Causality and Narrative in French Fiction from Zola to Robbe-Grillet Causality and Narrative in French Fiction from Zola to Robbe-Grillet Roy Jay Nelson Ohio State University Press Columbus Copyright © 1990 by the Ohio State University Press. All rights reserved. Library of Congress Cataloging-in-Publication Data Nelson, Roy Jay. Causality and narrative in French fiction from Zola to Robbe- Grillet / Roy Jay Nelson, p. cm. Bibliography: p. Includes index. ISBN 0-8142-0504-6 (alk. paper) 1. French fiction—20th century—History and criticism. 2. French fiction—19th century—History and criticism. 3. Causation in literature. 4. Narration (Rhetoric). 5. Zola, Emile, 1840—1902— Technique. 6. Robbe-Grillet, Alain, 1922 - —Technique. I. Title. PQ637.C37N45 1989 843.009*23—dc2O 89-34065 CIP The paper in this book meets the guidelines for permanence and durability of the Committee on Production Guidelines for Book Longevity of the Council on Library Resources. © Printed in the U.S.A. 987654321 For Anita Contents Preface ix Introduction xiii Part I: Rectilinear Causality in Narrative i 1 Causality in Fiction 3 2 Causal Options in the Récit: Gide 21 3 Causal Options in Histoire and Narration: Zolas L'Assommoir 37 4 Causal Chains and Textualization: Zola's La Bête humaine 65 5 Inference, Causality, and the Levels of Narrative 88 Part II: Nonrectilinear Causal Strategies 101 6 Track and Sidetrack 103 7 Associating with Proust 122 8 Myth and Mythos 136 9 The Broken Line 149 10 Mental-Representation Fiction 174 Conclusion 208 Appendixes 227 Works Cited 233 Index 239 Preface lthough it is a primary connector in the long, linear "syntag­ matic" dimension of stories, causation in fiction has attracted only Apassing attention from theorists. Despite the illuminating pages on the subject by Genette, Todorov, Prince, and others which enrich this study, no attempt has yet been made to construct a unified theory of the causal concept in narrative. Among the many reasons for the dearth of etiological theories, two, I think, deserve mention at the outset. First, the causal dynamic remained largely invisible to the relatively static visions of literary structuralism, which dominated early theoretical exploration. Sec­ ond, well-founded contemporary doubts about the real-world truth value of the causal concept raise questions about the validity of causal statements concerning fiction. The pages that follow are organized partly in response to these prob­ lems. A brief introductory summary of philosophical views of "real-life" causation, from Hume and Kant, through Nietzsche, Bergson, and Rus­ sell, to Mackie and Anscombe, leads among other things to the observation that causation is indeed not a truth but a hypothesis, not an entity to be found in nature, but rather a useful perceptual grid in the minds of human observers. This finding leads me to postulate that the reader serves as the "human observer" of fiction, and that causation arises most basically in stories from readerly perception. Thus the principles of reader-response criticism open doors that were closed to structuralist poetics. Chapter one then evokes briefly the notions of causal "functions" in narrative as developed by modern narratologists and seeks to show how this approach, eliminating the human observer (reader) from analyses, erroneously seeks "true" logical connections between events. Yet causes are present only in the discourse of a text, in the unavoidable causal implications of language, not in the events of x • Preface the story itself. This long-established distinction between "story" and "discourse" leads me to adapt for analytical purposes Genettes three-tier model of narrative, which includes a "narration" level as well, referring to the inferred authors "production" of the text. Inductive analysis of four relatively traditional French novels serves to demonstrate how causality—or readerly inference thereof—operates on all three levels, each of which assumes a dynamic role. At this early stage of causal research, it seems both prudent and honest to derive elements of causal theory chiefly by induction, through observation of existing fiction, rather than to propose a system apparently ex nihilo and to attempt to show subsequently that real texts exemplify it. Near the end of part I, I advance a three-level provisional model of the causal dynamic in "standard" novels and seek to demonstrate that the levels of narrative are actually definable by causal operations. Part II of this study attempts to apply this matrix, no longer to traditional fiction, but to a series of French modernist narratives. These further inductive analyses suggest that the twentieth-century French novel­ ists in question have become increasingly unsure of the objective existence of real-world causality, or at least of human ability to observe it. They are less likely than traditional writers to seek to imply its existence between events of their stories, but they must nonetheless use language, with all the causal expression that entails. They employ therefore various strategies of blocking and indeterminacy