, Jonah’s refusal of his mission. The Jewish
reading of Jonah is, by and large, one that accords with its role as the prophetic book
read on the day of Yom Kippur, often ignoring Jonah’s questionable behavior and
concentrating instead on the Ninevites who avert divine judgment by repentance,
fasting, and prayer.2 Christian commentaries on Jonah take off, as they must, from
Jesus’s identification of himself with Jonah in Matthew and Luke. In Matthew 12:
38–41, Jonah primarily prefigures the death and resurrection of Jesus by descending
into the depths and rising after three days; in Luke 11: 29–32, Jesus reads Jonah as a
sign for the prophet of God whose word compels spiritual change. Jonah is thus a
messianic antetype in ways that have been endlessly elaborated.3 In the Qur’an the
prophet Yunus teaches several moral lessons: Yunus was ‘‘one of those who are cast off’’
who would have been left in the belly of the fish till the last judgment, save that he
redeems himself through confession and prayer, to which Allah responds with
deliverance.4
What all these interpretations have in common is that they have agreed to ignore
the last chapter of Jonah, which finds the prophet camped outside Nineveh, waiting
to witness its destruction, made so miserable by the success of his mission that he
wishes he were dead. Jonah spells out his point to YHWH: didn’t I tell you why I was
refusing to come here? ‘‘Didn’t I already say before I fled for Tarshish that you are the
God of grace and compassion, slow to anger, full of kindness, repenting of evil?’’5
Then, since the Ninevites seem to have learned their lesson, God directs his mighty
acts toward Jonah, who has not. He creates a gourd to shade Jonah from the fierce
Assyrian sun, and once Jonah is glad of the shelter, makes a worm to destroy the
gourd, calling up at the same time a hot dry east wind. And when Jonah once more
rages and wishes he were dead, God asks: how could you care about a gourd, which
came up in a night without your labor and perished in a night, and not care about my
other creations, those thousands of inhabitants of Nineveh?
As Meir Sternberg argues persuasively in The Poetics of Biblical Narrative, this story’s
operation depends upon our jumping to exactly the wrong conclusion, as we read
Chapter 1, about the reason Jonah is fleeing God’s mission – we assume that ‘‘Jonah is
too tender-hearted to carry a message of doom to a great city’’ (Sternberg 1985: 318).
Jonah’s honesty in admitting to the sailors that the tempest is his own fault, and his
self-sacrificial willingness to be thrown overboard for their sake, feeds this misunderstanding of his character. The information we are given at the start of Chapter 4
‘‘shatters the entire model of the narrative world and world view. . . . Not that the plot
and the participants are suddenly transformed but they are suddenly recognized for
Some Aspects of Biblical Narratology
287
what they always have been . . . ’’ (Sternberg 1985: 319). Sternberg asserts that the
story moves generically from ‘‘a punitive affair between God and Nineveh’’ to ‘‘a story
of a prophet’s education’’ (p. 320).
Sternberg’s reading is thus in radical contrast with the edifying readings demanded
by the various orthodoxies, which evade any contact with the reversal that follows the
reader’s discovery of Jonah’s motives. I suspect that the orthodox readings have already
hijacked the public meaning of the book of Jonah beyond all repair by wise narratologists, but even Sternberg’s reading may be more edifying than it needs to be. That
is, Sternberg posits that Jonah is educated by his experience, but since Jonah makes no
reply to God’s final question, there is not a shred of textual evidence that anyone but
the reader learns a thing. What explodes into view, in contrast to the mercy of God, is
Jonah’s narcissistic ruthlessness. Jonah raises none of the usual theological and moral
issues of the prophetic books, like Israel’s infidelity to God or their exploitation of
their fellow humans, and the only prophecy Jonah utters (‘‘Yet forty days and Nineveh
shall be destroyed’’) takes up exactly five Hebrew words.6 If this makes Jonah unique
among the prophetic books, it is because, as Tom Paine pointed out over two centuries
ago, Jonah is not a prophetic book at all but an antiprophetic book, a satire on
bloodthirsty prophets.7 Prophets are probably the most popular object of satire in the
Hebrew Bible, from Balaam in Numbers 22, cursing for pay, to the petty revenge of
Elisha on the children who mock his baldness in 2 Kings 2, and so here. It would not
be necessary for the author of Jonah to have had a specific prophet in mind, but if he
did, Joel might be a strong candidate.8 There are linguistic links between Joel and
Jonah – lots are cast in both books with unpleasant results for the winner – and
Jonah’s characterization of YHWH seems to come word for word out of Joel.
Be that as it may, the antiprophetic satire in Jonah is, if anything, more lighthearted and goofy even than Elisha’s bears or Balaam struck dumb by his speaking ass.
And it is the animals here too that contribute crucially to the outrageous atmosphere.
I’m thinking of that wonderfully excessive moment when, after the people have fasted
and put on sackcloth, after the King of Nineveh9 has himself fasted and put on
sackcloth, the King issues an edict that the fast shall include ‘‘man and beast, flocks
and herds, none shall taste food or drink water, all shall be covered with sackcloth and
cry mightily unto YHWH’’ (Jonah 3: 7–8). One imagines all too easily the miserable
bullocks and rams, bleating and bellowing their repentance. And it is to this scene
that the author of Jonah alludes when, in the final verse, he has God refer to the
creatures he has spared as ‘‘more than one hundred twenty thousand people who do
not know their left hand from their right hand, and also much cattle’’ (4: 11, my italics).
If this reading of Jonah is far from wild or surprising, if configuring the story this
way seems obvious, we need to explain the apparent reluctance, even among secular
scholars, to read Jonah as a satiric fable,10 as opposed to ‘‘a didactic story about a
prophet’’ (Day 1990: 39), or ‘‘the story of a prophet’s education’’ (Sternberg 1995:
320) or some other formulation. John Day questions ‘‘whether a distinct class of
literature consciously dubbed ‘satire’ existed in ancient Israel, unlike say ancient
Greece’’ (p. 39). But the issue cuts deeper than the lack of a biblical Hebrew word
288
David H. Richter
for ‘‘satire.’’ I suspect that the real problem with calling Jonah what it is, a satirical
fable, is that it implicitly classifies it as fiction. The allusion to Jonah in two of the
Gospels may make the matter even more sensitive, although for Jesus to compare his
mission to that of a fictional character well known from scripture would be by no
means absurd.11 Indeed one might think that the faithful would rejoice in not having
to defend the veracity of a story that features a man who survives for three days and
nights in the belly of a fish,12 or a Bronze Age city so large that it would take a person
three days to cross it on foot – roughly the size of Los Angeles County – or a God who
brings into existence on cue gourds, worms, and huge fish as well as several varieties
of bad weather.13 But characterizing a biblical book as fiction can enrage, with
something like Jonah’s rage, that large segment of Bible readers who believe in its
literal truth.14 And the particular satire in Jonah – on ruthless prophets of doom –
may be even more rebarbative to evangelicals for whom predictions of imminent
disaster can be pleasure as well as good business.
For me, insisting that biblical narratives must be nonfictional is far too limiting.
I think the world has a great deal to gain from the notion that some biblical narratives
that have been absorbed into the Deuteronomistic ‘‘sacred history’’ that runs from
Joshua to 2 Kings may at some point have been designed to be read as fictions. I have
argued that the story of the concubine of Gibeah that ends the book of Judges may
have originated as a lampoon against King Saul, that, for instance, the faithless
concubine from Bethlehem in Judah may represent David, who sold his services
first to King Saul and then to his enemy, the Philistine King Achish of Gath.15
Other stories in the book of Judges show similar patterns.16 And I would argue more
generally that, given how little we can know about the generic system that characterized Israelite narratives two or three thousand years ago, our default position
should be to assume that they had the same mix of fictional and factual genres we
find in other cultures of the period, including lampoons and satires, whether or not
they had names for these genres, as the Greeks did.
Similarly, I would argue that, absent discoveries not yet made about rules of
reading in use two or three millennia ago, we should be willing to assume that
textual juxtapositions that generate suspicious readings today might well have done
so in the intended authorial audience of historical texts.
Repetition
In 1 Samuel the relationship of Israel’s last judge, Samuel, and its first king, Saul, is
roiled by Samuel’s deep ambivalence about giving up the power he has wielded as the
charismatic leader of Israel for a position as second fiddle, the prophet channeling the
word of God for the direction of the anointed monarch. This ambivalence is expressed
directly, as when Samuel resists giving Israel a king, but it is also enacted in episodes
Some Aspects of Biblical Narratology
289
in which Samuel puts Saul into a double bind. If Saul interprets what Samuel has
commanded in the name of God in one way, he can be condemned for not interpreting
them the other way. Saul of course never understands this. Trusting in Samuel, he
always responds as a sinner grateful to have had his errors pointed out and eager for
forgiveness. We too can be trusting readers of the literal word of the text, just as Saul
is a trusting reader of Samuel, but to do that we would have to ignore what we already
know about Samuel’s rancor against monarchy in general and Saul in particular. So it
is difficult to avoid suspicious readings of these episodes.
That is, the trusting reader understands that Saul has shown his unworthiness to be
King of Israel by failing to carry out the will of YHWH not once but twice, but a
suspicious reader might understand that Samuel revenges himself upon the king he
never wanted to crown by placing him in a pair of double binds, once by giving him
inadequate instructions, and once by giving him too many contradictory instructions,
so that whatever he does he can be put in the wrong.17 Interestingly, the reader’s
dilemma – to read as the trusting reader or the suspicious reader – places him or her in
the same double bind in which Samuel places Saul, who is forced to infer an intention
from ambiguous language.
We also need to point out that the reader’s stance, in the hermeneutic duel between
Samuel and Saul, will be opposed to that of the character whose position we support.
That is, those who adopt with respect to the narrative the ‘‘trusting’’ reading
associated with Saul will agree with Samuel that Saul has transgressed and should
be deposed as king, while those who adopt toward the narrative the hermeneutic of
suspicion characteristic of Samuel (who refuses to take at face value the professions of
Saul) will be more likely to view Saul as the victim of Samuel’s machinations.18
If the reader of sacred history is likely to start in a trusting frame of mind, the mere
fact of the doublet may in itself foster a suspicious mode of interpretation. Just as
people who give us a whole series of reasons for taking a given action may thereby
suggest they are concealing the true one, the fact that there have to be two depositions
of Saul may make his inadequacy seem less plausible than if there had been only one.
Especially since, despite having been deposed twice, Saul continues to rule Israel for
what is apparently many years, until his death in battle some 15 chapters later.
Indeed, a son of Saul’s succeeds to his throne, so that it is not until several years after
Saul’s death that David comes into his kingdom, at 2 Samuel 5.
But at the next turn of the screw, the suspicious readings of 1 Samuel begin to raise
questions, not only about what the text says but about how the text was composed.
That is, unless we refuse to notice them, we are likely to become puzzled by the
endlessly reduplicated events of 1 Samuel. For there are not only two depositions of
Saul, those depositions follow three different acclamations of Saul as king, which
themselves follow at least two different versions of Samuel resisting the demand of
Israel to be ruled by a king. And following Saul’s deposition are two different stories
about how David becomes a part of Saul’s entourage, two daughters of Saul who
290
David H. Richter
become David’s brides and then are sundered from him, the two flights of David from
Saul, the two brotherhood oaths of David with Jonathan, the two approaches by David
to King Achish of Gath, the two occasions when Saul takes an army to hunt David,
who spares Saul’s life when he is most vulnerable, and so on, including two versions of
the death of Saul. Doublets occur elsewhere in the Bible – Genesis begins with two
inconsistent accounts of the Creation – but nowhere else are so many collected
together.19
Some of these doublets are events that could quite easily happen twice in succession. Saul could throw his spear at David and miss him on Tuesday and repeat the
same performance on Wednesday, could offer David marriage with one daughter in
March and another daughter in April. Other doublets seem clearly to be offered as
alternative stories: we are told one version of the death of Saul by the narrator of
Samuel, an authoritative story, though one that, taken literally, could have had no
possible witness; the other version is the eyewitness account of an Amalekite – and we
all know how reliable those Amalekites are – who is attempting to curry favor with
David by bringing him Saul’s crown and bracelet.
In the case of some doublets it is clear that an attempt has been made to reconcile
contradictions: the two stories about David’s first meeting with Saul are harmonized
at 1 Samuel 17: 15: ‘‘But David went and returned from Saul to feed his father’s sheep
at Bethlehem.’’ With this redactorial smoothing, David can be the shepherd boy
freshly come with a food parcel from Bethlehem even though we have already seen
David as the personal attendant of Saul, his musician, and his armor-bearer.20 But
other doublets defy any usual mode of reconciliation, such as the two very different
episodes that are presented as the origin of the proverbial saying ‘‘Is Saul too among
the prophets?’’
These peculiarities can be resolved in at least three ways. In one trusting reading,
we can become reconciled to the fact that we are in a world in which everything
happens twice, where we are not to be bothered by repetitions or apparent contradictions. Or we can naturalize the contradictions of the lifeworld by thinking of 1
Samuel as a nouveau roman like those of Robbe-Grillet, in which things happen once
but are narrated multiple times in inconsistent ways, a narrative that problematizes
the way the world of events can be mapped by language. Or we can naturalize the
contradictions yet another way, by viewing the authorship of 1 Samuel as residing in a
redactor who has gathered materials from sources that include alternative versions of
most of the significant events, and, helpless to decide which of two accounts may be
the true one, has fashioned a narrative that allows him to include both, but who
intends for us to notice the contradictions.21 As the history becomes less transparent
and as the conflicts between the various versions of events begin to focus the interest
of the suspicious reader upon the redactor rather than the agents within the narrative,
1 Samuel begins to turn into a precursor of Rashomon, or of the historical novel by
Faulkner that is only too aptly named (Absalom, Absalom!), where the interest in the
story of Thomas Sutpen shifts in the course of the various tellings from the actions of
Some Aspects of Biblical Narratology
291
the tragic or demonic figure at the center of the tale to the motivations of the various
narrators for constructing the inconsistent versions we are offered.22
Temporal Order
With the accession of David this elaborate parade of doublets comes to an end but not
the occasions for suspicious readings. In the case of 2 Samuel, some of these issues have
to do with when in David’s reign things are done, and by whom. I want to
acknowledge that not all the chronological anomalies generate questions. Meir
Sternberg (1990) has worked out the basic mechanisms of representing time in the
Bible, and most of the anomalies come under his general exceptions to the basic rule
of first things first.23
But just as there are doublets that are impossible to make sequential (like David’s
two first meetings with King Saul) there are episodes whose problematic chronological relation to the rest of 2 Samuel forces a reconfiguration of our understanding of
King David’s reign. Consider the dark episode recounted in 2 Samuel 21. After three
years of famine David asks God why Israel is being punished, and is told that it is on
account of the bloodguilt on the House of Saul, who slew or attempted to slay the
Gibeonites out of zeal for Judah and Israel. It may seem strange that the episode of
Saul slaying the Gibeonites was not mentioned in its proper place during his reign,
stranger that God should have caused a famine in Israel for the sake of mere
Gibeonites,24 and that the famine should be delayed until what looks like late in
the reign of David, and strangest of all that atoning for Saul’s crime should involve
executing his entirely innocent descendents. But after a brief negotiation, King David
agrees to hand over seven blameless men to the Gibeonites who crucify them in Saul’s
home town of Gibeah and leave their bodies for the birds to peck at: two sons of Saul’s
concubine Rizpah bat Aiah and five sons of Michal, Saul’s daughter, David’s wife.25
But one stray remark during this revolting vengeance – that David ‘‘spared
Mephibosheth, the son of Jonathan the son of Saul’’ (21: 7) – gives us a clue about
the time frame by linking this episode to Chapter 9: 1, where David asks, ‘Is there yet
any that is left of the house of Saul, that I may show him kindness for Jonathan’s
sake?’ David’s question, isolated and unmotivated where it sits, leads to David
making generous provision for Mephibosheth at the royal court. But that scene
only makes sense if David had already sent off every other male descendent of Saul
to the Gibeonites for execution.