to eliminate from readers' constitution of the "story" level whatever causation may remain operative in the discourse. In the most recent Robbe-Grillet text analyzed, all "real-world" referentiality is effectively eliminated, thus making it impossible for readers to draw causal inferences about the story. On the basis of strengths and limitations now apparent in the provisional three-tier model, my conclusion proposes a more complex and uncertain one as applicable within limits to both tradi­ tional and modernist texts. Thus my response to the first problem—the inability of theory to observe causality in fiction—is an approach founded upon readerly response to texts. To the second problem, I reply that causal statements are not intended to be true: they are hypotheses. But they are inferences of such power that they inform the metonymical dimension of traditional stories and require those modern texts that would prefer to banish causal connec­ tions to adopt strategies of particular sophistication. In an effort to make these pages readily accessible to all readers of Preface • xi English, I have provided translations for the substantive quotations in other languages. Unless otherwise indicated, translations are, for better or worse, my own. To Anita, for her patience, help, and unfailing support; to my Univer­ sity of Michigan colleagues, for their interest and longanimity; among whom to Marcel Muller, for his numerous useful references; and among whom especially to Ross Chambers, who read the penultimate version, for his uncommonly valuable insights, suggestions, and encouragement: my warmest thanks. Introduction There are neither causes nor effects. Linguistically we do not know how to rid ourselves of them. But that does not matter. FRIEDRICH NIETZSCHE (trans. Kaufmann & Hollingdale) Tu causes, tu causes, c'est tout ce que tu sais faire. —Queneaus "Laverdure" ausality is fundamental to human thought and activity. Our every action implies a grasp of its practical application, and elements of C its principles are inherent in all our decisions. From traffic lights to birth control devices, our precautions affirm our belief that causation is not simply a notion, but a reality. The affirmation is hardly new. The first human fire builders were obviously aware of a predictable relationship between their actions and the results thereof, and early hominids, chipping away to sharpen one stone with another, must have envisioned the potential utility for the hunt of the spearhead in the making. Why stop with human beings? Some great apes select and wield tools. And who is to say that the osprey, diving upon a shallow-swimming fish, has not chosen its prey and fixed the angle of its dive with a conscious eye to the effects of its decisions? Everything points to causality as a conatus, a virtually instinctual link between mind and matter. Yet it has eluded objective observation and description, and no philosopher to date has provided a completely workable analysis of it, nor even an entirely satisfactory definition. Undefined, it remains interesting. The coexistence of reasonable yet differing views on the subject signifies the coexistence of reasonable yet dif­ fering worlds. And in fact a multiple cosmos awaits us in novels, each a xiv • Introduction "universe" with its own harmony arising from the character of the codes which inform it. The concept and the code of causality evolved over the period here under study (approximately the 1870s to the 1970s), and with them—perhaps because of them—evolved the structure of fiction. Yet define we must, for the sake of minimal clarity. Causality is the name of the relationship between causes and their effects. {Causation desig­ nates the same thing, but it also admits of more particular application: one may speak of the causation of a specific event, as the "act of causing" it or as the complex of its causes.) But what is a cause? A primary difficulty in defining is the multiplicity and diversity of relationships subsumed in this most general of terms. The relationships we perceive between the gravita­ tional force and "falling," between the presence of certain bacteria and "disease," and between a house fire and the flight of the occupants belong to different orders of reality and thought, their very distinctness militating against the choice of a common designation. Still, a most unscientific instinct persists in telling us they indeed have something in common. Present-day philosophers still find the term meaningful enough to pursue debates about what cause signifies. But contemporary formulations, and my own definition for the purposes of this study, are best deferred, in favor of a rapid survey of selected historical philosophical opinion, which forms the context of current approaches. Not surprisingly, a common factor in post-Enlightenment theories of causality is the required presence of a human observer. David Hume launched the "modern" arguments in A Treatise of Human Nature (1739), a book whose suspiciously anthropocentric title suggests the turn etiological analysis was to take.