Once we make this connection, we see that the Gibeonites’ vengeance at the end of
2 Samuel is chronologically out of place. It doesn’t come late in David’s reign over
unified Israel, as we might expect from its textual position, but in its earliest days.
But seeing this raises two other questions. The easy question is what rhetorical
purpose is served by telling it where it is told, in Chapter 21. David has himself
suffered grievously from the deaths of the nameless child of Bathsheba, his firstborn
292
David H. Richter
Amnon, and his favorite Absalom. Lately he barely survived the civil war fomented by
his own son. Coming after all this, the horror of David’s action, undertaken as it is for
the sake of public welfare, is moderated. Told in its proper position, at the beginning
of David’s reign, it would have a different valence. 2 Samuel begins with what seem to
be a series of lucky breaks that leave David on the throne with his hands relatively
clean. He has profited by the deaths in battle of Saul and Jonathan, and the
assassinations of Saul’s general Abner, and Saul’s son and successor Ishbosheth, but
he has had nothing personally to do with these deaths. Told here, the Gibeonite
episode would expose the bloodbath with which David’s reign began, the way he got
rid of every male descendent of Saul except for the lame, and thus harmless,
Mephibosheth.
The harder question is why it was included at all, since this is not an episode that
anyone would miss if it were not here. Its presence suggests that the redactor26 wants
to force the audience to reconfigure retroactively their understanding of David’s
character and reign just as it comes to a close. Those whose minds are shaken by
the Gibeonite episode may be struck dumb by the single shocking verse later in the
same chapter (21: 19), stuck amid a catalog of the mighty acts of David’s warriors:
‘‘There was again war with the Philistines at Gob; and Elhanan . . . the Bethlehemite
slew Goliath of Gath, the staff of whose spear was like a weaver’s beam.’’ This verse has
evident textual corruptions27 but it seems to state that the great deed of the lad with
slingshot and five smooth stones, that immortal moment from 1 Samuel 17 we can all
envision, may have happened at some other time and may have been accomplished by
somebody else entirely. A summary verse at the end of the chapter tells us that all the
monstrous enemies of Israel ‘‘fell by the hand of David and his servants’’ as though
every legendary deed somehow belongs to the monarch, no matter who actually
performed it, as though legend must triumph over history.28 But saying so deflates
legend, and indeed our last look at David, in the second chapter of Kings, shows the
decrepit former warrior, shivering in bed, croaking out his final instructions to
Solomon, a mixture of general pious cant about obedience to the commandments of
YHWH and a specific list of the people he wants his son to dispose of after he is gone.
It’s a chilling farewell, disgusting in its hypocrisy and more horrifying than anything
in The Godfather.
Biblical Narrative vs. Bible Stories
Given the temptations to suspicious readings and the ungovernable complications
that arise out of such readings, one can well understand why the religions that have
adopted Samuel and Jonah as two of their holy books have avoided reading the text as
a continuous narrative but have instead read it as pericopes (as biblical scholars call
the chunks of text that can be extracted for study or for preaching). Individual
pericopes can be coherent and morally edifying, as Jesus found the book of Jonah,
even though the narrative as a whole is not. We know Bible stories, that is, rather than
Some Aspects of Biblical Narratology
293
biblical narrative, because Bible stories are safer things, safer both for Jews who look
to David as the model of the righteous king whose lineage will return in the end of
days, and for Christians who look to David as one of the antetypes of Christ the King,
who came and who will come again. Within these broader cultural narratives, the
tragic story of Saul, doomed by the malice of Samuel and the contradictions of his own
nature, and the horrific tale of David, God’s anointed monster, have become stories we
can read as wholes only with the greatest difficulty.
NOTES
1 I am using ‘‘factual’’ to cover stories meant to
be taken for true, regardless of actual truth
(including lies and propaganda).
2 This picture can be complicated if one goes
into the midrashic details. The early midrash
Mekilta de Rebbe Ishmael argues at Pesiqta Bo
that Jonah was trying to protect the Israelites
from God’s anger. If the Ninevites heeded
Jonah’s prophecy and repented, would not
God punish the Israelites who ignored God’s
prophets even more severely for their sins? The
Mekilta considers Jonah wrong to defend the
honor of the child (Israel) against the honor of
the father (God), but his flaw lies in his priorities, not in his character. Ibn Ezra’s commentary (1951) defends Jonah by thematizing
the heavy burden of prophecy, comparing
Jonah’s reluctance to go to Nineveh with
Moses’s reluctance to return to Egypt and
take the Israelites out of slavery.
3 For example, Faussett, Jamieson, and Brown’s
nineteenth-century commentary emphasizes
the multiple allegorical dimension of Jonah
as ‘‘living symbol’’: Jonah’s engulfment by
the fish is ‘‘the emblem of death’’ which is ‘‘a
present type to Nineveh and Israel of the death
[of the soul] in sin, as his deliverance was of
the spiritual resurrection of repentance.’’ And
in addition to the symbolic significance of
Jonah to his contemporaries, he is a ‘‘future
type of Jesus’s literal death’’ as atonement ‘‘for
sin and resurrection by the Spirit of God.’’
4 Qur’an, Sura 10 (Yunus): 96–8; Sura 21
(The Prophets): 87–8; Sura 37 (The Rangers):
139–48; Sura 68 (The Pen): 48–50.
5 The phraseology here goes back to pre-exilic
documents like Exodus 34 (and also found in
6
7
8
9
Nehemiah 9 and Psalms 86, 103, and 140),
but it exactly parallels Joel 2: 13, especially
with the key phrase ‘‘v’nachem al ha-ràah’’
which is found nowhere else. (The verse
goes: ‘‘And rend your heart, and not your
garments, and turn unto the LORD
your God: for he [is] gracious and merciful,
slow to anger, and of great kindness, and
repenteth him of the evil.’’) This link may be
more than just coincidence – Jonah may be
satirizing Joel.
‘‘Yet forty days and Nineveh shall be
destroyed’’; in Hebrew: `od arba’im yom v’Nin’
ve nehapecheth.
See The Age of Reason (1795), Part II, chapter 1:
As the book of Jonah, so far from treating of
the affairs of the Jews, says nothing upon that
subject, but treats altogether of the Gentiles,
it is more probable that it is a book of the
Gentiles than of the Jews, and that it has
been written as a fable to expose the nonsense, and satirize the vicious and malignant
character, of a Bible-prophet, or a predicting
priest.
Joel’s subject emerges as a final battle among
the nations after which those like Egypt and
Edom that have plundered Israel, that have
‘‘cast lots for my people, given a boy for a
harlot, sold a girl for wine’’ ( Joel 3: 3) shall
be judged and punished, while Judah and
Jerusalem shall be restored to fruitfulness and
beauty. The dates of both Jonah and Joel are
held by the orthodox to be pre-exilic, but
secular scholars are more apt to place both
books in the postexilic period.
An off-kilter title like ‘‘The Queen of London’’
or ‘‘The President of Cincinnati.’’
294
David H. Richter
10 Contemporary secular scholars who have configured Jonah as a satire include E. M. Good
(1965), M. Burrows (1970), J. C. Holbert
(1981), and J. S. Ackerman (1981).
11 No more strange rhetorically than for the
newly elected governor of California to
speak of his political program in terms of
characters he has enacted in the movies.
12 Ibn Ezra (1951) ambiguously refers to
Jonah’s survival in the fish as a ‘‘neis,’’ which
literally means ‘‘sign’’ – either a miraculous
sign from God or, as I think likely, a symbolic event not meant to be taken literally.
13 Terry Eagleton (1990: 236) refers to
YHWH’s performance as being like a magician ‘‘having a ropy night on a stage in
Blackpool.’’
14 There is agreement even among the orthodox
and the fundamentalist Christians that the
Bible contains meshalim (parables), stories
whose literal truth is not guaranteed, as
long as these pericopes are bracketed off
from the main narrative, ascribed to an individualized speaker, and given a rhetorical
situation (e.g., Jotham in Judges 9, Nathan
in 2 Samuel 12, the Wise Woman of Tekoah
in 2 Samuel 14, as well as the many parables
of Jesus in the gospels). The problem with
reading Jonah even as a didactic fable is that
none of these features obtains here.
15 I argued (Richter 1999) that other narrative
elements of the story travestied or satirized
major events of Saul’s reign: his cutting up of
oxen to initiate his first campaign, his insistence on risking the entire tribal army to save
the frontier town of Jabesh-gilead while more
important areas of Canaan were not under
Israelite control, his superstitious reliance
on oracles, his listless lingering under the
pomegranate tree before the battle against
the Philistines, his fatal combination of rashness and inanition.
16 For example, the story of Micah in Judges
17–18 is an alternative version, almost cartoon-like in its extravagant breaches of both
secular morality and sacred tradition, of how
an alternative temple with a graven image
happened to wind up at the northern boundary of Israel at Dan. The story of Jephthah’s
rash vow in Judges 11 (to sacrifice to
YHWH whatsoever comes out to meet him
on his return home from battle) parallels
Saul’s rash vow to fast before and during the
battle of Michmash and to slay anyone who
breaks the fast, a vow that, except for the
intervention of the people, should have
resulted in the execution of Jonathan.
Jephthah’s obscure origins in Gilead link
him to Saul who has an equally obscure connection to Gilead on the other side of the
Jordan. The stories of both Gideon and Samson may allude to the monarchy under Solomon. Like Solomon, Gideon amasses huge
amounts of gold, enjoys many wives and
fathers dozens of children, making an ephod
which leads the Israelites to idolatry, and has
a royal reign of 40 years. Samson, like
Solomon, has a weakness for wine and for
non-Israelite women who lead him into sin;
Samson destroys a heathen temple while
Solomon builds the temple to YHWH.
17 Saul is given the contradictory instructions at
1 Samuel 10: 7–8: verse 7 tells him to ‘‘do as
occasion serve thee’’ while verse 8 tells him to
wait seven days for Samuel, who will come to
perform sacrifices before Saul’s battle with
the Philistines. Saul waits the seven days,
but since Samuel has not appeared, and
since the army with which he is supposed
to defeat the Philistines is deserting before
his eyes, he follows the ‘‘do as occasion serve
thee’’ part of the instructions and performs
the sacrifices preliminary to the battle by
himself. No sooner has he done so than Samuel appears, and tells Saul that he has already
blown his mission and the monarchy:
‘‘Thou . . . hast not kept the commandment
of the LORD thy God . . . [N]ow thy kingdom shall not continue: the LORD hath
sought him a man after his own heart, and
the LORD hath commanded him to be captain over his people, because thou hast not
kept that which the LORD commanded
thee.’’
The inadequate instructions are in 1 Samuel 15, where Saul is commanded to smite
the Amalekites and ‘‘devote them to destruction’’ (hecharamtem). Saul defeats and destroys
the Amalekites, slaying all except the king,
coming in triumph with the king and the
Some Aspects of Biblical Narratology
best of the cattle ‘‘to sacrifice to the Lord thy
God at Gilgal.’’ Samuel treats this as a poor
excuse for taking spoils in a holy war –
implying that Saul should have killed the
king and slain all the cattle right on the
battlefield – and deposes Saul for the offense.
But if that is what Samuel meant, perhaps he
should have said so, since Joshua is described
as doing exactly what Saul did (with full
apparent approval by the Deuteronomic narrator) at Joshua 11: 10–14, after which
Joshua and the people return, just as Saul
has now done, to Gilgal.
18 This chiastic relation between characters and
their hermeneutic has been pointed out before in Shoshana Felman’s (1977) analysis of
The Turn of the Screw and by Barbara Johnson’s
(1985) analysis of Billy Budd.
19 Some of the doublets in the story of King
Saul – the references to 1 Samuel unless
otherwise marked:
The Israelites demand a king; Samuel
harangues them on their sin in so doing: 8:
5–20;
The Israelites having demanded a king are
harangued by Samuel on their sin in so
doing: 12: 6–18
Saul chosen as ruler (nagid) by the LORD: 9:
16, Saul anointed as ruler (nagid) 10: 1
Saul chosen as king (melech) by the divination
ritual of the ephod: 10: 26
Saul is made king by the people (yimlichu):
11: 15
Saul
Saul
Saul
Saul
prophesies, the origin of the saying: ‘‘Is
among the prophets?’’: 10: 11–12
prophesies, the origin of the saying: ‘‘Is
among the prophets?’’: 19: 23–24
Saul deposed as king by Samuel before the
battle of Michmash: 13: 8–14
Saul deposed as king by Samuel after the
battle with the Amalekites: 15: 16–31
David joins Saul’s court as musician and
armor-bearer: 16: 18–21
David joins Saul’s court as warrior after slaying Goliath: 17: 57–58
Saul throws a spear at David but misses:
18: 10
Saul throws a spear at David but misses:
19: 10
295
Saul throws a spear at Jonathan but misses:
20: 33
Saul gives David his daughter Merab as his
wife: 18: 17; then gives her instead to Adriel:
18: 19
Saul gives David his daughter Michal as his
wife: 18: 27; then gives her instead to Phalti:
25: 44
Saul’s daughter helps David flee Saul’s assassins: 19: 11–12
Saul’s son helps David flee Saul’s assassins:
20: 41–2
Fleeing from Saul, David seeks the protection
of King Achish of Gath: 21: 10
Fleeing from Saul, David seeks the protection
of King Achish of Gath: 27: 2–3
Saul, pursuing David but in a vulnerable
position, is spared by David, who cuts his
robe: 24: 3–4
Saul, pursuing David but in a vulnerable
position, is spared by David, who takes his
spear: 26: 12
Saul, wounded, asks his armor-bearer to slay
him, but is forced to kill himself: 31: 3–4
Saul, wounded, asks a passing Amalekite to
slay him, who does so: 2 Samuel 1: 9–10
20 See 1 Samuel 16: 18–23. This reconciliation
only partly reconciles, it would seem, for in
the aftermath of the slaying of Goliath in
Chapter 17, Saul asks after David’s identity
as though he had never laid eyes on the lad
before. The harmonizer has not entirely gone
on strike, though, since there is a final note at
1 Samuel 18. 2 that ‘‘Saul took [David] that
day and would not let him go back to his
father’s house.’’
We might speculate that the language in
which Saul asks Abner about David (‘‘Who is
the father of that lad?’’ at 17: 55, repeated
literally at verses 56 and 58) may be deeply
meaningful in context, because Saul has been
told that his successor as king has already
been chosen and will not be of his own
family. Given that Saul has already met
David twice for the first time, his insistent
question about the identity of David’s father
may be taken metaphorically, as expressing
the desire to be the father of David, which he
accomplishes, in the only way he can do so
296
David H. Richter
retroactively, by marrying one of his daughters to David. Saul’s insistent question
‘‘Mi’’ (who?) as a form of wishful thinking
may be connected to the biblical idiom
‘‘Mi yiten li’’ – literally ‘‘Who will give me’’
but best translated as the optative ‘‘Would
that I were . . . ’’) This posited desire of Saul’s
to be David’s father is of course highly ambivalent, as one might expect from a man
welcoming his already anointed successor
into the royal family: Saul first offers his
elder daughter Merab, then gives her to
Adriel, then his younger daughter Michal,
who becomes David’s first wife until she too
is taken away from David after his defection
from Saul’s court and given to Palti ben
Laish. Nevertheless Saul several times refers
to David at key episodes as ‘‘my son’’ (24: 16,
26: 17 and 25) but David reciprocates by
calling Saul ‘‘my father’’ only once, at 24:
11. We may want to question just how affectionate David’s expressed filial attitude toward Saul is, because in the next chapter,
when David is shaking down a rich farmer
named Nabal for blackmail, David has his
messengers refer to him as ‘‘thy son David’’
(1 Samuel 25: 8).
21 Many biblical commentators, including
Robert Alter (2000), have argued that readers
knew that the historical books of the Bible
contained alternative versions of events and
that they were read the way we would read a
documentary political history that gives space
to different accounts whose contradictions
cannot be resolved. It is tempting to think
that people a few millennia ago read Samuel
differently from the way we do, but the actions
of the Samuel redactor in attempting to reconcile discrepant accounts would seem to
work against this view. But we cannot tell at
what stage prior to the tenth-century Masoretic recension these reconciliatory comments
were inserted into the text, and so Alter could
be right: the original historian may have assumed the ‘‘alternative version’’ mode of reading while a later redactor, after this mode of
reading had passed away, sensed a genuine
contradiction and inserted glosses. Harmonizing the contradictions in the book of Samuel
begins as early as the book of 1 Chronicles.
One keen reader of 1 Samuel, Ken Gordon,
has suggested to me (personal communication) that the doublets are not placed in the
text at random but function to highlight the
episode that separates them. The two depositions of Saul, for example, flank the episode
in which both Saul’s heroism and his quixotry hit their high point at the battle
of Michmash. The two episodes in which
David spares Saul when he has him in
his power flank the episode of David
and Nabal, which highlights David’s
crude extortions as a local warlord. But sometimes doublets come so far apart that a doublets-as-highlighter theory seems difficult to
sustain.
22 In ‘‘The Grand Chronology’’ Meir Sternberg
feels that, while from the genetic perspective
of source criticism, doublets present alternative versions of a single event, ‘‘when
we switch from the genetic to the poetic
perspective . . . each compels . . . a successive
reading. In the finished narrative they
have all been linearized into stages within a
well-plotted chronology’’ (Sternberg 1990:
126).
23 For example, it is during the siege of the
Ammonite city of Rabbah that King David
begins his affair with Bathsheba, and it is
under the wall of Rabbah that her husband
Uriah the Hittite perishes. Then the domestic story shifts to David’s marriage to Bathsheba, Nathan’s condemnation, the king’s
repentance, the death of the couple’s first
son, and the conception and birth of Solomon. Finally, Joab sends David word that
Rabbah has been taken and advises him to
be present at the victory. Are we to think that
Joab and his army have been waiting under
the walls of Rabbah throughout the many
months of Bathsheba’s two pregnancies?
Sternberg rightly suggests that the chiastic
structure is formal rather than chronological,
that ending the war story is a way of
indicating that the domestic drama of
David’s household, as far as it has come, is
complete.
24 The Gibeonites first appear in Joshua 9,
as Canaanites who pretend to be allies of
the Israelites from a distant land, and make
Some Aspects of Biblical Narratology
a covenant with them. When found out,
they cannot be slain because of the treaty
but are turned into a permanent underclass –‘‘hewers of wood and drawers of
water.’’
25 Thus the Masoretic text, but this is inconsistent with Michal’s barrenness stated at
2 Samuel 6: 23. Editors speculate that
‘‘Michal’’ is a scribal mistake here for
‘‘Merab’’ her elder sister, whom Saul married
to Adriel, who is named here as the father of
those children, and as Merab’s husband at 1
Samuel 18: 19. Another conjecture is that,
despite the word ‘‘bore’’ here (yaldah), Michal
for some unstated reason had raised her sister’s children and would for that reason have
been considered their mother. The King
James Bible and other translations translate
what the Gibeonites did to the descendents
of Saul as ‘‘hanged,’’ but the verb v’hoqàanum
refers to a slow method of execution involving impalement, where the body once dead
remains exposed, so I have translated it perhaps anachronistically as ‘‘crucified’’ (the
Vulgate has ‘‘crucifixerunt’’).
26 Many scholars argue that the so-called 2
Samuel epilogue (Chapters 21–4, containing
two narrative episodes, two lists of David’s
warriors, and two Davidic psalms, arranged
in chiastic order) is from a different ‘‘hand’’
than the rest of Samuel. From a source-critical point of view, it is often seen as shoveling
a batch of unrelated documents and episodes
at the end of a book, just before the narrative
close, which is also done elsewhere in the
Bible (for example the end of Deuteronomy).
The narrative close of 2 Samuel actually
comes at 1 Kings 2: 12, when the succession
of Solomon is assured. Robert Alter, who is
generally dismissive of source-critical readings, and insistent on taking a redactorial
standpoint, here finds he needs to reverse
course and discuss ‘‘alternate source’’ theories
because of the harm these episodes do not
just to the unity but to the coherence of his
‘‘David Story.’’
27 Elhanan’s patronymic here (son of Yaare-oregim) has somehow picked up the word for
‘‘weavers’’ (oregim) repeated in ‘‘weavers’
beam’’ further down in the verse; later in a
297
catalog of David’s 30 picked warriors, the
same Elhanan is referred to as Elhanan ben
Dodo of Bethlehem. Chronicles picks up
the story and revises it so that Elhanan
kills Goliath’s brother Lahmi (1 Chronicles
20: 5), a name that seems to be made up
(from Elhanan’s birthplace – in Samuel Elhanan is a Bethlehemite, in Hebrew ‘‘beth-halahmi’’). The revision in Chronicles underlies
the King James Version’s translation of the
verse, which goes ‘‘Elhanan . . . slew the brother
of Goliath the Gittite.’’ (The italics
on the words ‘‘the brother of’’ mean that they
are supplied by the translators but are missing
from the original text.) Some rabbinic commentators including Rashi rescue the situation by claiming that ‘‘Elhanan’’ (meaning
‘‘God is gracious’’) is the throne name of
David.
28 The 3,000th anniversary of the birth of
David, celebrated recently, has been the occasion for dozens of ‘‘debunking’’ biographies
(e.g., Steven L. MacKenzie’s (2000) King
David: A Biography; Baruch Halpern’s
(2001) David’s Secret Demons, David Rosenberg’s (1997) The Book of David). What I find
endlessly amusing about these books is their
common stance of unvarnished truth – that
they will heroically tell the real story about
David that the book of Samuel doesn’t tell.
But aside from a few factual references that
can be inferred from early psalms, and the
highly complimentary account of David in
Chronicles (which whitewashes David by
simply eliminating any story that shows his
blemishes), there just aren’t any sources
about David except the one in Samuel. And
all the ugly things the biographers tell about
David that he began his career as a freebooter
who lived by shaking down local landowners,
that he hired himself and his guerrilla band
out to the Philistines, the national enemy,
that he came to power by getting rid of
Saul and all of his descendents, that he was
an adulterer and a murderer, that he was a
terrible father, husband, and leader – all these
things are plainly written in the book of
Samuel for those who know how to read it,
and few of them need to be read ‘‘between the
lines.’’
298
David H. Richter
REFERENCES
AND
Ackerman, J. S. (1981). ‘‘Satire and Symbolism
in the Song of Jonah.’’ In B. Halpern and
J. Levenson (eds.), Traditions in Transformation
(pp. 214–46). Winona Lake, WI: Eisenbrauns.
Alter, R. (2000). The David Story: A Translation
with Commentary of 1 and 2 Samuel. New York:
W. W. Norton.
Burrows, M. (1970). ‘‘The Literary Character
of the Book of Jonah.’’ In H. Thomas Frank
and W. L. Reed (eds.), Translating and Understanding the Old Testament (pp. 80–107). New
York: Abington Press.
Day, J. (1990). ‘‘Problems in the Interpretation of
the Book of Jonah.’’ In A. S. Van de Woude
(ed.), Quest of the Past. Studies on Israelite Religion,
Literature and Prophetism (pp. 32–47). Leiden:
Brill.
Eagleton, T. (1990). ‘‘J. L. Austin and the Book of
Jonah.’’ In R. Schwartz (ed.), The Book and the
Text: The Bible and Literary Theory (pp. 231–6).
Oxford: Blackwell.
Fausset, A. R., Jamieson, R., and Brown, D.
(1871). A Commentary, Critical and Explanatory,
on the Old and New Testaments. Hartford, CT:
Scranton. <http://blueletterbible.org/Comm/
jfb/Jon/Jon001.html>
Felman, S. (1977). ‘‘Turning the Screw of Interpretation.’’ Yale French Studies 55–6: 94–207.
Good, E. M. (1965). Irony in the Old Testament.
Philadelphia: Westminster Press.
Halpern, B. (2001). David’s Secret Demons: Messiah,
Murderer, Traitor, King. Grand Rapids, MI:
Eerdmans.
Holbert, J. C. (1981). ‘‘Deliverance Belongs to
the Lord: Satire in the Book of Jonah.’’
Journal for the Study of the Old Testament 21,
59–81.
FURTHER READING
Ibn Ezra, Avraham ben Meir (1951). Commentary
on Jonah in Mikraot Gedolot: Nevi’im u-Ketuvim.
New York: Pardes.
Johnson, B. (1985). ‘‘Melville’s Fist: The Execution of Billy Budd.’’ In The Critical Difference.
Essays on the Contemporary Rhetoric of Reading
(pp. 79–109). Baltimore and London: Johns
Hopkins University Press.
McKenzie, S. L. (2000). King David: A Biography.
New York: Oxford University Press.
Mekilta de-rebbe Ishmael (1961). ed. and trans.
J. Lauterbach, 3 vols. Philadelphia: Jewish Publication Society of America.
Paine, T. (1794–5). The Age of Reason.
<http://libertyonline.hypermall.com/Paine/AORFrame.html>
Rabinowitz, P. J. ([1987] 1998). Before Reading:
Narrative Conventions and the Politics of Interpretation. Columbus: Ohio State University Press.
Richter, D. H. (1999). ‘‘Farewell My Concubine:
The Difficult, the Stubborn, and the Outrage of
Gibeah.’’ In M. L. Raphael (ed.), Agendas for the
Study of Midrash (pp. 101–22). Williamsburg,
VA: William and Mary Press. <http://www.
qc.edu/ENGLISH/Staff/richter/concubine.
html>
Rosenberg, D. (1997). The Book of David. New
York, Harmony.
Sternberg, M. (1985). The Poetics of Biblical Narrative: Ideological Literature and the Drama of
Reading. Bloomington: Indiana University
Press.
Sternberg, M. (1990). ‘‘Time and Space in Biblical
(Hi)story Telling: The Grand Chronology.’’ In
R. Schwartz (ed.), The Book and the Text: The
Bible and Literary Theory (pp. 83–145). Oxford:
Blackwell.
19
Why Won’t Our Terms Stay Put?
The Narrative Communication
Diagram Scrutinized and
Historicized
Harry E. Shaw
I would hazard the guess that few teachers and few students of narratology or narrative
poetics have failed to use the Narrative Communication Diagram, in one of its
familiar guises or another – for instance, the one proposed by Seymour Chatman
(Chatman 1978: 151) many years ago, which runs as follows:
Real Author ! Implied Author ! (Narrator) ! (Narratee) ! Implied Reader ! Real Reader
Versions of the diagram differ, but in whatever form, it has proven its usefulness.1
Perhaps for that very reason, it seems fitting to subject it to some scrutiny. I believe
that the diagram has gained wide acceptance and use in part because its terms are
malleable. Critics with quite different views of narratology and of narrative itself can
avail themselves of the diagram, simply by giving its various terms rather different
meanings.
Some, echoing a familiar criticism of narratology, might conclude that such critics
are simply finding devious ways around a system that is too abstract and constraining.
This, however, is not my own view, even though I do not hope to find a way of ironing
out the differences and coming up with a definitive, all-purpose diagram. My hope
instead is to uncover the conditions of possibility for disagreements, overt and latent,
about the diagram and in doing so to suggest what is at stake in different ways of
imagining its terms. Along these lines, I’ll make two local suggestions: first, that
users of the diagram bring to it two different implicit models of the communication
situation; and second, that the terms the diagram seeks to describe necessarily become
hazier as we move from left to right. On a more general level, my argument will be
that we are likely to imagine the diagram in one way and not another depending on
our pre-existing beliefs about some large and fundamental matters involving the
300
Harry E. Shaw
scope of narratology itself, and also – and this is the heart of my argument – on the
kind of fiction we think of as normative. (To say this is to say that our use of the
diagram turns out to be historical in several senses.) I will therefore conclude my
essay, not with a definitive version of the model, but with the admission that the
version of the model I recommend in these pages reflects my own critical priorities on
both these scores.
The notion of a narrative communication diagram evokes a simple underlying
image, the image of someone telling a story to someone else. But this image is
susceptible to different emphases, and I think that in practice those who have
developed and employed its components have tended to rely on two significantly
disparate ways of imagining what’s central about situations in which communication
occurs. If you observe two people conversing, you can take an objective stance in
which you concentrate on what they are saying, on who is talking and who is
listening, on whether the person who is telling the story is recounting his or her
own experience or that of another (and in the latter case, if he or she does so by
enacting the other person’s words and actions or summarizing them), and so on. We
might call such an emphasis, which looks on from the outside and gauges the flow
and the parameters of the narrative situation, an emphasis on the modality of
‘‘information.’’ (Here I echo Genette, who consistently speaks in terms of the transfer
or flow of information in his seminal discussions of narrative focalization.) The
external observation that characterizes the information perspective begins with immediate perceptions, but it need not be confined to them. We might, for instance,
wish to go beyond immediate perceptions in an attempt to assess the reliability of a
tale-teller by measuring what he or she says against other evidence we observe in the
text.
It would, however, also be possible to engage differently with two figures in a
story-telling situation in a different way. Instead of focusing on the flow of information as seen from outside, we might adopt a role familiar to all of us when we
ourselves participate in such situations, the role of imagining what the person
addressing you has in mind as he or she tells you the story, what effects and purposes
the teller wishes to achieve. Does the story-teller seem to be crafting a version of the
story designed to appeal to the particular listener being addressed? Might the storyteller even be attempting to inveigle the listener into adopting a certain kind of role
and thereby acquiescing in certain beliefs and values? And what do the story-teller’s
intentions reveal about his or her own values? When we connect with the narrative
situation in this way, we have moved from issues of information to issues of ‘‘rhetoric.’’
Both approaches are and must be based on the reading of signs; the latter, however,
attempts to use its reading of signs to reconstruct and indeed to inhabit a world of
internality.
I believe that the terms contained in the communication diagram that have caused
most controversy have done so because their nature changes according to whether one
thinks of them in terms of information or of rhetoric. The most dramatic example is
the category of the implied author. This category seems clearly to belong to the
The Narrative Communication Diagram
301
rhetoric mode, and not only because it was invented (by Wayne Booth) in a book
entitled The Rhetoric of Fiction. The implied author is a critical reconstruction of the
mind behind the rhetorical purposes informing a work of narrative. It serves to focus
the question ‘‘What would you have to have believed and valued to have created this
work?’’ by constructing an anthropomorphic agent of belief, and thereby invites those
who use it to employ the vast and subtle panoply of techniques, most of them
unformulated and hardly conscious, that we draw upon in everyday life when we
try to assess the values and intentions of others. It has the particular pedagogical
virtue of allowing this cornucopia of hermeneutic expertise to be focused on a work of
art, but at the same time preventing it from flowing over (as it would naturally be
inclined to) into a scrutiny of the agent we all know actually made the work of art,
namely its author; instead, the focus stays on the work itself. Given these manifest
virtues, it is unsurprising that the category of the implied author has gained wide
acceptance, and it is hard to feel that there isn’t something perverse about the
insistence of Genette (and others) that it is a mere redundancy. But if we shift from
issues of rhetoric to issues of information, the perversity vanishes, for there it is
redundant. We never, the information argument would go, read the implied author’s
words, or hear his or her voice telling the story. (I myself believe that this is not
entirely true, but I work from the rhetoric perspective.) When we assess, say, the
narrator’s reliability, the information critic would continue, we respond to all kinds of
objective cues we hear in the narration. There is, however, no necessity for creating an
anthropomorphic source for these cues and for any discrepancies we find among them:
we can simply note and assess them, without having to concoct a special agent from
which they proceed. We know that someone wrote the story; we know that we are
hearing a narrator tell the story. That (if you inhabit the information camp) is all we
know, and all we need to know. Thus if we think for the moment only of the terms
that fall on the left side of the narrative communication diagram, we can see that the
implied author is different in kind from the others. The real author and the narrator
harmonize with both the information and the rhetoric views of the diagram. The
implied author fits only the rhetoric view.
I move now to a perturbing force that affects the diagram as a whole in a more
systematic way, a force described by what I have come to call ‘‘Genette’s Law.’’ (I call it
a ‘‘law’’ to stress the fact that what is at issue here depends not on the presuppositions
one brings to the diagram, but on an inherent property of the diagram itself.) Genette
identifies what he calls ‘‘the vectorality of narrative communication. The author of a
narrative, like every author, addresses a reader who does not yet exist at the moment
the author is addressing him, and who may never exist. . . . the implied reader is the
idea, in the real author’s head, of a possible reader’’ (Genette 1988: 149). The
implications of this seemingly innocuous, commonsensical observation are significant, for it tends to destroy the symmetry of the communication diagram, rendering
‘‘audience’’ terms on the right side of the diagram less substantial than ‘‘narrator’’
terms on its left side. It is amusing that Genette’s own formulation, made from the
information camp, echoes the hallmark of what I’ve called the rhetoric camp, by
302
Harry E. Shaw
raising the question of what is ‘‘in the head’’ of a narrative entity – to be sure with a
difference, since entering the head is viewed an inescapable plight, not a useful and
illuminating hermeneutical procedure. Such is the power of a general perspective to
alter the force of individual terms and locutions.
Genette’s Law suggests that the items in the narrative communication diagram will
become more diffuse as we move from left to right. How should we act on this
insight? I believe that it is possible to obtain a high degree of solidity and specificity
for the terms ‘‘narratee’’ and ‘‘implied reader’’ if we define those terms from the
information perspective, but that the effects of Genette’s law cannot be evaded from
the rhetoric perspective. Turning first to the information perspective, it is possible to
stabilize both the terms ‘‘narratee’’ and ‘‘implied reader’’ by severely narrowing their
scope. When narrowed in this way, the term ‘‘implied reader’’ becomes (as I think it
already is for many analysts) a way of specifying matters buried so deeply in culture
that they precede and undergird the realm of the conscious persuasion that characterizes the implied author. What level of linguistic competence does the narrative
seem to assume? What level of cultural knowledge? What set of shared values? Talk
about the implied reader would in this scheme of things contain a large element of the
empirical: an attempt to define the implied reader in the classroom might well begin
by suggesting that implied readers of older texts would already know the lore
provided for modern students in the footnotes of modern editions.
The effect of restricting the narratee to the realm of objectively observable information would be more dramatic. Under such a scheme, narratees would need to be
specifically described recipients of a narrative, palpably addressed by the narrator; a
narratee could not simply be a figure the narrator has in mind in telling the story. Two
kinds of narratees would then emerge. One would occur in frame tales, where one
character acts as a narrator and addresses another character, who thereby becomes the
narratee. The other would occur when a narrator suddenly addresses someone specific,
as when the narrator in Tristram Shandy evokes a narratee by exclaiming, ‘‘How could
you, Madam, be so inattentive in reading the last chapter?’’ (chap. 20). This sort of
narratee is often a ‘‘scapegoat’’ narratee, singled out to exemplify traits the actual
audience should avoid.
If we turn to the rhetoric perspective, the scene changes: we will find it impossible
to stabilize and concretize the terms narratee and implied reader by restricting their
meaning. (At least, I haven’t been able to find a way to do so.) At first blush, this
seems disconcerting. After the initial shock, though, it turns out not to be a problem,
and indeed not to be surprising. This is because from the rhetoric perspective, what
we’re interested in anyway occurs in the mind of the implied author (or sometimes, as
I’ll argue more fully in a moment, in the mind of the narrator). To be sure, it is crucial
that we keep in mind the possibility that a narrative may be addressed to different
audiences simultaneously, and particularly the possibility that (guided by the implied
author), the narrator may be addressing one kind of audience through another. But
whether the best way to keep this vital distinction in mind is to reify it into set
categories of audiences seems to me doubtful. I suspect that in the end, the sentences
The Narrative Communication Diagram
303
we write either to describe a given passage or to give a more general account of the
narratological situation they exemplify will in any case involve what entities on
the left side of the diagram have in mind with regard to audiences. If this is so,
why not simply shift the focus back on those entities and their rhetorical attempts to
imagine and move their audiences?2 And as it happens, for this purpose, an attempt to
specify different kinds of narratees and other audience entities can have a significant
heuristic value. Robyn Warhol, for instance, helps clarify what she means by the
‘‘engaging’’ narrator by outlining different possible relationships between the narrator, narratee, and reader (Warhol 1989: 25–30). In such discussions, however, I would
argue that the concept of the narratee functions most fruitfully precisely when the
insights it enables fold back into insights about the purposes of the narrator. Warhol’s
contribution (and I consider it a very significant one) is, after all, to have added to our
narratological discourse the figure, not of the engaged narratee, but of the engaging
narrator.
Let me further solidify what I have in mind here by borrowing an example adduced
by Warhol (1989: 34). When Stowe’s narrator addresses the ‘‘mothers of America’’ in
Uncle Tom’s Cabin, urging them to put themselves in the place of slave mothers, the
audience she actually has in mind includes the mothers of America, but extends
beyond them: she is addressing all possible readers through the specific narratees she
singles out, inviting them all to focus their feelings about slavery by imagining that
they are mothers, even if they are not female, or not mothers, or not even Americans
(one recalls the profound effect Uncle Tom’s Cabin had on the male workers of
England). It’s important to realize that more than one audience is at issue here, and
it is helpful toward this end to describe the ‘‘mothers of America,’’ in information
terms, as a (collective) narratee. So much is clear. But how much help could terms on
the right side of the diagram give us in describing the nonmother non-American parts
of the audience Stowe actually wishes to address? Do they constitute a second narratee
group? To pursue this possibility, we would have to specify their difference from
the first narratee group, and to do that we would soon find ourselves in the mind of
the implied author. How many different groups do we wish to specify with the term
narratee? Perhaps we should instead describe them by employing the term ‘‘implied
reader.’’ If we do so, however, that term will begin to splinter, because in fact what is
in the mind of Stowe’s implied author is a set of different audiences. For some
members of the audience, she is preaching to the converted and thus needs only to
encourage further solidarity. For others, her task is different: she needs to win them
over. To discuss such possibilities, we will again find ourselves in the mind of the
implied author. Perhaps that is where we should stay.
From the rhetoric point of view, then, there would appear to be no way to increase
the solidity of the narratee and implied reader by limiting their scope, as we can from
the information point of view; the only way to gain greater solidity and specificity is
to turn to figures who reside on the left side of the diagram, untouched by the
diffusive effects of Genette’s Law. Normally, if we’re working along these lines, we
will interest ourselves in the mind of the implied author. Most attempts to discuss the
304
Harry E. Shaw
rhetorical aims of a given passage or scene as they involve imagined audiences can
be handled by considering what the implied author has in mind; indeed, the category
of the implied author itself was invented as a way of describing the implications of
the overall rhetorical intent of a given work. But there are novels in which
the narrative presence is so fully developed that it will be necessary to shift our
attention to the narrator’s own rhetorical purposes. I have argued elsewhere that such a
situation obtains in the novels of George Eliot. Eliot sometimes creates a narrator
whom she self-consciously places in situations that seem constrained by history in
just the same way as her readers are so constrained in their everyday lives (Shaw 1999:
236–55). Eliot does this partly because she wants her narrator’s ethical judgments
to affect real-world historical readers in a way that the words of a being exempt
from such constraints simply could not. We tend to trust those who have
‘‘been there,’’ who know from the inside what the difficulties of life feel like,
more than those who have not, and for Eliot, those difficulties include being in
history.
George Eliot’s historicized narrator assumes a stance that allows her to wield ethical
authority, as a result of accepting a definite and limited position in history. We find a
similar situation in the novels of Thackeray, though what hems his narrator in is not
quite the fully imagined historicity of Eliot, but instead a sense of the ethical
situation all human beings face in a world which I think we can fairly identify as
being that of urban modernity, but which Thackeray presents in more timeless terms,
as the world of ‘‘Vanity Fair.’’ Thackeray’s narration is so multivalent and complex in
its shifts of tone that our sense of what kind of narrative entity is addressing us makes
a very great difference in how we interpret passages and even individual sentences. My
own view is that we are likely to miss what Thackeray is up to if we do not keep our
attention on what the narrator (not just the implied author) ‘‘has in mind.’’ Thackeray’s narrator is a figure who finds himself trying to mount an adequate response to a
world characterized sometimes by nobility and kindness, but more often by cruelty,
stupidity, blindness, self-absorption, and (outrageously, given the narrator’s own
feeling of inadequacy in reacting to all of this) ethical smugness. The narrator enacts
a complex attempt, sometimes self-contradictory and at best only partially successful,
at taking all this in and reacting to it in an adequate way, and by doing so tries to
goad us into seeing just where we are in the world.
It matters a great deal that it is the narrator who tries to do this, not a disembodied
implied author positioned somewhere above the ethical situation. The narrator’s
difficulties, and his authority, arise precisely from the fact that he finds himself in
the same fix as that of the characters he is describing. To be sure, he appears to
understand how the world of ‘‘Vanity Fair’’ works better than they do, and he is
probably neither as good as the best of them nor as bad as the worst. Yet in his desires,
motivations, and likely actions, he is just as ‘‘creatural’’ and limited as all of them are
– and as we his readers are as well, if only we will admit it. Consider the following
passage from Vanity Fair:
The Narrative Communication Diagram
305
We must pass over a part of Mrs. Rebecca Crawley’s biography with that lightness and
delicacy which the world demands – the moral world, that has, perhaps, no particular
objection to vice, but an insuperable repugnance to hearing vice called by its proper
name. There are things we do and know perfectly well in Vanity Fair, though we never
speak them: as the Ahrimanians worship the devil, but don’t mention him: and a polite
public will no more bear to read an authentic description of vice than a truly refined
English or American female will permit the word breeches to be pronounced in her
chaste hearing. And yet, Madam, both are walking the world before our faces every day,
without much shocking us. If you were to blush every time they went by, what
complexions you would have! It is only when their naughty names are called out
that your modesty has any occasion to show alarm or sense of outrage, and it has been
the wish of the present writer, all through this story, deferentially to submit to the
fashion at present prevailing, and only to hint at the existence of wickedness in a light,
easy, and agreeable manner, so that nobody’s fine feelings may be offended. I defy any
one to say that our Becky, who has certainly some vices, has not been presented to the
public in a perfectly genteel and inoffensive manner. In describing this syren, singing
and smiling, coaxing and cajoling, the author, with modest pride, asks his readers all
round, has he once forgotten the laws of politeness, and showed the monster’s hideous
tail above water? No! Those who like may peep down under waves that are pretty
transparent, and see it writhing and twirling, diabolically hideous and slimy, flapping
amongst bones, or curling round corpses; but above the water-line, I ask, has not
everything been proper, agreeable, and decorous, and has any the most squeamish
immoralist in Vanity Fair a right to cry fie? When, however, the syren disappears and
dives below, down among the dead men, the water of course grows turbid over her, and
it is labour lost to look into it ever so curiously. They look pretty enough when they sit
upon a rock, twanging their harps and combing their hair, and sing, and beckon to you
to come and hold the looking-glass; but when they sink into their native element,
depend on it those mermaids are about no good, and we had best not examine the
fiendish marine cannibals, reveling and feasting on their wretched pickled victims. And
so, when Becky is out of the way, be sure that she is not particularly well employed, and
that the less that is said about her doings is in fact the better.
(Thackeray [1848] 1994: 637–8)
What is in the head of the narrator in this passage would appear to include a wish not
simply to indicate Becky’s downward spiral after Rawdon discovers her with Lord
Steyne, but also and more centrally to ensure that the reader doesn’t imagine that
downward spiral too glibly. The passage condemns what Becky does and becomes, but
it also directs anger and disdain toward those who enjoy glutting their prurience on
spectacles they don’t want fully represented, the better to enjoy them in an unhealthy
half-light. When the narrator suggests that ‘‘those who like may peep down under
[the] waves’’ he catches his readers’ voyeuristic tendencies; when he adds that the
waves are ‘‘pretty transparent’’ he among other things admits to a certain complicity
on his own part in those tendencies. The element of prurient exaggeration likely to be
present in our imaginings, and the sheer absurdity of male fear of the nether parts of
females, is underlined by the parody into which the description of the ‘‘syren’’
306
Harry E. Shaw
descends, with its talk of mermaids ‘‘twanging’’ their harps, ‘‘fiendish marine’’
cannibals, and ‘‘pickled’’ (pickled?) victims. Becky is guilty, but our ways of imagining
her guilt make us, all of us including the narrator, guilty too – which is one reason
why Becky outrages us, which is not to say that certain aspects of what she does
should not outrage us.
To take in the full tonal complexity of the ‘‘syren’’ passage, I am arguing, it is useful
to impute it to the mind of an anthropomorphic narrative agent full of anger, sorrow,
pity, and incredulity at the human spectacle, and particularly at the spectacle of those
whose smug moralism would exempt themselves from its complexities and their own
complicity in human weakness. If we as readers of Vanity Fair have been engaged in
the process of imagining this sort of mind in action, we will have come to expect its
workings to be ongoing and cumulative, as the narrator pursues an endless attempt to
come to grips with life’s complexities. As a result of this expectation, it will not come
as a surprise when, in the second paragraph following the ‘‘syren’’ passage, the narrator
tells us, in quite a different tone and with a significant transformation of the
underlying imagery, that Becky’s ‘‘abattement and degradation did not take place all
at once: it was brought about by degrees, after her calamity, and after many struggles
to keep up – as a man who goes overboard hangs on to a spar whilst any hope is
left, and then flings it away and goes down, when he finds that struggling is in vain’’
(p. 638). This is a very different vision of drowning from the one in the ‘‘syren’’
passage, where women like Becky drag men down into the water, to their pickled
doom. Here it is Becky who is the victim, and the metaphoric death is real. I will add
in passing that, from our own historical vantage point, it is the more striking that at
the moment of his greatest generosity toward Becky’s plight, when he has left behind
him the scapegoated female narratees with which the ‘‘syren’’ passage opens, the
narrator dignifies that plight by imagining h